vehicular edge

Blockchain-based vehicular edge computing (VEC) is regarded as a promising computing paradigm that can enhance the computing capabilities of mobile vehicles while ensuring security during task offloading. However, the blockchain consensus for secure task offloading inevitably increases the communication and computation resource consumption. More importantly, the frequent handover among roadside units during the fast movement of vehicles also raises the communication cost for blockchain consensus. To address these issues, this article proposes intelligent secure task offloading and caching (ISTOC) scheme for VEC networks. Specifically, we first establish a digital twin-assisted VEC network that migrates the blockchain consensus process from the physical space to the cyber space, supporting the dynamic handover of vehicles. Correspondingly, we propose a lightweight blockchain scheme named diffused delegated Byzantine fault tolerance (d2BFT). Then, aiming at simultaneously reducing the task processing latency and improving the blockchain transaction throughput, we formulate the joint blockchain, communication, computation, and caching (B3C) optimization problem subject to task division, communication bandwidth, computing frequency, cache storage, task deadline, and blockchain stability. Due to the nonconvexity of B3C, we transform it into a Markov decision process, and propose a multiagent double actor-critic (MADAC) algorithm in light of the distributed characteristic of blockchain. Through offline training and online execution, we jointly optimize the task division, communication bandwidth, computing frequency and cache storage allocation, block size, and block generation interval for ISTOC. Experimental results show that the proposed MADAC-based ISTOC scheme can stably converge with a much higher reward than the benchmark schemes based on MADDPG, soft actor-critic, deep deterministic policy gradient, and TD3. The improvement of MADAC-ISTOC over SAC-ISTOC is more than 25.93%.

Edge computing improves the Internet of Vehicles (IoV) by offloading heavy computations from in-vehicle devices to high-capacity edge servers, typically roadside units (RSUs), to ensure rapid response times for intensive and latency-sensitive tasks. However, maintaining Quality of Service (QoS) remains challenging in dense urban settings and remote areas with limited infrastructure. To address this, we propose an software-defined networking (SDN)-driven model for uncrewed aerial vehicle (UAV)-assisted vehicular edge computing (VEC), integrating RSUs and UAVs to provide computing services and gather global network data via an SDN controller. UAVs serve as adaptable platforms for mobile-edge computing (MEC), filling gaps left by traditional MEC frameworks in areas with high vehicle density or sparse network resources. An optimal offloading mechanism, designed to minimize the Age of Information (AoI) while balancing energy consumption and rental costs, is implemented through a soft actor-critic (SAC)-based algorithm that jointly optimizes UAV trajectory, user association, and offloading decisions. Experimental results demonstrate the model’s superior performance, achieving up to 87.2% energy savings in energy-limited settings and a 50% reduction in time-sensitive scenarios, consistently outperforming traditional strategies across various task sizes.

In the digital twin mobile edge network (DTMEN), the maintenance of the vehicle twin model and vehicular task processing in the server require the support of computing resources. In addition, they are performed simultaneously. Therefore, how to allocate resources for twin maintenance and task processing under limited server resources is crucial. However, current research tends to ignore the aspect of resource competition for twin maintenance. In this study, we analyze the delays of these two affected by resource allocation under a generic DTMEN to construct the optimization problem. For this problem, we transform the problem using a Markov decision process. Meanwhile, we propose a multiagent reinforcement learning (MADRL)-based twin maintenance and task processing resource collaborative scheduling (TMTPRCS) algorithm to solve the problem. Experiments show that our proposed approach is effective in terms of resource allocation compared to other alternative algorithms.

Vehicle edge computing can effectively ensure the quality of experience for user vehicles (UVs), but road side units (RSUs) with limited resources may not be able to handle intensive tasks under high traffic conditions. In this case, worker vehicles (WVs) with idle resources can share resources to alleviate the pressure on RSUs. However, selfish WVs may be reluctant to share idle computation resources without any rewards. In addition, the optimization problems in previous research are relatively simple and cannot be applied to complex scenarios. To address the above challenges, we propose an incentive-driven partial offloading framework aiming to maximize social welfare. In particular, the computing service provider (CSP) managing RSUs first determines resource prices and offloading rates with UVs, while also determining contract terms with WVs. Then, it generates the optimal task scheduling strategy and notifies the UVs to offload tasks to the corresponding WVs. Considering that maximizing social welfare is a mixed-integer nonlinear programming (MINLP) problem, we design the hybrid proximal policy optimization (HPPO)-based task offloading and resource allocation algorithm (HORA) with a hybrid action space to directly solve the original problem. Finally, extensive simulation results show that HORA outperforms other baseline methods across various scenarios, and the contract terms meet the constraints of individual rationality (IR) and incentive compatibility (IC).

In conventional vehicular edge computing (VEC), vehicles at edge nodes often face issues such as congestion and overhead, particularly when numerous vehicles offload their tasks to a single edge node. This scenario results in heightened processing delays and increased energy consumption. Additionally, the unpredictability of the task offloading process at the edge node presents a major challenge for vehicles in determining their offloading strategies within a dynamic environment. In this paper, we propose a jointed approach for task offloading and resource allocation aimed at minimizing the overall latency and energy consumption of all vehicles. This is accomplished through task profiling, channel allocation, and resource distribution for both vehicles and roadside units (RSUs). We introduce a framework for partial task offloading and resource allocation based on a decentralized learning algorithm known as the Distributed Partial Task Offloading and Resource Allocation (DPTORA) scheme. This approach provides flexibility in task processing, allowing each vehicle to choose whether to execute its task locally, partially offload it, or distribute it among multiple RSUs using vehicle-to-infrastructure (V2I) connections in a dense network. We develop algorithms based on the Generalist Pursuit Learning Algorithm (GPLA) and the Distributed Partial Task Offloading (DPTO) scheme to effectively address the optimization problem. Additionally, we provide a sub-optimal solution with low computational complexity. Extensive simulations validate the effectiveness of our proposed scheme in decreasing time latency and energy consumption while facilitating partial task offloading and resource allocation through a decentralized learning Approach in dynamic VEC networks.

Vehicular Edge Computing (VEC) has garnered substantial attention owing to its capacity to provide ample computational resources for computation-intensive tasks. However, how to flexibly allocate computing tasks within vehicles and efficiently manage the resources consumed by tasks has emerged as a challenge. To tackle this issue, this research advances the proposition of employing an auxiliary vehicle (AV) for task offloading and introduces a novel Auxiliary Vehicle Algorithm (AVA). AVA integrates both federated learning and multi-agent reinforcement learning to fully utilize computing resources in the vehicular environment, and simultaneously achieves task delay reduction, energy consumption minimization, and task completion rate augmentation. Moreover, we establish a federated learning framework to judiciously determine the proportion of resource allocation of AV through the implementation of inventive mechanisms. Experiment results validate that our approach not only leads to the improvement of key system performance indicators, but also ensures the comprehensive exploitation of the computing resources of mobile vehicles.

Vehicular edge computing (VEC) provides an effective task offloading paradigm by pushing cloud resources to the vehicular network edges, e.g., road side units (RSUs). However, overloaded RSUs are likely to occur especially in urban aggregation areas, possibly leading to greatly compromised offloading performance. Inspired by this, this article explores this situation by introducing an unmanned aerial vehicle (UAV) to address the VEC overload problem. Specifically, we formulate a novel online UAV-assisted vehicular task offloading problem to minimize vehicular task delay under the long-term UAV energy constraint. To solve the formulated problem, we first decouple the long-term energy constraint based on the Lyapunov optimization technique. In this way, the problem can be solved in a real-time manner without requiring future information. Then, we construct a Markov chain based on Markov approximation optimization to find out the close-to-optimal UAV-assisted offloading strategies. Furthermore, we derive a mathematical analysis to rigorously demonstrate the offloading performance of the proposed algorithm. Additionally, the simulation results show that the proposed method outperforms the baselines by significantly reducing the vehicular task delay constrained by the long-term UAV energy budget under various system parameters, such as the energy budget and computation workloads.

Vehicular Edge Computing (VEC) is the transportation version of Mobile Edge Computing (MEC) in road scenarios. One key technology of VEC is task offloading, which allows vehicles to send their computation tasks to the surrounding Roadside Units (RSUs) or other vehicles for execution, thereby reducing computation delay and energy consumption. However, the existing task offloading schemes still have various gaps and face challenges that should be addressed because vehicles with time-varying trajectories need to process massive data with high complexity and diversity. In this paper, a VEC-based computation offloading model is developed with consideration of data dependency of tasks. The minimization of the average response time and average energy consumption of the system is defined as a combinatorial optimization problem. To solve this problem, we propose a Mobility-aware dependent task offloading (MESON) Scheme for urban VEC and develop a DRL-based algorithm to train the offloading strategy. To improve the training efficiency, a vehicle mobility detection algorithm is further designed to detect the communication time between vehicles and RSUs. In this way, MESON can avoid unreasonable decisions by lowering the size of the action space. Moreover, to improve the system stability and the offloading successful rate, we design a task priority determination scheme to prioritize the tasks in the waiting queue. The experimental results show that MESON is superior compared to other task offloading schemes in terms of the average response time, average system energy consumption, and offloading successful rate.

Mobile edge computing (MEC) offers promising solutions for various delay-sensitive vehicular applications by providing high-speed computing services for a large number of user vehicles simultaneously. In this paper, we investigate non-orthogonal multiple access (NOMA) assisted secure offloading for vehicular edge computing (VEC) networks in the presence of multiple malicious eavesdropper vehicles. To secure the wireless offloading from the user vehicles to the MEC server at the base station, the physical layer security (PLS) technology is leveraged, where a group of jammer vehicles is scheduled to form a NOMA cluster with each user vehicle for providing jamming signals to the eavesdropper vehicles while not interfering with the legitimate offloading of the user vehicle. We formulate a joint optimization of the transmit power, the computation resource allocation and the selection of jammer vehicles in each NOMA cluster, with the objective of minimizing the system energy consumption while subjecting to the computation delay constraint. Due to the dynamic characteristics of the wireless fading channel and the high mobility of the vehicles, the joint optimization is formulated as a Markov decision process (MDP). Therefore, we propose an asynchronous advantage actor-critic (A3C) learning algorithm-based energy-efficiency secure offloading (EESO) scheme to solve the MDP problem. Simulation results demonstrate that the agent adopting the A3C-based EESO scheme can rapidly adapt to the highly dynamic VEC networks and improve the system energy efficiency on the premise of ensuring offloading information security and low computation delay.

With the rapid development of vehicular networks, the computational capabilities and application scenarios of vehicles are becoming increasingly diverse, leading to a continuous emergence of complex computational tasks. Facing these tasks, a single vehicle node often struggles to handle them effectively; thus, it is necessary to offload tasks to other vehicles with computational resources through Vehicle-to-Vehicle (V2V) communication. However, due to the mobility of vehicles and the limitations of computing and communication resources, efficiently completing these complex computational tasks presents a significant challenge. To address this, this paper proposes an innovative optimization scheme that combines Digital Twin (DT) technology with vehicular edge computing. It constructs digital twins of vehicles through Roadside Units (RSUs) and utilizes these digital twins to optimize task offloading strategies. The scheme aims to jointly optimize transmission power, task offloading ratios, and computational resource allocation to minimize the impact of communication constraints and vehicle mobility on task completion delay. The paper models the wireless communication channel between vehicles using the Nakagami-m fading model, taking into account both transmission delay and computation delay in the overall task completion time. To solve this non-convex optimization problem, we introduce a joiSACnt optimization framework based on the Soft Actor-Critic (SAC) algorithm for efficient task allocation and dynamic transmission power adjustment. The simulation results show that the proposed scheme significantly reduces the maximum task delay and improves overall communication efficiency, particularly when compared with baseline schemes without power optimization and digital twin modules, as well as the DQN and DDPG algorithms. It demonstrates better task processing efficiency and communication performance, providing an effective solution for task handling in vehicular networks.

Vehicular edge computing enables real-time decision-making by offloading vehicular computation tasks to edge servers along roadways. This paper focuses on optimizing offloading and scheduling these tasks, with an emphasis on task prioritization to maximize task completion within deadlines while minimizing latency and energy consumption across all priority levels. We propose a prioritized Deep Q-Network (DQNP) that optimizes long-term rewards through a priority-scaled reward system for each priority level, guiding the deep reinforcement learning (DRL) agent to select optimal actions. The model dynamically adjusts task selection based on environmental conditions, such as prioritizing tasks with higher deadlines in poor channel states, ensuring balanced and efficient offloading across all priority levels. Simulation results demonstrate that DQNP outperforms existing baseline algorithms, increasing task completion by 14%, particularly for high-priority tasks, while reducing energy consumption by 8% and maintaining similar latency. Additionally, the model mitigates resource starvation for lower-priority tasks, achieving task selection rates of 27%, 32%, and 42% for low-, medium-, and high-priority tasks, with completion ratios of 88%, 87%, and 86%, respectively, reflecting balanced resource allocation across priority classes.

The increasing complexity of vehicles has led to a growing demand for in-vehicle services that rely on multiple sensors. In the Vehicular Edge Computing (VEC) paradigm, energy-efficient task scheduling is critical to achieving optimal completion time and energy consumption. Although extensive research has been conducted in this field, challenges remain in meeting the requirements of time-sensitive services and adapting to dynamic traffic environments. In this context, a novel algorithm called Multi-action and Environment-adaptive Proximal Policy Optimization algorithm (MEPPO) is designed based on the conventional PPO algorithm and then a joint task scheduling and resource allocation method is proposed based on the designed MEPPO algorithm. In specific, the method involves three core aspects. Firstly, task scheduling strategy is designed to generate task offloading decisions and priority assignment decisions for the tasks utilizing PPO algorithm, which can further reduce the completion time of service requests. Secondly, transmit power allocation scheme is designed considering the expected transmission distance among vehicles and edge servers, which can minimize transmission energy consumption by adjusting the allocated transmit power dynamically. Thirdly, the proposed MEPPO-based scheduling method can make scheduling decisions for vehicles with different numbers of tasks by manipulating the state space of the PPO algorithm, which makes the proposed method be adaptive to real-world dynamic VEC environment. At last, the effectiveness of the proposed method is demonstrated through extensive simulation and on-site experiments.

Due to the high mobility of vehicles, service migration is inevitable in vehicular edge computing (VEC) networks. Frequent service migrations incur prohibitive migration cost including the computing cost (e.g., increased computing delay) and communication cost (e.g., occupied backhaul bandwidth). Yet existing service migration schemes are usually designed without considering the impact of the computing cost. This paper considers the impact of computing and communication cost jointly, and proposes a computing and communication cost-aware service migration scheme for VEC networks (i.e., CA-migration). Taking the service delay as a QoS metric for VEC networks, this paper formulates a migration optimization problem aiming to maximize the services’ satisfaction degree of delay (i.e., the probability that the service delay is smaller than the service delay requirement), where both the communication cost and computing cost affect the services’ satisfaction degree. Since the optimization problem is a constrained non-linear integer programming problem, it is difficult to solve. Moreover, the VEC networks are highly dynamic. Thus, a fast transfer reinforcement learning (fast-TRL) method combining transfer learning and reinforcement learning is proposed to provide an adaptive service migration scheme in dynamic VEC networks. Simulation results show that compared with existing schemes, the proposed CA-migration scheme can increase the satisfaction degree by up to 30%, and needs 25% less training time to obtain the optimal service migration policy.

Vehicle edge computing (VEC) acts as an enhancement to provide low latency and low energy consumption for internet of vehicles (IoV) applications. Mobility of vehicles and load difference of roadside units (RSUs) are two important issues in VEC. The former results in task result reception failures owing to vehicles moving out of the coverage of their current RSUs; the latter leads to system performance degradation owing to load imbalance among the RSUs. They can be well solved by exploiting flexible RSU-RSU cooperation, which has not been fully studied by existing works. In this paper, we propose a novel resource management scheme for joint task offloading, computing resource allocation for vehicles and RSUs, vehicle-to-RSU transmit power allocation, and RSU-to-RSU transmission rate allocation. In our scheme, a task result can be transferred to the RSU where the vehicle is currently located, and a task can be further offloaded from a high-load RSU to a low-load RSU. To minimize the total task processing delay and energy consumption of all the vehicles, we design a twin delayed deep deterministic policy gradient (TD3)-based deep reinforcement learning (DRL) algorithm, where we embed an optimization subroutine to solve 2 sub-problems via numerical methods, thus reducing the training complexity of the algorithm. Extensive simulations are conducted in 6 different scenarios. Compared with 4 reference schemes, our scheme can reduce the total task processing cost by 17.3%-28.4%.

Vehicular edge computing (VEC) has driven the proliferation of computation-intensive and delay-sensitive vehicular services by deploying computing and energy resources at the edge. However, the exploitation of edge resources faces challenges due to unpredictable environmental dynamics and partial observability. To this end, this paper investigates the computing energy efficiency (CEE) problem in twin-timescale VEC scenarios by dynamically adjusting the offloading policy. Based upon modeling the problem as a decentralized partially observable Markov decision process (Dec-POMDP), a pair of non-cooperative and cooperative offloading solutions are proposed relying on multi-agent reinforcement learning (MARL), respectively. Specifically, the non-cooperative solution employs multi-agent independent proximal policy optimization (IPPO) to enable vehicular user equipments (VUEs) to learn their policies in a fully distributed manner without any information sharing. By contrast, the cooperative solution combines the multi-agent shared PPO with graph attention networks (MAPPO-GAT), where the relationship among agents is learned cooperatively and the historical learning experience is shared. Additionally, we compare the computational complexity and analyze the convergence. Simulation results show that in terms of the trade-off between offloading delay and offloading energy consumption, the proposed cooperative solution is superior to the non-cooperative counterpart with the cost of moderate training overhead for cooperative learning.

The rapid increase in the number of connected vehicles has led to the generation of vast amounts of data. As a significant portion of this data pertains to vehicle-to-vehicle and vehicle-to-infrastructure communications, it is predominantly generated at the edge. Considering the enormous volume of data, real-time applications, and privacy concerns, it is crucial to process the data at the edge. Neglecting the management of processing resources in vehicular edge computing (VEC) could lead to numerous challenges as a substantial number of vehicles with diverse safety, economic, and entertainment applications, along with their data processing, emerge in the near future [1]. Previous research in VEC resource allocation has primarily focused on issues such as response time and privacy preservation techniques. However, an approach that takes into account privacy-aware resource allocation based on vehicular network architecture and application requirements has not yet been proposed. In this paper, we present a privacy and latency-aware approach for allocating processing resources at the edge of the vehicular network, considering the specific requirements of different applications. Our approach involves categorizing vehicular network applications based on their processing accuracy, real-time processing needs, and privacy preservation requirements. We further divide the vehicular network edge into two parts: the user layer (OBUs) is considered for processing applications with privacy requirements, while the allocation of resources in the RSUs and cloud layer is based on the specific needs of different applications. In this study, we evaluate the quality of service based on parameters such as privacy preservation, processing cost, meeting deadlines, and result quality. Comparative analyses demonstrate that our approach enhances service quality by 55% compared to existing state-of-the-art methods.

As vehicular edge computing and networks (VECONs) emerge, Internet of vehicles (IoVs) is gaining significant advantages. To provide users with dependable services, VECONs need to offer efficient and secure task offloading services. Unfortunately, little effort has been put into addressing security issues such as the risk of privacy exposure when offloading data. Modeling and optimizing offloading dependability are affected by these issues. To address these concerns, a dependable three-layer offloading framework involving blockchain technology is presented in this paper, with fully homomorphic encryption (FHE) algorithm incorporated in order to mitigate privacy exposure risks. The framework divides tasks reasonably and introduces a novel measure of offloading dependability called cost from energy consumption, delay, and privacy exposure risk (cEDP) for constructing dependable offloading optimization. Additionally, a self-learning based dependable privacy preservation offloading (SDPPO) algorithm is presented that is capable of solving the optimal offloading problem in a decentralized manner. Simulation results show that the SDPPO algorithm outperforms its counterparts in terms of faster convergence speed, better cEDP convergence results, and improved scalability and tunability characteristics.

Benefiting from the outstanding advantages in speeding up task processing and saving energy consumption, vehicular edge computing has entered a period of rapid development. Given the sharp increase in application services, it is vital to fully utilize all available computation resources to guarantee personalized requirements from different users. Specially, a lot of idle vehicle resources can be exploited for task execution to improve the service experience. On the other hand, most works focus on the system performance and fail to guarantee diversified user demands. To this end, we propose a novel distributed collaborative computing scheme for task completion rate maximization (TCRM) in vehicular networks by taking into account both vertical and horizontal collaboration. The novelty of horizontal collaboration lies in the full use of available one-hop vehicle resources for task computing. In order to simultaneously guarantee the system-level performance and the user-level performance, TCRM aims to maximize the task completion rate while minimizing the energy consumption by intelligent resource optimization and task allocation. A TD3-based algorithm combined with the Dirichlet distribution is proposed to obtain the optimization decisions. Extensive simulations demonstrate that TCRM significantly improves performance compared to baseline algorithms.

The surge in mobile vehicles and data traffic in Vehicular Edge Computing and Networks (VECONs) requires innovative approaches for low latency, stable connectivity, and efficient resource usage in fast-moving vehicles. Existing studies have identified that utilizing digital twins (DTs) can effectively improve service quality in VECONs. However, it still faces substantial challenges posed by large-scale complex DT communications in sustaining real-time collaborative endeavors. In particular, within the dynamic VECONs, the decision regarding DT migration plays a pivotal role in sustaining the quality of services. In this paper, we propose an adaptive DT migration (ADM) algorithm to minimize the overall migration costs when DTs deliver services. Specifically, 1) We formulate ADM as a combinatorial optimization problem in VECONs, comprehensively considering communication latency and migration latency under complex DT communications, vehicular mobilities, and dynamic states of edges; 2) An ADM algorithm based on off-policy actor-critic reinforcement learning is proposed to make migration decisions. Moreover, the ADM agent employs warm-up policies to address exploration challenges in sparse state spaces; 3) Simulations based on real-world, large-scale urban vehicular mobility datasets demonstrate that our method outperforms existing algorithms by approximately 39% on average, and it can achieve results close to the optimal.

In the evolution of the Internet of vehicles (IoV), the increasing demand for vehicular computation tasks presents significant challenges, particularly in the context of constrained local computation resources and high processing delays. To mitigate these challenges, multi-access edge computing (MEC) offers a potential solution by leveraging edge servers for low-latency processing. However, it also encounters issues such as sub-channel competition and workload imbalance owing to the uneven distribution of vehicle densities. This paper introduces a novel IoV architecture that incorporates multi-task and multi-roadside unit (RSU) capabilities, enabling edge-to-edge collaboration for efficient task offloading among RSUs. The optimization problem is formulated with the objective of minimizing the overall task delay, which is further divided into two sub-problems: communication resource allocation and load balancing. Considering the non-deterministic polynomial (NP)-hard nature of these sub-problems, we propose a two-stage deep reinforcement learning-based communication resource allocation and load balancing (DRLCL) algorithm to address them sequentially. Based on realistic vehicle trajectories, comprehensive evaluation results demonstrate the superiority of the proposed algorithm in reducing system delay compared to existing state-of-the-art baselines, offering an effective approach for optimizing the performance of vehicular edge computing (VEC) networks.

With the explosion of connected devices and Internet-of-Things (IoT) services in the smart city, the challenge to meet the demands of urban computing is increasingly prominent. Recent advances in vehicle-to-everything (V2X) communications promote urban Internet-connected vehicles to become excellent candidates for computing tasks. However, due to the limited computing capacity of vehicles, conducting computation in the vehicular networks themselves is insufficient to satisfy the demands of smart city applications. Edge computing, which delivers computing tasks to edge servers (e.g., base stations, BSs, or roadside units, RSUs) with plenty of computing resources, could be a possible solution. However, the static deployment of edge servers may cause severe load unbalance among servers in both real-time communication and computation, thereby decreasing the system performance. This paper explores a two-hop vehicle-assisted edge computing network framework in which vehicles are able to offload the tasks beyond their capabilities to underloaded edge servers relaying via neighbor vehicles. According to the state of the time-varying vehicular environment and the dynamic traffic loads among RSUs, we formulate the task offloading, relay node selection, and resources allocations problem as a Markov decision process (MDP) aiming at maximizing the performance of the computation offloading capacity with the considerations of load balancing and latency constraints. We propose a deep reinforcement learning (DRL) algorithm with a DNN as Q action-value function approximator to solve this problem. Extensive simulation results reveal that the proposed scheme can significantly improve the system performance compared to other state-of-the-art algorithms.

Intelligent Transportation Systems (ITSs) are transforming the global monitoring of road safety. These systems, including vehicular networks and transportation infrastructure, are vulnerable to several security issues, which could disrupt services and potentially cause harm to the users. It is crucial to establish robust security measures to protect against evolving attacks and ensure the safe and reliable operation of ITS. Artificial Intelligence (AI)-based Intrusion Detection Systems (IDS) are mainly used to enhance the security of ITS. The adoption of AI-based techniques to secure ITS against new emerging threats has been limited due to a lack of realistic and recent data on these types of attacks ( $i.e.$ , zero-day attacks). In this context, we introduce a novel Edge-based Framework that uses Federated Learning (FL) and blockchain to secure ITS against new emerging threats. In particular, our proposed framework consists of (1) a novel distributed Edge-based architecture that allows multiple Edge nodes to securely collaborate while preserving their privacy; and (2) a decentralized and secure reputation system based on blockchain technology to maintain the reliability and trustworthiness of the FL process within the ITS; This system manages reputation data for individual nodes (such as vehicles), guaranteeing the integrity of the FL training process. Experiment results using the UNSW-NB15 dataset show that our proposed framework achieves high accuracy and F1 score (99%) in detecting new threats while ensuring the privacy and reliability of the whole ITS. These results demonstrate the effectiveness of our proposed framework in securing ITS.

Recently, integrated visible light communication (VLC) and sensing has emerged as a promising technology for vehicular edge computing (VEC). As the explosive increase of latency-sensitive and computation-intensive vehicular applications, limited on-board computing resources are unable to satisfy diversified requirements. This paper proposes a digital twin-aided virtualization VEC network architecture, and studies joint task offloading and resource allocation for integrated VLC and sensing. To determine the associations between channels and subtasks, we first propose a communication-sensing task offloading mechanism, where vehicles send communication and sensing data to edge servers (ESs) via the street lamps, and ESs adaptively assign subtasks for various channels. Then, we formulate a joint task offloading and resource allocation optimization problem by minimizing the overall latency. To tackle the non-convex problem, we decouple it into a communication subproblem and a sensing subproblem. We then propose a successive convex approximation and alternating minimization algorithm and an alternating minimization algorithm to solve the two subproblems, respectively. Finally, an overall algorithm framework is proposed. The convergence and complexity analysis indicates that the proposed algorithms are convergent and time-efficient. Numerical results verify the superiority of the proposed algorithms. Moreover, the effects of key parameters on latency performance are discussed.

With the advent of vehicular ad hoc networks (VANETs), vehicular edge computing (VEC) facilitates the execution of vehicular tasks through the Internet. In the VEC architecture, vehicles request task offloading, and a central decision center allocates resources. Effective task offloading algorithms provide optimal and equitable decisions based on objectives, such as task latency and system overhead; however, current task offloading algorithms for VEC face challenges in adapting to complex and dynamic road environments. This paper proposes a task-offloading algorithm based on deep reinforcement learning (RL) to address the challenges of task offloading in VEC. During the task offloading process, the privacy of vehicular task data may be compromised. This study introduces a novel task-offloading algorithm for VEC that employs differential privacy (DP) principles to safeguard the confidentiality of vehicular tasks during the offloading process. The proposed algorithm introduces noise in accordance with the privacy budget during the training process. The study provides a theoretical analysis of privacy, and experimental results based on Attari demonstrate that the proposed DP-based RL algorithm exhibits superior convergence compared to existing algorithms. Veins-based simulation experiments on VEC demonstrate that the proposed DP-based task-offloading algorithm can achieve practical offloading while preserving privacy.

Advanced in the proliferation of the Internet of Things (IoT), a plethora of functions have been integrated in vehicular networks and thereby transfered it into a smart network. However, the contradiction between the limited on-vehicle computing resource and the massive data collected by these IoT devices hinders the broader adoption of vehicular network as a vast variety of on-vehicle applications are latency-sensitive. To address this issue, vehicular edge computing has become a promising technology as it can offload a large number of tasks from its proximal vehicles. However, the offloading methods recently utilized are inefficient while dealing with multi-user vehicular networks under dynamic scenarios. To design a superior offloading method that can effectively and efficiently offload tasks from vehicles to servers, multiple objectives and constraints with various topologies should be considered. In this paper, instead of constructing a typical multi-user and multi-server vehicular edge computing scenario, a complex scenario with more uncertainties, i.e. urban scenario, is modeled. We propose a Hybrid Architecture Matching Algorithm (HAMA) to minimize the average time latency subject to the constraint on energy consumption and evaluate the proposed algorithm in the above two scenarios. Moreover, HAMA is constructed based on hybrid centralized-distributed architecture, which can process the centralized collected information on a distributed manner. Experimental results demonstrate that the matching algorithm can significantly reduce average time latency, achieving up to a 68% improvement compared to local execution.

In this correspondence, we consider vehicular edge computing systems in which computing tasks arrive randomly to vehicles over time and are offloaded to one of the edge servers. Unlike previous work that assumes ideal and global server information, we study a practical online framework for acquiring edge server information. We propose an online offloading and scheduling algorithm that efficiently integrates the Moore–Hodgson algorithm into this practical framework to satisfy hard deadlines required by tasks. Numerical experiments demonstrate that the proposed algorithm outperforms conventional first-come-first-serve scheduling and D-Dedas scheduling. Furthermore, the results show that the performance of proactive broadcasting of server information closely approaches that of ideal server information as the number of vehicles increases.

Modern vehicles have become typical consumer electronics with the development of sensing, transmission, and computation technologies. The emerging intelligent transportation systems (ITSs) yield lots of deep neural network (DNN) based tasks, requiring intensive computation. In view of this, this paper makes the first exploration of accelerating the processing of DNN-based tasks while maintaining decent system reliability in vehicular edge computing (VEC) via task partitioning and offloading. Specifically, we present a specific scenario, where vehicles partition and offload their tasks to nearby vehicles and roadside infrastructures, while they may fail to receive results due to unstable vehicular communications. Then, we model the task delay by considering task properties and node capacities. On this basis, the Partitioning and Offloading Problem (POP) is formulated as a bi-objective optimization problem, to maximize both the acceleration ratio and service ratio of DNN-based tasks in VEC. Further, we propose a Distributed Partitioning and Offloading Solution (DPOS), where a delay-priority-oriented offloading strategy is designed to help edge nodes make offloading decisions, and a stacking-based partitioning strategy is designed to assist client vehicles to make partitioning decisions. Finally, we give a comprehensive performance evaluation, which demonstrates the superiority of the proposed solution.

Due to the flexible deployment and availability of line-of-sight (LoS) link, unmanned aerial vehicle (UAV) is able to assist the roadside unit (RSU) to provide timely computing resources to the covered vehicle users in the temporary congestion roads. In UAV-assisted vehicular edge computing networks (VECNs), caching necessary data in RSU/UAV reduces task execution delay and bandwidth cost significantly. However, for the constrained storage and computation capacities of UAV and the dynamic requests of users, the efficient caching data selection and replacement schemes are needed. In this article, we propose a novel joint hybrid caching and replacement scheme in a scene that a single UAV assists RSU to cover a set of vehicle users with large number of iterative calculation tasks. In particular, both the content caching and service caching are considered for the RSU and UAV. To minimize the whole task completion delay of users, we joint optimize the hybrid caching data selection of UAV and the task offloading strategy of users, a deep Q-network (DQN)-based solution is proposed to improve the utility of UAV. Then, we design an optimal evaluation function, and propose a caching replacement scheme for both the RSU and UAV. For RSU and UAV can update its caching data separately, a double-DQN (DDQN)-based solution is proposed. Extensive simulation results show that the proposed algorithms have good convergence, and the designed schemes reduce the task completion delay efficiently.

In this paper, we present a framework that integrates digital twin (DT) technology into space-air-ground integrated networks (SAGINs) to enhance vehicular edge computing (VEC). Our objective is to efficiently offload tasks in ultra-reliable low-latency communications (URLLC)-enabled vehicular networks, focusing on minimizing overall latency for requested tasks by reducing transmission time for task offloading and edge processing requirements. The proposed framework leverages DT-assisted SAGINs to minimize task offloading latency, expand network coverage, and reduce energy consumption. Key components of our framework include partial task offloading, distributed edge computing, latency modeling, energy consumption analysis, mobility, and channel modeling. We formulate a non-convex optimization problem considering various network constraints to achieve the system objective. To solve this optimization problem, we develop a novel multi-agent deep reinforcement learning (DRL) algorithm, enabling intelligent decision-making by individual agents. Through extensive simulations, we validate the effectiveness of our proposed system in advancing VEC by integrating DT technology into SAGINs.

Mobile edge computing has been a promising solution to enable real-time service in vehicular networks. However, due to high dynamics of mobile environment and heterogeneous features of vehicular services, traditional expert-based or learning-based strategies has to update handcrafted parameters or retrain learning model, which leads to intolerant overhead. Therefore, this paper investigates the problem of multi-task offloading (MTO), where there exist multiple offloading scenarios with varying parameters, such as task topology, resource requirement and transmission/computation capability. The objective is to design a unified solution to minimize task execution time under different MTO scenarios. Accordingly, we develop a Seq2seq-based Meta Reinforcement Learning algorithm for MTO (SMRL-MTO). Specifically, a bidirectional gated recurrent units integrated with attention mechanism is designed to determine offloading action by encoding sequential offloading actions and showing different preferences to different parts of input sequence. Particularly, a meta reinforcement learning framework is designed based on model-agnostic meta learning, which trains a meta policy offline and fast adapts to new MTO scenario within a few training steps. Finally, we conduct performance evaluation based on task generator DAGGEN and realistic vehicular traces, which shows that the SMRL-MTO reduces task execution time by 11.36% on average compared with greedy algorithm.

This paper introduces a novel approach to Vehicle Edge Computing (VEC), addressing the need for low-latency, high-reliability applications in vehicle networks. By leveraging nearby Multi-access Edge Computing (MEC) resources, VEC enhances data processing speed and reliability for applications like autonomous driving, real-time traffic management, and infotainment systems. The proposed solution models a multi-user non-cooperative computation offloading game in vehicular MEC networks, where each vehicle adjusts its offloading probability based on factors like distance to the MEC access point, communication model, and competition for resources. Additionally, a best response-learning algorithm is designed based on the computation offloading game model. The approach focuses on maximizing each vehicle’s utility while ensuring convergence to a single, stable equilibrium under defined conditions. To demonstrate the effectiveness and performance of the proposed algorithms, comprehensive experiments were performed. Numerical results demonstrate the fast convergence and improved performance achieved.

In an internet of vehicle (IoV) scenario, vehicular edge computing (VEC) exploits the computing capabilities of the vehicles and roadside unit (RSU) to enhance the task processing capabilities of the vehicles. Resource management is essential to the performance improvement of the VEC system. In this paper, we propose a joint task offloading and resource allocation scheme to minimize the total task processing delay of all the vehicles through task scheduling, channel allocation, and computing resource allocation for the vehicles and RSU. Different from the existing works, our scheme: 1) considers task diversity by profiling the tasks of the vehicles by multiple attributes including data size, computation amount, delay tolerance, and task type; 2) considers vehicle classification by dividing the vehicles into 4 sets according to whether they have task offloading requirements or provide task processing services; 3) considers task processing flexibility by deciding for each vehicle to process its tasks locally, to offload the tasks to the RSU via V2I (Vehicle to Infrastructure) connections, or to the other vehicles via V2V (Vehicle to Vehicle) connections. An algorithm based on the Generalized Benders Decomposition (GBD) and Reformulation Linearization (RL) methods is designed to optimally solve the optimization problem. A heuristic algorithm is also designed to provide the sub-optimal solution with low computational complexity. We analyze the convergence and complexity of the proposed algorithms and conduct extensive simulations in 6 scenarios. The simulation results demonstrate the superiority of our scheme in comparison with 4 other schemes.

Vehicular Edge Computing (VEC) has gained increasing interest due to its potential to provide low latency and reduce the load in backhaul networks. In order to meet drastically increasing computation demands from emerging ever-growing vehicular applications, e.g., autonomous driving, abundant computation resources of individual vehicles can play a crucial role in task execution in a VEC scenario, that can further contribute in considerably improving user experience. This is however an extremely challenging task due to high mobility of vehicles that can easily lead to intermittent connectivity, thereby disrupting on-going task processing. In this paper, we propose a task offloading scheme by exploiting multi-hop vehicle computation resources in VEC based on mobility analysis of vehicles. In addition to the vehicles within one hop from the task vehicle that generates computation tasks, certain multi-hop vehicles that meet the given requirements in terms of link connectivity and computation capacity, are also leveraged to carry out the tasks offloaded by the task vehicle. An optimization problem is formulated for the task vehicle to minimize the weighted sum of execution time and computation cost of all tasks. A semidefinite relaxation approach with an adaptive adjustment procedure is proposed to solve the formulated optimization problem for obtaining the corresponding offloading decisions. The simulation results show that our proposed offloading scheme can achieve significant improvement in terms of response delay by at least 34% compared with the other algorithms (e.g., local processing and random offloading).

Vehicular Edge Computing (VEC) is enjoying a surge in research interest due to the remarkable potential to reduce response delay and alleviate bandwidth pressure. Facing the ever-growing service applications in VEC, how to effectively aggregate and flexibly schedule ubiquitous network resources for implementing diverse tasks and meeting differentiated demands from numerous vehicular users remains haunting. Toward this end, we investigate collaborative task computing and on-demand resource allocation. The collaborative computing framework in VEC is provided to support deep collaboration and intelligent management of heterogeneous resources widely distributed in vehicles, edge servers and cloud. Based on this framework, the joint optimization problem of distributed task offloading and multi-resource management is formulated with the aim to maximize the system utility by making the optimal task and resource scheduling policy, the novelty of which lies in the exploration of available vehicle resources and the consideration of service migration. In view of the dynamics, randomness and time-variant of vehicular networks, the asynchronous deep reinforcement algorithm is leveraged to find the optimal solution. Extensive simulation experiments are implemented to demonstrate the superiority of our proposed algorithm in terms of response latency compared with full offloading and random offloading.

With the rapid advancement in technology, numerous advanced vehicular applications have emerged that generate large volumes of data that need to be processed on the fly. The vehicles' computing resources are limited and constrained in processing the huge amount of data generated by these applications. Cloud data centers, which are large and capable of processing the generated data, tend to be far away from the vehicles. The long distance between the cloud and the vehicles results in large transmission delays, making the cloud less suitable for executing such data. To address the long-standing issue of huge transmission delays in the cloud, edge computing, which deploys computing servers at the edge of the network, was introduced. The edge computing network shortens the communication distance between the vehicles and the processing resources and also provides more powerful computation compared to the vehicles' computing resources. The advantages offered by the vehicular edge network can only be fully realized with robust and efficient resource allocation. Poor allocation of these resources can lead to a worse situation than the cloud. In this paper, a hybrid Marine Predatory and Particle Swarm Optimization Algorithm (MPA–PSO) is proposed for optimal resource allocation. The MPA–PSO algorithm takes advantage of the effectiveness and reliability of the global and local search abilities of the Particle Swarm Optimization Algorithm (PSO) to improve the suboptimal global search ability of the MPA. This enhances the other steps in the MPA to ensure an optimal solution. The proposed MPA–PSO algorithm was implemented using MATLAB alongside the conventional PSO and MPA, and the proposed MPA–PSO recorded a significant improvement over the PSO and MPA.

Vehicular edge computing (VEC) effectively reduces the computational burden on vehicles by offloading tasks from resource-constrained vehicles to edge nodes. However, non-uniformly distributed vehicles offloading a large number of tasks cause load imbalance problems among edge nodes, resulting in performance degradation. In this paper, we propose a deep reinforcement learning-based decision scheme for task offloading and load balancing with the optimization objective of minimizing the system cost considering the split offloading of tasks and the load dynamics of edge nodes. First, we model the mutual interaction between mobile vehicles and Mobile Edge Computing (MEC) servers using a Markov decision process. Second, the optimal task-offloading and resource allocation decision is obtained by utilizing the twin delayed deep deterministic policy gradient algorithm (TD3), and server load balancing is achieved through edge collaboration using a server selection algorithm based on the technique for order preference by similarity to the ideal solution (TOPSIS). Finally, we have conducted extensive simulation experiments and compared the results with several other baseline schemes. The proposed scheme can more effectively reduce the system cost and increase the system resource utilization.

In the Internet of Vehicles, a variety of vehicle applications provide convenience for people to drive and travel. Among the tasks generated by these applications, some are computing-intensive and delay-sensitive, and need data caching to complete the task. Against this background, our paper considers a MEC-assisted vehicular network where vehicles need to complete the tasks during driving. Through jointly optimizing the offloading decision, MEC server computing resource allocation and data caching decision, we propose an optimization problem to minimize the total cost of vehicles, which can improve the quality of service (QOS). Because the optimization problem is a Mixed-Integer Non-Linear Programming (MINLP) problem, we divide it into two subproblems. We propose an algorithm JOMAD to solve the MINLP problem, mainly using discrete Binary Particle Swarm Optimization (BPSO) to solve two subproblems respectively. Simulation results demonstrate that the proposed scheme achieves less cost than other benchmarks.

No abstract available

No abstract available

No abstract available

Leveraging the abundance of computational resources, the cloud-edge collaborative architecture provide stronger data processing capabilities for vehicular networks, which not only significantly enhances the timeliness of offloading operations for delay-sensitive tasks but also substantially mitigates resource expenditure associated with non-delay-sensitive tasks. Addressing the communication scenarios characterized by diverse task types, this paper introduces cloud-assisted mobile-edge computing (C-MEC) networks, underscored by a novel optimization scheme. The scheme incorporates a utility function that is correlated with offloading delays during the transmission and computation phases, effectively balancing resource allocations and enhancing the operational efficiency of vehicular networks. However, the mobility of vehicles introduces channel uncertainty, adversely affecting the offloading stability of C-MEC networks. To develop a practical channel model, a first-order Markov process is employed, taking into account vehicular mobility. Additionally, probability constraints regarding co-channel interference are imposed on signal links to ensure the offloading quality. The Bernstein approximation method is utilized to transform the original interference constraints into a tractable form, and the Successive Convex Approximation (SCA) technique is meticulously applied to address the non-convex robust optimization problem. Furthermore, this paper proposes a robust iterative algorithm to ascertain optimal power control and task scheduling strategies. Numerical simulations are conducted to assess the effective of the proposed algorithm against benchmark methods, with a particular focus on robustness in task offloading and utility in resource allocation.

Numerous applications of vehicles are computation- intensive and delay-sensitive. In order to deal with the problem caused by limited wireless and computation capability in the Mo- bile Edge Computing (MEC) enabled vehicular network, a Joint Optimization of Wireless and Computation Allocation (JOWCA) algorithm is proposed to minimize global delay of MEC-enabled vehicular network. The JOWCA algorithm consists of vehicle- to-vehicle (V2X) matching and MEC computation capability allocation. In the V2X matching, a Graph-based Interference Cancellation (Graph-IC) scheme is proposed to allocate Resource Blocks (RBs) for vehicle-to- infrastructure (V2I) links and vehicle- to-vehicle (V2V) links to mitigate co-channel interference. The Graph-IC contains an adaptive interference threshold modified Heuristic Clustering (HC) algorithm and Hungarian algorithm. In the MEC computation capability allocation, the optimal solution of V2I link offloading ratio and MEC computation capability scheduling are obtained by applying Karush-Kuhn- Tucker (KKT) condition. Simulation shows that the proposed scheme can effectively reduce the global delay of the MEC- enabled vehicular network.

By providing storage and computational resources at the network edge, which enables hosting applications closer to the mobile users, Multi-Access Edge Computing (MEC) uses the mobile backhaul, and the network core more efficiently, thereby reducing the overall latency. Fostering the synergy between 5G and MEC brings ultra-reliable low-latency in data transmission, and paves the way towards numerous latency-sensitive automotive use cases, with the ultimate goal of enabling autonomous driving. Despite the benefits of significant latency reduction, bringing MEC platforms into 5G-based vehicular networks imposes severe challenges towards poorly scalable network management, as MEC platforms usually represent a highly heterogeneous environment. Therefore, there is a strong need to perform network management and orchestration in an automated way, which, being supported by Software Defined Networking (SDN) and Network Function Virtualization (NFV), will further decrease the latency. With recent advances in SDN, along with NFV, which aim to facilitate management automation for tackling delay issues in vehicular communications, we study the closed-loop life-cycle management of network services, and map such cycle to the Management and Orchestration (MANO) systems, such as ETSI NFV MANO. In this paper, we provide a comprehensive overview of existing MANO solutions, studying their most important features to enable network service and resource orchestration in MEC-enhanced vehicular networks. Finally, using a real testbed setup, we conduct and present an extensive performance analysis of Open Baton and Open Source MANO that are, due to their lightweight resource footprint, and compliance to ETSI standards, suitable solutions for resource and service management and orchestration within the network edge.

The Internet of Vehicles (IoV) represents a paradigm shift in vehicular communication, aiming to enhance traffic efficiency, safety, and the driving experience by leveraging interconnected vehicles. Despite its promise, the IoV faces challenges such as efficient task offloading, energy management, and data security. Mobile Edge Computing (MEC) emerges as a solution to some of these challenges by bringing computational resources closer to the vehicular network’s edge, yet it raises critical concerns regarding resource management, service continuity, and scalability in dynamic vehicular environments. Addressing both IoV and MEC challenges necessitates robust and dynamic optimization mechanisms. In response to these challenges, our study introduces a multi-objective approach using Double Deep Q-Networks (DDQN). This algorithm combines the strengths of Deep Neural Networks (DNNs) and Deep Learning (DL) techniques, enabling dynamic decision-making that can adapt to changing conditions. By considering multiple objectives, the DDQN algorithm allows for a sophisticated trade-off analysis, efficiently balancing between the different objectives to optimize overall system performance. Through the use of Blockchain technology, known for its secure, decentralized structure, our model enhances the integrity of data, providing a reliable and efficient solution for IoV-MEC systems. We conducted a comparative analysis of our model against the standard Deep Q-Network (DQN) and Deep Deterministic Policy Gradient (DDPG) algorithms, which are prevalent in this field. Our model demonstrated significant improvements over these traditional methods: energy consumption was reduced by 26.4%, latency decreased by 6.87%, and the cost was minimized by 7.41%.

In the contemporary landscape of vehicular networks, characterized by disruptive innovations such as autonomous driving, real-time traffic monitoring, and integrated infotainment services, there is a greater need for substantial computational power and precise operations. To cope with such demand, mobile edge computing (MEC) emerges as a pivotal mechanism that can boost the capacities of vehicular networks, leveraging edge servers, offering ultralow latency, and providing support for operations requiring near real-time access to rapidly growing and varying data. In this paper, we exploit deep reinforcement learning (DRL) techniques to minimize the overall service delay of the vehicle in a vehicular network using MEC. We propose a dual-stage DRL-assisted resource optimization strategy; firstly, the strategy optimizes the transmit power of the vehicle using the deep deterministic policy gradient (DDPG) algorithm after the vehicle initiates computational offloading, which contributes to minimizing the propagation delay. Secondly, once the data are transferred to the roadside unit (RSU), the strategy uses the deep Q-network (DQN) algorithm to allocate an optimal number of processing cores for computational tasks, thus minimizing the computational delay. Additionally, we present the energy efficiency (EE) and spectral efficiency (SE) analysis of the vehicle to examine the trade-off between the EE and SE. We compare the performance of the proposed strategy against the benchmark schemes, including always offload, never offload, random offload, and deep neural network (DNN)–based cooperative offloading schemes. The simulation results demonstrate that the proposed resource allocation strategy outperforms the benchmark schemes and improves the system performance by reducing the overall service delay.

Mobile Edge Computing (MEC)-enabled vehicular networks have emerged as a promising approach to enhancing the performance and efficiency of the Internet-of-Vehicles (IoV) applications. By leveraging some vehicles to act as transmission relays, multi-hop task offloading addresses the problem of intermittent connectivity between vehicles and edge servers to cope with the issues of network congestion or obstacles. However, two critical issues, i.e., uncooperative behaviors of selfish vehicles and network resource dynamics, resulting from multi-vehicle concurrent offloading are not fully considered in the existing work. To fill this gap, this paper proposes a novel and efficient task offloading scheme, namely MMTO, that exploits the multi-hop computational resources to maximize the system-wide profit, and supports incentive compatibility of vehicular users and concurrent offloading. Specifically, an iterative hierarchical estimation algorithm is designed to estimate the offloading delay and energy cost in order to iteratively optimize the offloading decisions. An energy-efficient routing approach is then proposed to schedule the transmission paths for the offloading vehicles. Furthermore, an effective reward-driven auction-based incentive mechanism is designed for incentivizing relayers and calculators to engage in collaboration. Both simulation and field experiments are conducted; extensive results demonstrate that MMTO outperforms the state-of-the-art approaches in terms of the system-wide profit improvement and overall task delay reduction.

The rise of modern vehicles and intelligent transportation systems has led to an exponential increase in vehicular data, especially from Controller Area Networks (CAN). While cloud infrastructures are commonly employed for storing and processing such data, transmitting high-frequency CAN messages from millions of vehicles poses a significant challenge to centralized systems, particularly during peak usage hours. This paper explores the potential of Multi-access Edge Computing (MEC) in conjunction with Kubernetes orchestration to alleviate the computational and network burden on central cloud infrastructures. Using real-world and publicly available CAN datasets, we analyze the scale of vehicular data generation and assess traffic density in urban settings to estimate the cumulative data load. A Kubernetes-based architectural framework is proposed and deployed on MEC nodes within the 5GTN network to collect, losslessly compress, and forward CAN data efficiently. Experimental results indicate that the proposed solution significantly reduces cloud resource consumption, specifically bandwidth usage by 17%, network packet and CPU usage by 99% and network activity by 100% when MECs are utilized as delegates, while ensuring scalable, secure, and reliable data collection. This study demonstrates the viability of leveraging MEC and a container orchestration technology for sustainable, distributed vehicular data management.

In vehicular communication and mobile edge computing (MEC) networks, limited storage resources and challenges related to vehicles data security pose significant concerns. To address these issues while reducing communication latency and enhancing network security, blockchain technology is introduced. Additionally, deep reinforcement learning (DRL) is leveraged to optimize content caching strategies. By formulating a Markov decision process (MDP) model and applying the proximal policy optimization (PPO) algorithm, efficient cache management and optimal resource allocation are achieved. Simulation results demonstrate that the proposed approach effectively improves cache hit rates and significantly reduces latency in vehicular communication environments, yielding superior performance.

The emergence of 5G networks has enabled the deployment of a two-tier edge and vehicular-fog network. It comprises Multi-access Edge Computing (MEC) and Vehicular-Fogs (VFs), strategically positioned closer to Internet of Things (IoT) devices, reducing propagation latency compared to cloud-based solutions and ensuring satisfactory quality of service (QoS). However, during high-traffic events like concerts or athletic contests, MEC sites may face congestion and become overloaded. Utilizing offloading techniques, we can transfer computationally intensive tasks from resource-constrained devices to those with sufficient capacity, for accelerating tasks and extending device battery life. In this research, we consider offloading within a two-tier MEC and VF architecture, involving offloading from MEC to MEC and from MEC to VF. The primary objective is to minimize the average system cost, considering both latency and energy consumption. To achieve this goal, we formulate a multi-objective optimization problem aimed at minimizing latency and energy while considering given resource constraints. To facilitate decision-making for nearly optimal computational offloading, we design an equivalent reinforcement learning environment that accurately represents the network architecture and the formulated problem. To accomplish this, we propose a Distributed-TD3 (DTD3) approach, which builds on the TD3 algorithm. Extensive simulations, demonstrate that our strategy achieves faster convergence and higher efficiency compared to other benchmark solutions.

Increasing demands for Quality of Experience (QoE) lead to massive connectivity and intensive computation in future vehicular networks. This article proposes a device-to-device (D2D)-based mobile edge computing (MEC) network architecture to provide effective communication connections and sufficient computing abilities for vehicular networks. However, the available communication and computing resources are limited in the D2D-based vehicular MEC networks, and an imbalanced resource allocation always leads to suboptimal optimization of overall performances. To address this challenge, we formulate a Lyapunov optimization method-based resource allocation framework to balance communication and computing by compromising energy efficiency (EE) and time delay. However, the long-term resource allocation framework is ineffective when it ignores the dynamic characteristics of vehicular networks, i.e., channel state changes due to the movement of vehicles and a dynamic queue backlog with data fluctuations. Considering the time-varying channel state and dynamic queue backlog, the proposed framework aims to balance resource allocations while primarily maintaining network stability. Finally, we propose a Lyapunov optimization-based long-term dynamic resource allocation algorithm to develop real-time allocation strategies. Simulation results illustrate that the proposed algorithm balances communication and computing resources by tuning the control parameter V. Furthermore, the results confirm that the proposed algorithm outperforms baseline algorithms in real-time transmission and offloading ability.

Vehicular cloud computing is gaining popularity thanks to the rapid advancements in next generation wireless communication networks. Similarly, Edge Computing, along with its standard proposals such as European Telecommunications Standards Institute (ETSI) Multi-access Edge Computing (MEC), will play a vital role in these scenarios, by enabling the execution of cloud-based services at the edge of the network. Together, these solutions have the potential to create real micro-datacenters at the network edge, favoring several benefits like minimal latency, real-time data processing, and data locality. However, the research community has not yet the opportunity to use integrated simulation frameworks for the easy testing of applications that exploit both the vehicular cloud paradigm and MEC-compliant 5G deployment environments. In this paper, we present our simulation tool as a platform for researchers and engineers to design, test, and enhance applications utilizing the concepts of vehicular and edge cloud. Our platform significantly extends OMNet++ and Simu5G, and implements our ETSI MEC-compliant architecture that leverages resources provided by far-edge nodes. In addition, the paper analyzes and reports performance results for our simulation platform, as well as provides a use case where our simulator is used to support the design, test, and validation of an algorithm to distribute MEC application components on vehicular cloud resources.

Multi-access Edge Computing (MEC) leverages cloud computing capabilities at the network's edge, giving underpowered devices the ability to offload tasks to more robust edge nodes. This distributed architecture improves latency and enhances bandwidth, which is crucial for edge-deployed applications, particularly in dynamic vehicular environments, where effective orchestration is essential due to continuous vehicle and device mobility. In this paper, we delve into Mobility Management within the MEC environment and introduce a novel architecture that enhances the acquisition and use of mobility data for improved system-level orchestration. With increasing support from the mobility data, different orchestration strategies were established and evaluated, specifically regarding their impact on application resource usage and response time within a simulated vehicular scenario. Results show the importance of precise mobility data in MEC orchestration, emphasising its role in improving network performance and edge resource consumption.

No abstract available

Task offloading is of paramount importance to efficiently orchestrate vehicular wireless networks, necessitating the availability of information regarding the current network status and computational resources. However, due to the mobility of the vehicles and the limited computational resources for performing task offloading in near-real-time, such schemes may require high latency, thus, become even infeasible. To address this issue, in this paper, we present a Trajectory Prediction-based Pre-offloading Decision (TPPD) algorithm for analyzing the historical trajectories of vehicles to predict their future coordinates, thereby allowing for computational resource allocation in advance. We first utilize the Long Short-Term Memory (LSTM) network model to predict each vehicle’s movement trajectory. Then, based on the task requirements and the predicted trajectories, we devise a dynamic resource allocation algorithm using a Double Deep Q-Network (DDQN) that enables the edge server to minimize task processing delay, while ensuring effective utilization of the available computational resources. Our simulation results verify the effectiveness of the proposed approach, showcasing that, as compared with traditional real-time task offloading strategies, the proposed TPPD algorithm significantly reduces task processing delay while improving resource utilization.

Beam selection for millimeter-wave links in a vehicular scenario is a challenging problem, as an exhaustive search among all candidate beam pairs cannot be assuredly completed within short contact times. We solve this problem via a novel expediting beam selection by leveraging multimodal data collected from sensors like LiDAR, camera images, and GPS. We propose individual modality and distributed fusion-based deep learning (F-DL) architectures that can execute locally as well as at a mobile edge computing center (MEC), with a study on associated tradeoffs. We also formulate and solve an optimization problem that considers practical beam-searching, MEC processing and sensor-to-MEC data delivery latency overheads for determining the output dimensions of the above F-DL architectures. Results from extensive evaluations conducted on publicly available synthetic and home-grown real-world datasets reveal 95% and 96% improvement in beam selection speed over classical RF-only beam sweeping, respectively. F-DL also outperforms the state-of-the-art techniques by 20-22% in predicting top-10 best beam pairs.

Communication and computation services supporting Connected and Automated Vehicles (CAVs) are characterized by stringent requirements, in terms of response time and relia-bility. Fulfilling these requirements is crucial for ensuring road safety and traffic optimization. The conceptually simple solution of hosting these services in the vehicles increases their cost (mainly due to the installation and maintenance of computation infrastructure) and may drain their battery excessively. Such disadvantages can be tackled via Multi-Access Edge Computing (MEC), consisting in deploying computation capability in network nodes deployed close to the devices (vehicles in this case), such as to satisfy the stringent CAV requirements. However, it is not yet clear under which conditions MEC can support CAV requirements and for which services. To shed light on this question, we conduct a simulation campaign using well-known open-source simulation tools, namely OMNeT++, Simu5G, Veins, INET, and SUMO. We are thus able to provide a reality check on MEC for CAV, pinpointing what are the computation capacities that must be installed in the MEC, to support the different services, and the amount of vehicles that a single MEC node can support. We find that such parameters must vary a lot, depending on the service considered. This study can serve as a preliminary basis for network operators to plan future deployment of MEC to support CAV.

The mobile edge computing (MEC) technology can simultaneously provide high-speed computing services for multiple vehicular users (VUs) in vehicular edge computing (VEC) networks. Nevertheless, due to the open feature of the wireless offloading channels and the high mobility of the vehicles, the security and stability of the offloading process would be seriously degraded. In this paper, by utilizing the physical layer security (PLS) technique and spectrum sharing architecture, we propose a deep reinforcement learning based joint secure offloading and resource allocation (SORA) scheme to improve the secrecy performance and resource efficiency of the multi-user VEC networks, where the VU offloading links share the frequency spectrum preoccupied with the vehicle-to-vehicle (V2V) communication links. We use Wyner’s wiretap coding scheme to obtain the achievable secrecy rate and guarantee that confidential information cannot be decoded by multiple mobile eavesdroppers. We aim at minimizing the system processing delay while securing the wireless offloading process, by jointly optimizing the transmit power, the frequency spectrum selection and the computation resource allocation. We formulate the optimization problem as a multi-agent collaborative optimal decision problem and solve it with a double deep Q-learning algorithm. Besides, we set a punishment mechanism for the rate degradation to guarantee the communication quality of each V2V link. Simulation results demonstrate that multiple VU agents adopting the SORA scheme can rapidly adapt to the highly dynamic VEC networks and cooperate to improve the system delay performance while increasing the secrecy probability.

Video surveillance in intelligent transportation systems is advancing rapidly, with video analytics technology being used to enhance the security of the Internet of Vehicles (IoV) system. However, the sheer volume of video data from cameras and the computational intensity of video analysis pose significant challenges to the IoV network. To address this, mobile edge computing (MEC) has been introduced to offload video tasks from cameras to mobile edge servers/groups formed by vehicles. However, the resource-constrained nature of edge servers and vehicle groups necessitates the design of effective offloading strategies. Additionally, ensuring the security of user data during transmission and computation is a pressing issue. Moreover, the heterogeneous devices in the IoV system may be reluctant to participate in the collaborative processing of video tasks due to mistrust and lack of incentives. To tackle these challenges, we propose a cooperative computing offloading and resource allocation framework that integrates blockchain and MEC to provide secure and low-latency computing offloading services for the IoV system. We also design an efficient incentive mechanism to promote the collaborative processing of video tasks. Our framework formulates computing offloading and resource allocation as a joint optimization problem to maximize the system revenue, and we propose an algorithm based on the alternating direction method of multipliers (ADMM) to solve the distributed optimization problem with fast convergence and low complexity. Simulation results demonstrate that compared to the typical baselines, our scheme can achieve the maximum system revenue and effectively reduce the system delay.

Multiaccess edge computing (MEC) has emerged as a promising technology for time-sensitive and computation-intensive tasks. However, user mobility, particularly in vehicular networks, and limited coverage of Edge Server result in service interruptions and a decrease in Quality of Service (QoS). Service migration has the potential to effectively resolve this issue. In this paper, we investigate the problem of service migration in a MEC-enabled vehicular network to minimize the total service latency and migration cost. To this end, we formulate the service migration problem as a Markov decision process (MDP). We present novel contributions by providing optimal adaptive migration strategies which consider vehicle mobility, server load, and different service profiles. We solve the problem using the Double Deep Q-network algorithm (DDQN). Simulation results show that the proposed DDQN scheme achieves a better tradeoff between latency and migration cost compared with other approaches.

As important services of the future sixth-generation (6G) wireless networks, vehicular communication and mobile edge computing (MEC) have received considerable interest in recent years for their significant potential applications in intelligent transportation systems. However, MEC-enabled vehicular networks depend heavily on network access and communication infrastructure, often unavailable in remote areas, making computation offloading susceptible to breaking down. To address this issue, we propose an MEC-enabled vehicular network assisted through aerial-terrestrial connectivity to provide network access and high data-rate entertainment services to a vehicular network. We present a time-varying, dynamic system model where high altitude platforms (HAPs) equipped with MEC servers, connected to a backhaul system of low-earth orbit (LEO) satellites, are used to provide computation offloading capability to the vehicles, as well as to provide network access for vehicle-to-vehicle (V2V) communications. Our main objective is to minimize the total computation and communication overhead of the joint computation offloading and resource allocation strategies for the system of vehicles. Since our formulated optimization problem is a mixed-integer non-linear programming (MINLP) problem, which is NP-hard, we propose a decentralized value-iteration-based reinforcement learning (RL) approach as a solution. In our Q-learning-assisted analysis, each vehicle acts as an intelligent agent to form optimal strategies for offloading and resource allocation. We further extend our solution to deep Q-learning (DQL) and double deep Q-learning to overcome the issues of dimensionality and the over-estimation of the value functions, as in Q-learning. Simulation results prove the effectiveness of our solution in successfully reducing system costs compared to baseline schemes.

No abstract available

Mobile edge computing (MEC) has been an effective paradigm for supporting computation-intensive applications by offloading resources at network edge. Especially in vehicular networks, the MEC server, is deployed as a small-scale computation server at the roadside and offloads computation-intensive task to its local server. However, due to the unique characteristics of vehicular networks, including high mobility of vehicles, dynamic distribution of vehicle densities and heterogeneous capacities of MEC servers, it is still challenging to implement efficient computation offloading mechanism in MEC-assisted vehicular networks. In this article, we investigate a novel scenario of computation offloading in MEC-assisted architecture, where task upload coordination between multiple vehicles, task migration between MEC/cloud servers and heterogeneous computation capabilities of MEC/cloud severs, are comprehensively investigated. On this basis, we formulate cooperative computation offloading (CCO) problem by modeling the procedure of task upload, migration and computation based on queuing theory, which aims at minimizing the delay of task completion. To tackle the CCO problem, we propose a probabilistic computation offloading (PCO) algorithm, which enables MEC server to independently make online scheduling based on the derived allocation probability. Specifically, the PCO transforms the objective function into augmented Lagrangian and achieves the optimal solution in an iterative way, based on a convex framework called Alternating Direction Method of Multipliers (ADMM). Last but not the least, we implement the simulation model. The comprehensive simulation results show the superiority of the proposed algorithm under a wide range of scenarios.

Internet of Things (IoT) systems require highly scalable infrastructure to adaptively provide services to meet various performance requirements. Combining Software-Defined Networking (SDN) with Mobile Edge Cloud (MEC) technology brings more flexibility for IoT systems. We present a four-tier task processing architecture for MEC and vehicular networks, which includes processing tasks locally within a vehicle, on neighboring vehicles, on an edge cloud, and on a remote cloud. The flexible network connection is controlled by SDN. We propose a CPU resource allocation algorithm, called Partial Idle Resource Strategy (PIRS) with Vehicle to Vehicle (V2V) communications, based on Asymmetric Nash Bargaining Solution (ANBS) in Game Theory. PIRS encourages vehicles in the same location to cooperate by sharing part of their spare CPU resources. In our simulations, we adopt four applications running on the vehicles to generate workload. We compare the proposed algorithm with Non-Cooperation Strategy (NCS) and All Idle Resource Strategy (AIRS). In NCS, the vehicles execute tasks generated by the applications in their own On-Board Units (OBU), while in AIRS vehicles provide all their CPU resources to help other vehicles’ offloading requests. Our simulation results show that our PIRS strategy can execute more tasks on the V2V layer and lead to fewer number of task (and their length) to be offloaded to the cloud, reaching up to 28% improvement compared to NCS and up to 10% improvement compared to AIRS.

The next-generation mobile cellular networks are dedicated to providing a valued and unique service experience by supporting ultrareliable and low-latency communication (URLLC), high throughput, and high availability. Multiaccess edge computing (MEC) is an emerging network solution that provides services and computing functions on edge nodes to provide users with a reliable and high-quality service experience. However, achieving satisfactory Quality of Service (QoS) for diverse service requests in a mobile environment is challenging because of the densely deployed yet resource-constrained MEC servers. A solution to ensure continued service quality is to migrate the services according to the mobility of users. However, in a highly mobile environment such as vehicular communications, this may result in a repeated relocation of services, incurring high operational costs and poor utilization of network resources. Moreover, each service has its own set of communication requirements, such as delay and bandwidth. Meeting these requirements in a highly dynamic and complex vehicular environment is an exacting challenge. Software-defined networking (SDN) concepts are leveraged in MEC to provide a unified control plane interface that performs effective network and service mobility management, to manage the heterogeneity of service requests within the resource-constrained MEC servers. We conducted various Proof-of-Concept (PoC) experiments in an overlapped vehicle-to-everything (V2X) networking environment to demonstrate the feasibility of our proposed system that ensures interconnection and federation among distributed MEC servers and mobile networks.

Roadside units (RSUs), which have strong computing capability and are close to vehicle nodes, have been widely used to process delay- and computation-intensive tasks of vehicle nodes. However, due to their high mobility, vehicles may drive out of the coverage of RSUs before receiving the task processing results. In this paper, we propose a mobile edge computing-assisted vehicular network, where vehicles can offload their tasks to a nearby vehicle via a vehicle-to-vehicle (V2V) link or a nearby RSU via a vehicle-to-infrastructure link. These tasks are also migrated by a V2V link or an infrastructure-to-infrastructure (I2I) link to avoid the scenario where the vehicles cannot receive the processed task from the RSUs. Considering mutual interference from the same link of offloading tasks and migrating tasks, we construct a vehicle offloading decision-based game to minimize the computation overhead. We prove that the game can always achieve Nash equilibrium and convergence by exploiting the finite improvement property. We then propose a task migration (TM) algorithm that includes three task-processing methods and two task-migration methods. Based on the TM algorithm, computation overhead minimization offloading (COMO) algorithm is presented. Extensive simulation results show that the proposed TM and COMO algorithms reduce the computation overhead and increase the success rate of task processing.

The rapid development of the Internet of Vehicles (IoV) along with the emergence of intelligent applications have put forward higher requirements for massive task offloading. Even though Mobile Edge Computing (MEC) can diminish network transmission delay and ease network congestion, the constrained heterogeneous resources of a single edge server and the highly dynamic topology of vehicular edge networks may compromise the efficiency of task offloading, including latency and energy consumption. Vehicular edge networks are also vulnerable to malicious outside attacks. In this paper, we propose a new blockchain-enabled digital twin vehicular edge network (DTVEN) where digital twin (DT) is exploited to monitor network communication, computation, and caching (3C) resources management in real time to provide rich data for offloading decision-making, and blockchain is utilized to secure fair and decentralized offloading transactions among DTs. To ensure 3C resources sharing across edge servers, we design a DT-assisted edge cooperation scheme, which makes full use of edge resources in vehicular networks. Furthermore, a DT-based smart contract is built to achieve a quick and effective consensus process. Then, we apply a task offloading algorithm based on an improved cuckoo algorithm (ICA) and a resource allocation scheme based on greedy strategy to minimize network cost by comprehensively taking into account latency and energy consumption. Numerical results demonstrate that our proposed scheme outperforms the existing schemes in terms of network cost.

The rapid development of connected cars and latency-sensitive applications highlights the importance of efficient and scalable resource allocation in Mobile Edge Computing (MEC)-assisted vehicular networks. This paper proposes a novel Stackelberg game theory framework integrating optical communication technology to address the hierarchical interaction issues between MEC servers and connected vehicles. By combining MEC with optical communication for high-speed data transmission, the proposed Stackelberg game framework enhances resource pricing and task offloading strategies by modeling the MEC server as a leader and the vehicles as followers, while ensuring stability. Experimental results based on OMNeT++ network simulation and Simulation of Urban Mobility (SUMO) traffic generation show that the proposed framework improves resource utilization by 31.4% and reduces latency by 99 percentage points compared to baseline methods. This study demonstrates the potential of optical communication and game theory in advancing MEC-assisted vehicular networks.

When the computation capacity of connected vehicles cannot meet the ultra-low latency requirement of autonomous driving applications, the emergency of Multi-access Edge Computing (MEC) solves this problem effectively, where connected vehicles can offload its computation tasks to edge servers to reduce the latency. In this paper, we investigate the task splitting and computation resource allocation strategy considering the mobility of vehicles based on MEC enabled vehicular network to minimize the system delay. Within the computing framework of alternating direction method of multipliers (ADMM), we propose a computation resource allocation scheme by partially offloading vehicles’ computation tasks to an edge server to balance their computation loads. By analyzing the impact of the number of vehicles and the computation capacity on the average system delay through numerical simulation, we verify that the proposed scheme outperforms the baseline schemes.

With the recent advances in the fifth-generation cellular system (5G), enabling vehicular communications has become a demand. The vehicular ad hoc network (VANET) is a promising paradigm that enables the communication and interaction between vehicles and other surrounding devices, e.g., vehicle-to-vehicle (V2V) and vehicle-to-everything (V2X) communications. However, enabling such networks faces many challenges due to the mobility of vehicles. One of these challenges is the design of handover schemes that manage the communications at the intersection of coverage regions. To this end, this work considers developing a novel seamless and efficient handover scheme for V2X-based networks. The developed scheme manages the handover process while vehicles move between two neighboring roadside units (RSU). The developed mechanism is introduced for multilane bidirectional roads. The developed scheme is implemented by multiple-access edge computing (MEC) units connected to the RSUs to improve the implementation time and make the handover process faster. The considered MEC platform deploys an MEC controller that implements a control scheme of the software-defined networking (SDN) controller that manages the network. The SDN paradigm is introduced to make the handover process seamless; however, implementing such a controlling scheme by the introduction of an MEC controller achieves the process faster than going through the core network. The developed handover scheme was evaluated over the reliable platform of NS-3, and the results validated the developed scheme. The results obtained are presented and discussed.

The advancement of 5G technology has brought the prosperous development of Internet of Vehicles (IoV). IoV services are not only computational intensive but also extremely sensitive to the delay. As a promising computing paradigm, mobile edge computing (MEC) can be applied to IoV scenarios. However, due to the limited resources of a single MEC server, it is difficult to cope with the suddenly increased computation loads caused by emergencies, or the intensive resource requests from busy regions. Therefore, we propose a novel regional intelligent management vehicular system with dual MEC planes, in which MEC servers in the same region cooperate with each other to achieve resource sharing. We classify computing tasks into different types according to their delay tolerances and focus on the optimization problem of resource allocation for different type tasks. And then, we design a resource allocation algorithm based on deep reinforcement learning, which can adapt to the changeable MEC environment to process high-dimensional data. Simulation results confirm that our proposed scheme is feasible and effective.

Mobile-edge computing (MEC)-enabled Internet of Things (IoT) networks have been deemed a promising paradigm to support massive energy-constrained and computation-limited IoT devices. Energy harvesting (EH) further enhances the operating capabilities of IoT devices that normally only possess very limited energy support. Nevertheless, many studies show that IoT devices using EH can experience uncertainty and unpredictability, which can complicate the EH-based IoT network design. Furthermore, with many new services in 5G and the forthcoming 6G eras, such as autonomous driving and vehicular communications, mobility consideration in IoT networks becomes more and more important. In this article, we study the computing offloading and resource allocation problems in an IoT network that supports both mobility and EH. The long-term average sum service cost of all the mobile IoT devices (MIDs) is minimized by optimizing the harvested energy, task-partition factors, the central process unit frequencies, the transmit power, and the association vector of MIDs. An online mobility-aware offloading and resource allocation (OMORA) algorithm is proposed based on the Lyapunov optimization and semidefinite programming (SDP). This online algorithm optimizes the offloading scheme without the need to have prior knowledge of the user mobility, EH model, and channel condition. Theoretical analysis shows that the proposed OMORA algorithm can achieve asymptotic optimality. Simulation results demonstrate that the proposed algorithm can effectively balance the system service cost and energy queue length, and outperform other offloading benchmark algorithms on the system service cost and packet losses.

Multi-access edge computing (MEC) is a key enabler to reduce the latency of vehicular network. Due to the vehicles mobility, their requested services (e.g., infotainment services) should frequently be migrated across different MEC servers to guarantee their stringent quality of service requirements. In this paper, we study the problem of service migration in a MEC-enabled vehicular network in order to minimize the total service latency and migration cost. This problem is formulated as a nonlinear integer program and is linearized to help obtaining the optimal solution using off-the-shelf solvers. Then, to obtain an efficient solution, it is modeled as a multi-agent Markov decision process and solved by leveraging deep Q learning (DQL) algorithm. The proposed DQL scheme performs a proactive services migration while ensuring their continuity under high mobility constraints. Finally, simulations results show that the proposed DQL scheme achieves close-to-optimal performance.

No abstract available

No abstract available

Modern vehicles increasingly rely on exchanging real-time data with other vehicles (V2V) and infrastructure (V2I) making them vulnerable to a growing set of cyber security threats, including spoofing, data manipulation, and denial of service (DoS) attacks. This paper introduces a hybrid AI architecture combining edge computing with cloud based intelligence for improving the security and the resilience of vehicle to everything (V2X) communication systems. Based on the need for prompt responses and the reduction of latency, the proposed approach takes advantage of edge computing to allow onboard AI systems to detect and mitigate threats locally in real time. At the same time, the cloud is also aggregating anonymized threat data from many edge devices using federated learning to enhance global threat models and disseminate security updates across the network. By connecting scalable cloud intelligence offering to localized decision making, this architecture provides an approach to addressing evolving threats, while still reducing communication delays and defending privacy. The study shows that this layered multi-layer AI framework can strengthen the vehicle's cybersecurity of connected vehicles and help with the secure and reliable communication required for intelligent transportation systems.

Environmental perception is a key element of autonomous driving because the information received from the perception module influences core driving decisions. An outstanding challenge in real-time perception for autonomous driving lies in finding the best trade-off between detection quality and latency. Major constraints on both computation and power have to be taken into account for real-time perception in autonomous vehicles. Larger object detection models tend to produce the best results, but are also slower at runtime. Since the most accurate detectors cannot run in real-time locally, we investigate the possibility of offloading computation to edge and cloud platforms, which are less resource-constrained. We create a synthetic dataset to train object detection models and evaluate different offloading strategies. Using real hardware and network simulations, we compare different trade-offs between prediction quality and end-to-end delay. Since sending raw frames over the network implies additional transmission delays, we also explore the use of JPEG and H.265 compression at varying qualities and measure their impact on prediction metrics. We show that models with adequate compression can be run in real-time on the cloud while outperforming local detection performance.

Cloud-based systems could be a solution for enabling one of the emerging technologies, Cellular-Vehicle-to-Everything (C-V2X) communication. To eliminate the limitations of centralized infrastructure elements, the Edge Cloud architecture could be the key in enhancing 5G systems' service capabilities by placing computational resources to the edge of the network, close to the users. To evaluate and validate new systems in this domain is to use model-based simulation tools. Thus, we introduce the Cloud-in-the-Loop (CiL) simulator concept. The implemented framework models the physical movement of vehicles, and based on this information, it orchestrates a complete distributed cloud system and executes various measurement scenarios. Here we focus on the distortions of a Kubernetes-based Edge Cloud environment caused by the application relocation mechanisms initiated due to user (i.e., vehicles) mobility.

By leveraging the use of wireless communication technologies and edge computing capabilities, Cooperative Intelligent Transport Systems (C-ITS) aim to improve safety and traffic management in mobility use cases. However, the deployment of C-ITS poses some critical challenges. Specifically, in heterogeneous systems, it is necessary to guarantee interoperability among the various available wireless technologies. This paper presents a cloud native infrastructure architecture for vehicular communications that guarantees the interoperability between cellular technologies (4G/5G), and specific Vehicle-to-Everything (V2X) communication technologies, such as LTE-PC5 and IEEE 802.11p wireless communications standards. Such interoperability is demonstrated through the implementation of an Edge Infrastructure where a vehicle equipped with one of the aforementioned radio access technologies, sends cooperative awareness messages, and such messages are received in vehicles provisioned with different wireless technologies.

Modern autonomous driving and intelligent transportation systems face critical challenges in managing real-time data processing, network latency, and security threats across distributed vehicular environments. Conventional cloud-centric architectures typically struggle to meet the low-latency and high-reliability requirements of vehicle-to-everything (V2X) applications, particularly in dynamic and resource-constrained edge environments. To address these challenges, this study introduces the V2X-Car Edge Cloud system, which is a cloud-native architecture driven by DevSecOps principles to ensure secure deployment, dynamic resource orchestration, and real-time monitoring across distributed edge nodes. The proposed system integrates multicluster orchestration with Kubernetes, hybrid communication protocols (C-V2X, 5G, and WAVE), and data-fusion pipelines to enhance transparency in artificial intelligence (AI)-driven decision making. A software-in-the-loop simulation environment was implemented to validate AI models, and the SmartX MultiSec framework was integrated into the proposed system to dynamically monitor network traffic flow and security. Experimental evaluations in a virtual driving environment demonstrate the ability of the proposed system to perform automated security updates, continuous performance monitoring, and dynamic resource allocation without manual intervention.

Compared with traditional cloud computing, service providers in edge computing are closer to the requester of the demand, and can better meet the requirements of communication transmission delay. The Internet of Vehicles is one of the important application scenarios of 5G, the vehicle edge nodes of it have a large number of free resources, but its highly dynamic characteristics will lead to environmental uncertainty, which will affect task offloading decisions. In this article, in the vehicle edge cloud computing system, a V2V or V2I communication method is selected based on probability to achieve low-latency and high-reliability commu ication requirements, and the Multi-Armed Bandit (MAB) method is used to learn the surrounding environment of the vehicle so that it can adapt dynamically changing environments to make task offloading decisions. After a large number of simulations, our solution can adaptively offload tasks to minimize the expected delay and obtain a sub-linearly increasing learning regret curve in a high-speed mobile vehicle networking scenario.

Vehicle-to-everything V2X applications usually have strict latency requirements that can be difficult to meet in traditional cloud-centric networks. By pushing resources to edge servers located closer to the users, end-to-end latency can be greatly reduced. Task offloading in edge-cloud environments refers to the optimization of which tasks should be offloaded between the edge and cloud. Moreover, the use of containers to virtualize applications can further reduce resource and time consumption and, in turn, the latency of V2X applications. Even though Kubernetes has become the de facto container orchestrator, the offloading of Kubernetes pods has not been previously studied in the literature, to the best of the authors’ knowledge. In this paper, a theoretical optimization framework is proposed for edge-to-cloud offloading of V2X tasks implemented as Kubernetes pods. First, an utility function is derived in terms of the cumulative pod response time, weighted by the priority levels and resource usage requirements of pods. Based on the optimal theoretical solution to this problem under memory and central processing unit (CPU) constraints, an edge-to-cloud offloading decision algorithm (ECODA) is proposed. Numerical simulations demonstrate that ECODA outperforms first-in, first-served (FIFS) in terms of utility, average pod response time, and occupancy of edge resources. Further, ECODA achieves a good trade-off between performance and computational complexity, and therefore it can help achieve strict latency requirements of V2X applications.

The evolution toward sixth‑generation (6G) networks introduces transformative capabilities in intelligent transportation, particularly throughultra‑reliable, low‑latency vehicle‑to‑everything (V2X) communication. As autonomous and connected vehicles generate vast amounts of data at the edge, conventional centralized learning approaches are increasingly constrained by privacy, bandwidth, and latency limitations. In this paper, we present a federated reinforcement learning (FRL) framework that enables distributed edge agents—such as vehicles and roadside units—to collaboratively learn real‑time decision policies for navigation, collision avoidance, and traffic optimization, without sharing raw data. Our approach models the V2X environment as a decentralized multi‑agent Markov decision process (MDP) and introduces an adaptive aggregation mechanism that accounts for node mobility and communication variability. We implement and evaluate the framework using a co‑simulation environment that integrates SUMO for traffic dynamics and ns‑3 for network emulation. Experimental results demonstrate that our FRL method outperforms centralized baselines by reducing average decision latency by 32 percent, while preserving data privacy and achieving robust convergence under intermittent connectivity. This work advances the deployment of edge AI in future vehicular ecosystems, providing a scalable, privacy‑preserving foundation for real‑time intelligence in 6G‑enabled V2X systems.

Deploying V2X services has become a challenging task. This is mainly due to the fact that such services have strict latency requirements. To meet these requirements, one potential solution is adopting mobile edge computing (MEC). However, this presents new challenges including how to find a cost efficient placement that meets other requirements such as latency. In this work, the problem of cost-optimal V2X service placement (CO-VSP) in a distributed cloud/edge environment is formulated. Additionally, a cost-focused delay-aware V2X service placement (DA-VSP) heuristic algorithm is proposed. Simulation results show that both CO-VSP model and DA-VSP algorithm guarantee the QoS requirements of all such services and illustrates the trade-off between latency and deployment cost.

In cloud-native environments, executing vehicle-to-everything (V2X) tasks in edge nodes close to users significantly reduces service end-to-end latency. Containerization further reduces resource and time consumption, and, subsequently, application latency. Since edge nodes are typically resource and energy-constrained, optimizing offloading decisions and managing edge energy consumption is crucial. However, the offloading of containerized tasks has not been thoroughly explored from a practical implementation perspective. This paper proposes an optimization framework for energy-aware offloading of V2X tasks implemented as Kubernetes pods. A weighted utility function is derived based on cumulative pod response time, and an edge-to-cloud offloading decision algorithm (ECODA) is proposed. The system's energy cost model is derived, and a closed-loop repeated reward-based mechanism for CPU adjustment is presented. An energy-aware (EA)-ECODA is proposed to solve the offloading optimization problem while adjusting CPU usage according to energy considerations. Simulations show that ECODA and EA-ECODA outperform first-in, first-served (FIFS) and EA-FIFS in terms of utility, average pod response time, and resource usage, with low computational complexity. Additionally, a real testbed evaluation of a vulnerable road user application demonstrates that ECODA outperforms Kubernetes vertical scaling in terms of service-level delay. Moreover, EA-ECODA significantly improves energy usage utility.

The highly established transport infrastructure, effective, precise, and timely vehicle-to-vehicle communication is a necessary process. Edge computing is transforming V2X communication by virtue of near-source real-time processing of data and thus accelerating decision-making and the system in turn. In this paper, an improved framework of Reinforcement Learning and Random Forest algorithms for improving data filtering, bandwidth utilization, and system scalability is presented. Compared to the classical cloud-based systems, the emerging edge computing paradigm allows for faster response times in mission-critical applications such as collision avoidance, intelligent traffic signaling, and real-time navigation. The system makes use of Reinforcement Learning for real-time decision-making tasks and Random Forest algorithms for optimal data clustering with VB.NET. The approach also minimizes unnecessary data transfer to the cloud, thereby minimizing bandwidth usage and maximizing performance. Experimental results corroborate that the target model attains 20% reduced latency, 15% increased bandwidth efficiency, and 10% improved prediction accuracy over conventional V2X systems. In addition, network convergence with the 5G network infrastructure facilitates improved communication speed and quality and facilitates seamless end-to-end connectivity among and to connected and autonomous vehicles. While the system has positive impacts, current global issues such as hardware limitation and energy consumption are still present. Future research will further extend the hardware support, accelerate data processing mechanisms, and improve scalability to large-scale V2X environments. In conclusion, this paper presents a realistic and effective solution to improving V2X communication and making transportation systems safer, smarter, and more efficient.

Abstract The Internet of Vehicles (IoV) refers to the growing number of connected vehicles, requiring efficient task scheduling and computation platforms. Our research focuses on communication types and quality of service improvements, particularly in Vehicle-to-Everything (V2X) communication. We evaluated network parameters such as latency and task processing between autonomous vehicles and edge servers. The analysis includes existing solutions for vehicle-to-edge communication and optimization methods to improve latency and resource management. Finally, we propose improvements in resource orchestration for balanced, effective computation at edge servers, enabling novel use cases in IoV.

SUMMARY We propose a component placement mechanism for latency-constrained applications in a distributed system comprising mobile edge and cloud datacenters. It maximizes the achievable processing rate of requests of an application while satisfying the acceptable maximum end-to-end latency to process each request, by placing the components of the application to optimal locations. We evaluated it by simulation and confirmed that it can find an optimal placement according to given situations. As a case study, we applied the mechanism to a V2X application and confirmed its effectiveness.

Vehicle-to-everything (V2X) edge computing accomplishes the goal of low latency by offloading tasks to edge computing servers, but how to reduce the computing latency of vehicle terminals while ensuring low-energy consumption and load balance of servers is still a challenge. In order to address this issue, this article proposes an adaptive computation offloading strategy based on adaptive alternating direction method of multipliers (AADMM). First, a distributed framework for multiple vehicles and multiple road side units (RSUs) is constructed by comprehensively considering the weights of delay and energy consumption, with the optimization objective of minimizing the total system cost. Second, the original variables and dual variables are updated alternately, and the step size is dynamically adjusted based on the magnitude of variable updates, thereby progressively approaching the optimal solution. Finally, simulation experiments show that our proposed strategy can effectively reduce the system cost compared with other traditional algorithms under the comprehensive consideration of delay and energy consumption, and our proposed algorithm has better performance in terms of number of vehicles, speed of vehicles, size of the task, etc.

A computing system for autonomous vehicles must efficiently process vast amounts of data from various sensors in real-time and have a backup design to handle system failures. Adding more computing devices improves performance and redundancy, but increases costs and energy consumption. With the development of vehicle-to-everything (V2X) communication technologies and edge computing, such as multiaccess edge computing (MEC), computation offloading has been introduced. This process involves the use of cloud or edge servers to handle calculations normally performed by on-board computers, thereby reducing their workload and offering redundancy. This article proposes an offloading strategy that optimizes the allocation of software components (SWCs) in autonomous driving software. By optimizing SWC allocation, SWCs can be effectively assigned to specific computing units. This article established a cost function and constraints based on SWC-specific features to address the offloading decision problem as an allocation optimization issue. This article also defines safety-related metrics, including the response time requirements and failure risks of SWCs, to set criteria for offloading decisions. The proposed SWC optimization method is tested and validated using a real autonomous driving vehicle demonstration involving both cloud and vehicular edge environments.

We propose a novelty system to manage Traffic Priority at city intersections by means of our Mobility-Hub (M-Hub), a next-generation Traffic Light Controller that leverages the power of cloud-edge continuum computing, Digital Twin, and Cellular Vehicle-to-Everything (C-V2X) technologies to transform traffic management into a dynamic and intelligent system. M-Hub acts as an open-edge computing platform, enabling real-time data processing, 3rd parties containerized applications and decision-making at the network edge. COGNIT is an open-source cloud-edge continuum framework, that offers many improvements for next-generation Intelligent Transportation Systems (ITS). The continuum allows for the integration of diverse data sources, including vehicular data from C-V2X communication, real-time traffic information from detectors or cameras, and other environmental data, to seamlessly generate Digital Twins in the ACISA smart mobility platform, SATURNO. By combining this data with advanced traffic optimization algorithms implemented in the COGNIT infrastructure, M-Hub can dynamically adjust traffic signal timings, optimize traffic flow, and reduce congestion with optimal use of computational resources.M-Hub has the potential to revolutionize urban mobility, enhancing safety, improving efficiency, and reducing environmental impact.

With the further development of roadside fusion perception, edge computing and C-V2X technology, vehicle-road collaboration will enter the 3.0 era. With the help of edge computing technology, micro cloud computing is mainly used to analyze and calculate road traffic conditions, large-scale vehicle guidance strategies, intelligent traffic scheduling and so on in Intelligent Vehicle Infrastructure Cooperative Systems.In this paper, combined with edge computing characteristics, a complete sensor is deployed on the driving route of self-driving vehicles, the state of traffic participants is sensed and computing units are set on the road edge, and the driving route of self-driving vehicles in the traffic area is cooperatively planned by using edge computing technology.Furthermore, the development and application of roadside edge equipment and edge cloud based on intelligent network connection are studied, so as to meet the practical needs of reducing the input of single-vehicle sensors, reducing the calculation cost of single-vehicle, and being able to use 5G technology for remote manual takeover when autonomous vehicles enter complex traffic scenes such as traffic intersections.

The rapid advancement of edge and cloud computing platforms, vehicular ad-hoc networks, and machine learning techniques have brought both opportunities and challenges for next-generation connected and automated vehicles (CAVs). On the one hand, these technologies can enable vehicles to leverage more computing power from edge and cloud servers and to share information with each other and surrounding infrastructures for better situation awareness and more intelligent decision making. On the other hand, the more distributed computing process and the wireless nature of V2X (vehicle-to-everything) communication expose vulnerabilities to various disturbances and attacks. In this paper, we discuss the security and safety challenges for edge- and cloud-enabled CAVs, particularly when they are under environment interferences, execution errors, and malicious attacks, and we will introduce our recent work and future directions in developing system-driven, end-to-end methodologies and tools to address these challenges and ensure system resiliency under uncertainties.

Co-operative automated driving (CAD) is a key fifth generation mobile networks (5G) use case in which autonomous vehicles communicate over vehicle-to-vehicle (V2V) links requiring a wide range of rate-reliability-delay performance. One key 5G enabler for CAD sidelink radio resource management (RRM) in a multi-operator environment is the virtualization of RRM at the cloud server. This, however, is challenging due to an increase in control plane delay, signaling overhead and complexity. This paper introduces an edge cloud-enabled end-to-end vehicle-to-everything (V2X) architecture to support sidelink RRM in CAD scenarios. Analyzing the problem of a cloud-based sidelink resource allocation for CAD, a utility-based multi-objective optimization problem is described and is translated to three tasks: 1) a vehicle cluster formation as a solution to the clique partitioning problem ensuring vehicle reachability on the control plane, 2) an inter-cluster resource block pool (RB-pool) allocation as a solution to a max-min fairness problem and 3) an intra-cluster resource allocation. The proposed solution framework aims to achieve high modularity, low complexity and decouples cluster formation and RB-pool assignment from the intra-cluster optimum resource allocation, which may be performed on different time scales at different edge cloud entities. Simulation results in a realistic vehicular deployment show significant gains in terms of sidelink throughput and delay while maintaining high link quality.

With the rapid advancement of autonomous driving technologies, Vehicle-to-Everything (V2X) networks are recognized as pivotal in enhancing the efficiency and safety of intelligent transportation systems. However, the highly dynamic and mobile nature of V2X environments poses considerable challenges for task offloading, particularly in achieving low delay, high energy efficiency and task success rate maximization. To address these issues, this study introduces the Dynamic Task Offloading Framework (DTOF) and the Hybrid Integrated Offloading Algorithm (HIOA). The DTOF incorporates dynamic task segmentation with cross-layer resource allocation, thereby improving adaptability under high mobility conditions, with vehicle speeds ranging from 20 to 120 km/h. The HIOA integrates deep reinforcement learning (DRL) with game-theoretic methods to achieve near-optimal multi-objective optimization, encompassing delay minimization and energy efficiency improvement across vehicle densities of 30 to 80 vehicles/km$^{2}$. Specifically designed to satisfy the delay requirement for Level 4 and beyond autonomous driving, the HIOA ensures both low delay and energy efficiency. Hybrid simulations combining SUMO traffic modeling and a 5 G New Radio (NR) channel model demonstrate that the HIOA achieves superior performance compared to existing approaches. Under typical operating conditions, it reduces service access delay by 25%,lowers energy consumption by 18% and elevates task success rate by 30%. Moreover, the HIOA maintains robust performance under peak traffic scenarios (80 vehicles/km$^{2}$) and amid infrastructure impairments (30% RSU failure rate). This work substantially augments the efficiency and adaptability of V2X, establishing a solid groundwork for further development of autonomous driving technologies. Future research will investigate federated learning for cross-domain collaboration and refine the integration of game-theoretic and DRL mechanisms to further reduce computational complexity.

With the rapid development of vehicular networks, vehicle-to-everything (V2X) communications have huge number of tasks to be calculated, which brings challenges to the scarce network resources. Cloud servers can alleviate the terrible situation regarding the lack of computing abilities of vehicular user equipment (VUE), but the limited resources, the dynamic environment of vehicles, and the long distances between the cloud servers and VUE induce some potential issues, such as extra communication delay and energy consumption. Fortunately, mobile edge computing (MEC), a promising computing paradigm, can ameliorate the above problems by enhancing the computing abilities of VUE through allocating the computational resources to VUE. In this paper, we propose a joint optimization algorithm based on a deep reinforcement learning algorithm named the double deep Q network (double DQN) to minimize the cost constituted of energy consumption, the latency of computation, and communication with the proper policy. The proposed algorithm is more suitable for dynamic scenarios and requires low-latency vehicular scenarios in the real world. Compared with other reinforcement learning algorithms, the algorithm we proposed algorithm improve the performance in terms of convergence, defined cost, and speed by around 30%, 15%, and 17%.

With the rapid evolution of communication technology, 5G has begun to enter the stage of large-scale application. Autonomous driving technology is also developing rapidly. At present, it has entered the stage of L2.x and has begun to move towards commercial trial. At present, the vehicle-road coordination technology based on 5G MEC and C-V2X is an important evolution direction of the Internet of Vehicles. Vehicle-Infrastructure coordination means that vehicles communicate with surrounding vehicles, roadside traffic infrastructure, and cloud services to obtain information on the status of surrounding vehicles, roadside traffic signals, traffic signs, etc., in order to improve the safe driving ability of vehicles. To realize L3 level and above vehicle-infrastructure collaborative autonomous driving, it is necessary to collaborate with multiple data sources of “people-vehicle-road-environment”. And one of the requirements of vehicle-infrastructure coordination is to present auxiliary information from multiple data sources obtained by rapid inference of AI technology to vehicles in a more intuitive and comprehensive way. It is also a low-latency, high-speed mobile service that requires end-to-end ultra-low latency and high reliability of network communications. In response to the requirements of vehicle- infrastructure collaborative autonomous driving, this paper proposes a method for dynamic band-width adjustment and TSN channel division based on 5G+TSN network, and applies the method to the cloud-side collaborative vehicle networking system. The system in this paper also applies AI + digital twin technology to generate 3D rendering video streams that are mostly source fusion to assist driving. Finally, the system is tested based on two application scenarios to verify the feasibility of the system.

Along with the development of technology and transportation, the internet of vehicles (IoV) has emerged. However, as the number of vehicles on the road increases, the number of computational tasks that need to be processed increases, and thus energy consumption and time latency are a number of relevant factors that we need to take into account in this process. The combination of vehicle-to-everything (V2X) communication technology and mobile edge computing in IoV provides a feasible solution for offloading and processing of computing tasks in vehicles. In this paper, a predictive vehicle task offloading method (PVTO) is proposed to offload computing tasks to the edge server with vehicle-to-vehicle communication (V2V) and vehicle-to- infrastructure communication (V2I), followed by a multi-objective optimization using genetic algorithm, and then a simple additive weighting algorithm (SAW) and a multiple criteria decision making (MCDM) to solve the optimal offloading strategy. Finally, the effectiveness of PVTO is demonstrated by experimental comparison.

Vehicle-to-everything (V2X) communications have a great potential of enabling future intelligent vehicle applications, and exploiting vehicle mobility is of great importance in designing efficient V2X protocols and applications. Thus, this paper proposes a novel edge-assisted algorithm that makes use of the resources in both cloud and edge sides of vehicular networks to predict vehicle mobility. The proposed algorithm adopts a hybrid architecture of convolutional and recurrent neural networks, and enables computationally efficient transfer learning in each vehicle to generate its customized mobility prediction model. Extensive evaluations have been conducted by using a real taxi mobility data set that is obtained from a testbed deployed in Tokyo, Japan. The results have validated that, compared with other state-of-art algorithms, our proposal improves the prediction F1 score of vehicle mobility by more than 30%, especially for those vehicles that own a strong individual mobility preference.

No abstract available

The densification of 5G deployments and Internet of Things (IoTs) are envisioned to generate data beyond the handling capacity of currently deployed networks. Moreover, the computations and storage on remote cloud servers which are typically situated far away from the end user bring in additional challenges of increased network latency and bottlenecked network bandwidths. Mobile Edge Computing (MEC) is an emerging technology that provides cloud and IT services including computation, networking and storage closer to the mobile subscribers. Therefore, the key architectural change in 5G is of integrating MECs with the the local cellular base stations which can lower latency and improve throughput. This paper explores the feasibility of co-locating MEC with the base station, and evaluate the performance of two sample use-cases of multimedia applications and Vehicle-to-Vehicle/Infrastructure (V2X) using the commercial OPNET simulator. The results thus obtained demonstrate the adoption feasibility in the use-cases presented.

The introduction of Electric Vehicles (EVs) leads to new concern on the E-Mobility. Making charging reservation, by considering the EV's arrival time and its expected charging time at Charging Stations (CSs) has been studied to predict the dynamic status of CSs. In this paper, we propose a Mobile Edge computer Geared v2x for E-mobility Ecosystem (MEGEE), as a decentralized alternative to the conventional centralized cloud based architecture. MEGEE enables the Vehicular Delay/Disruption Tolerant Networking (VDTN)-driven anycasting for information delivery, and Mobile Edge Computing (MEC) functioned CSs for information mining and aggregation. MEGEE efficiently and timely processes essential charging reservations and charging control information, through the Internet of EVs and MEC servers. Our studies show that MEGEE can achieve the close charging performance as performed by the centralized system, while offers a significant saving in communications cost.

Edge computing and artificial intelligence (AI) are important components of vehicle-to-everything (V2X) systems and key enablers of their future evolution. From a standard perspective, multi-access edge computing (MEC) offers cloud computing capabilities at the network edge and an IT environment for providing value-added services to applications and services at the edge (through the use of application programming interfaces, i.e., APIs). This article introduces the latest advances of MEC standards related to V2X. It further explains how to implement roadside MEC in real-world networks based on these latest standards to carry AI and sensor fusion computing workloads to support a wide range of V2X use cases. In particular, our work shows how MEC hosts can be conveniently deployed in road infrastructure to support the critically important computing workloads mentioned above. We also propose some promising directions for future standards development as key enablers for achieving interoperability and scalability of MEC platforms.

The Internet of Vehicles (IoVs) makes communications between numerous devices that use various protocols susceptible to hacker incursions and attacks, which can compromise privacy and seriously jeopardize driving safety. Many studies have been proposed to detect intrusions hitherto, but two major limitations remain. First, traditional Vehicles-to-Cloud (V2C) have difficulty in figuring out the decentralized distribution of data and computational power in IoVs. Second, the majority of studies suffer from unbalanced data in which the attacks only make up a small part and fail to detect low-probability attacks. To address these limitations, we design a Federated Learning-Edge Cloud (FL-EC) communication architecture for IoVs with a Feature Select Transformer (FSFormer) for effective intrusion detection: In FL-EC, mobile users collect and encrypt data before uploading it to edges for training, with edges and cloud functioning as clients and servers in FL, ensuring privacy and efficient data transmission. In FSFormer, we propose a Feature Attention mechanism to search and promote significant features. Furthermore, the Feed-Forward Network is replaced with a Routing module for a deeper but less-parameter network. Extensive experiments show that our model effectively boosts the detection rate of low-probability attacks and outperforms five baseline models in almost all scenarios.

Vehicular-to-everything (V2X) communications are vital part of transportation systems that enable the exchange of information between vehicles, infrastructure, pedestrians, and networks. This system pays prime role for road safety and cooperative driving, but it also suffers from problems such as cyber-attacks, fake data, and lack of reliable information. Artificial Intelligence (AI) provides methods (anomaly detection, adaptive defense, trust management) to fill this gap. This study highlights a hybrid edge-cloud framework that enables lightweight intrusion detection modules on vehicles and roadside side units as well as improved privacy-preserving models through federated updates. Key points of research include taxonomy between AI techniques and V2X threats, an integrated trust story mechanism, and a comprehensive evaluation framework based on accuracy, latency, and robustness.

SUMMARY In future 6G Vehicle-to-Everything (V2X) Network, task offloading of mobile edge computing (MEC) systems will face complex challenges in high mobility, dynamic environment. We herein propose a Multi-Agent Deep Reinforcement Learning algorithm (MADRL) with cloud-edge-vehicle collaborations to address these challenges. Firstly, we build the model of the task offloading problem in the cloud-edge-vehicle system, which meets low-latency, low-energy computing requirements by coordinating the computational resources of connected vehicles and MEC servers. Then, we reformulate this problem as a Markov Decision Process and propose a digital twin-assisted MADRL algorithm to tackle it. This algorithm tackles the problem by treating each connected vehicle as a agent, where the observations of agents are defined as the current local environmental state and global digital twin information. The action space of agents comprises discrete task offloading targets and continuous resource allocation. The objective of this algorithm is to improve overall system performance, taking into account collaborative learning among the agents. Experimental results show that the MADRL algorithm performed well in computational efficiency and energy consumption compared with other strategies. key words: multi-agent deep reinforcement learning, mobile edge computing, task offloading, cloud-edge-vehicle system.

This study proposes an edge computing framework for Artificial Intelligence (AI) optimized traffic management in autonomous vehicles aimed at enhancing real-time responsiveness, reducing network latency and improving traffic flow efficiency. Using Vehicle-to-Everything (V2X) connectivity, the system architecture incorporates a three-layered method with vehicular, edge and cloud levels. Computationally expensive activities are offloaded from autonomous cars to neighbouring edge nodes. While Reinforcement Learning (RL) agents placed at the network's periphery make real-time adjustments to traffic signals, Federated Learning (FL) allows for collaborative method refinement decentralising sensitive data. Furthermore, for precise short-term traffic flow prediction, hybrid deep learning models are used which combine Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) algorithms. Through experimental validation utilising SUMO and CARLA simulations it has been shown that the edge-based system outperforms typical cloud-based models in terms of inference latency reduction (by more than 50%), traffic throughput increases by 22% and predict accuracy improvement by 3.1%. The effectiveness and scalability of the suggested method were confirmed by the fact that resource utilisation was uniform across edge devices. In addition to demonstrating the promise of edge-AI systems for future autonomous transportation infrastructures, the findings show that these systems are better at intelligent traffic management.

For the effective operation of interconnected autonomous vehicles, it is imperative to ensure extremely low latency, high reliability, and significant processing capabilities, which are essential for situational awareness, reasoning, and vehicle-to-everything (V2X) communication. The incorporation of 5G technology with Multi-Access Edge Computing (MEC) presents itself as a promising framework to meet these requirements. However, the relatively lower computational capacity of MEC compared to central cloud servers poses substantial challenges in efficiently scheduling applications associated with autonomous connected vehicles (ACVs). This manuscript presents a comprehensive scheduling framework that aims at optimizing the allocation of vehicular applications on MEC hosts based on various criteria, including physical distance, traffic density, resource availability, and computational demand. We have developed five heuristic algorithms—<inline-formula> <tex-math notation="LaTeX">$BSR$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$THS$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$FFA$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$SFA$ </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">$FAS$ </tex-math></inline-formula>—to address the intrinsic NP-hard problem. To further enhance the quality of the solutions obtained, we propose a novel Variable Neighborhood Search (VNS) mechanism that utilizes the proposed heuristics as initial solutions to explore alternative execution schedules. This approach results in superior alternatives: <inline-formula> <tex-math notation="LaTeX">$\widehat {BSR}$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$\widehat {THS}$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$\widehat {FFA}$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$\widehat {SFA}$ </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">$\widehat {FAS}$ </tex-math></inline-formula>. Experimental evaluations conducted across 1,800 problem instances and 45 distinct <inline-formula> <tex-math notation="LaTeX">$(n_{app}, n_{vm})$ </tex-math></inline-formula> pairs demonstrate the robustness and overall superiority of the VNS-based methods. In particular, <inline-formula> <tex-math notation="LaTeX">$\widehat {THS}$ </tex-math></inline-formula> attains the optimal solution in 65.7% of all instances, with an average deviation of only 0.001. These results highlight the effectiveness of the proposed framework in improving resource utilization and reducing execution delays for ACV applications within MEC environments. Future research may explore extending this VNS framework to broader domains such as cloud and edge computing, as well as integrating adaptive learning-based neighborhood mechanisms or hybrid metaheuristics to further enhance scalability and performance.

The emergence of Internet of Vehicles (IoV) technology provides a wider range of application scenarios for edge computing based on Vehicle-to-everything (V2X). It is essential to ensure the high availability and reliability of services in IoV systems. Currently, cloud service providers have established a data center level of fault tolerance, such as redundancy and checkpoints, guaranteeing the reliability of cloud infrastructure and reducing phenomena such as service termination or downtime. However, current computing systems reactively handle failures. Especially in edge computing, this approach not only lacks flexibility but also consumes excessive system resources, which is not conducive to ensuring the reliability in resource-constrained systems and poses security risks to end users. To mitigate this problem, we propose a Proactive Fault-tolerance Driven Task Scheduling System. Different from the traditional reactive strategies, the proposed framework predicts the possible system crashes by monitoring the critical state indicators of the computing system. According to the prediction results, a class of tasks or services that are most likely to be terminated are rescheduled in advance. Extensive experiments are conducted, and evaluation results demonstrate that our proposed proactive fault tolerance framework can effectively improve the long-term performance of the IoV edge system.

Vehicular Edge Computing (VEC) has emerged as an attractive and feasible paradigm within the Internet of Vehicles (IoV), where computational tasks are offloaded to Road Side Units (RSUs) to reduce both processing delays and resource consumption in vehicles. However, significant challenges arise from the dynamic nature of Vehicular Edge Computing Network (VECN), which is influenced by factors such as vehicle mobility, the limited coverage range of RSUs during task offloading, the real-time unpredictability of RSU workload, and traffic congestion causing network bottlenecks at RSUs. Another challenge is resource management on the RSU side. This paper focuses on minimizing the overall system service delay, providing more effective task scheduling and resource allocation strategies for vehicular tasks. It outlines a scenario where multiple vehicles offload tasks to a nearby RSU, establishing a task pool. In this setup, RSU serves as the coordinator responsible for task scheduling. Specifically, we introduce the Vehicular Earliest Deadline First (V-EDF) algorithm, an enhanced version of the EDF algorithm tailored for vehicular task scheduling. The main enhancements of this algorithm focus on improving resource allocation and prioritizing high-priority tasks while factoring in elements such as task priority, deadlines computational requirements and other relevant parameters. We conducted experimental validation of V-EDF using real-time vehicular aperiodic task sets and evaluated the effectiveness of our proposed algorithm through comparison and analysis.

With advances in vehicular communications technology such as Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V), the computation of vehicular tasks has become distributed facilitated by computation offloading from the vehicular platform to road side units (RSUs) or the cloud infrastructure. In this work, due to communication overheads, we do not consider the cloud and only consider offloading horizontally across RSU nodes. However, the advantage of offloading depends upon how the tasks are scheduled across the RSUs. Unlike existing horizontal offloading works, this work explores the benefits of online task partitioned scheduling for computation offloading from multiple vehicles to RSUs or edge nodes. Specifically, we propose a new efficient timeslot-based online hybrid partitioned scheduling algorithm, which splits some tasks into subtasks and schedules them across RSU nodes while considering vehicle flow constraints. We compared and evaluated the effectiveness of our proposed hybrid partitioned scheduling algorithm with the fully partitioned algorithm. We also compared the performance of the aforementioned algorithms with an optimal scheduling algorithm utilizing several experiments conducted on a real-world vehicular dataset.

Owing to the Internet of Vehicles (IoV) quick growth, both academics and the sector have paid close emphasis to vehicular edge computation (VEC). Nevertheless, because of the unbalanced congestion and the strict delay requirements, task offloading in various junction situations continues to struggle from inefficient resource allocated and poor operation implementation standards. This study proposes a task-offloading technique using a fuzzy decision-making method to deal with ambiguity and uncertainties to solve these problems. Roadside Utilities (RSUs) placed alongside remote roadways typically have limited energy resources, thus they must offer energy-effective planning assistance with the distribution of duties to VEC. However, planning decisions for regional task execution incur computational costs, and assigning duties to edge automobiles incurs transmission costs, making energy usage management difficult. Task data transmission to edge automobiles results in increased RSU power usage even while task offloading lowers response delay. To meet task schedule and supply restrictions, this study proposes an energy-effective automobile planning issue for offloading functions to mobile edge units. This study develops a planning technique depending on on-policy deep reinforcement learning (DRL) and a fuzzy-based DRL to address the extremely complex problem brought on by a rise in the number of automobiles under RSU service. When contrasted to the Q-learning method, this FRL not only speeds up the learning procedure but also enhances long-term payoff.

Vehicular edge intelligence, distinct from traditional edge intelligence, exhibits unique characteristics, including the mobility of vehicles, uneven spatial and temporal distribution of vehicles, and variability in the AI models deployed on vehicles, Roadside Units (RSUs), and edge servers (ESs). In this paper, we propose a Deep Reinforcement Learning (DRL)-based resource orchestration scheme for task inference in vehicle-RSU-edge collaborative networks. In our approach, vehicles’ inference tasks can be processed on the vehicles, RSUs, or ESs, encompassing a total of 9 possible scenarios based on the cross-RSU mobility of vehicles. The scheme jointly optimizes task processing decision-making, transmission power allocation, computational resource allocation, and transmission rate allocation. The objective is to minimize the total cost, which involves a trade-off between task processing latency, energy consumption and inference error rate across all vehicle tasks. We design a DRL algorithm that decomposes the original optimization problem into sub-problems and efficiently solves them by combining the Softmax Deep Double Deterministic Policy Gradients (SD3) algorithm with multiple numerical methods. We analyzed the complexity and convergence of the algorithm. Specifically, we demonstrated its low complexity and fast, stable convergence, which prove its effectiveness in solving the problem. And we demonstrate the superiority of our scheme by comparing it with 5 benchmark schemes across 6 different scenarios.

Mobile Edge Computing (MEC) and caching at vehicular network edge have been recognized as promising technologies in the context of autonomous driving. Roadside Units (RSUs) deployed on both sides of road are regarded as computing nodes and caching nodes, catering to vehicles' requests. Existing edge computing and caching technologies face two challenges: 1) Vehicles' requests keep changing, making content popularity hard to predict. 2) Passive computing and caching technologies struggle to meet demands of computation-intensive and latency-sensitive requests when considering task offloading. To tackle these challenges, we propose a proactive edge computing and caching scheme to optimize task offloading. This scheme involves RSUs proactively sensing and identifying potential tasks that may be requested. Subsequently, it performs edge computing and caches the content based on its predicted popularity to respond to vehicle requests. The primary obstacle of our solution lies in selecting appropriate edge computing and caching nodes to minimize task computation delay and maximize caching benefits. To achieve this objective, we formulate a 0–1 mathematical model and transform it into a Markov Decision Process. Subsequently, we propose a solution based on deep reinforcement learning. Through extensive simulations, we demonstrate that our scheme effectively reduces long-term average computation delay and improves overall response ratio to vehicles' requests.

Nowadays, autonomous driving is one of promising technologies in Internet of Vehicles. However, traditional environment sensing technologies via vehicle-mounted sensors are unable to provide sufficient information due to limited perception capabilities. The emerging solution based on Road-side Units (RSUs) for perceiving environment actively to assist autonomous driving can be expected in the foreseeable future. Specifically, RSUs are deployed to proactively acquire information of the environment. The sensing tasks are processed in an edge computing architecture. Finally, the results are sent to vehicles for reference to autonomous driving. One of its issues is task offloading, that is, how to assign computing tasks from RSUs to either the cloud, the vehicles of autonomous driving, or the local. In this article, we propose a novel strategy for offloading tasks of environment perception. To solve the problem, we formulate task offloading as a 0-1 mathematical model with the objective of minimizing the task delay. Due to the dynamics of the environment, we first present the problem as Markov Decision Process, and a reinforcement learning based approach is provided. Extensive simulation results demonstrate that our strategy can effectively reduce the long-term average delay of the tasks.

No abstract available

The rapid advancement of the Internet of Vehicular has enabled a wide range of vehicular applications, intensifying the demand for low-latency and high-efficiency computation. However, the limited onboard computing capabilities and the highly dynamic nature of vehicular networks present significant challenges for task offloading and resource allocation. To tackle these challenges, Vehicular Edge Computing (VEC) incorporates Roadside Units (RSU) not only as computing nodes but also as relay nodes capable of multi-hop task forwarding, effectively extending the offloading scope and mitigating transmission constraints in sparse or high-mobility scenarios. Moreover, due to the diversity of task demands and fluctuating resource availability, it becomes essential to dynamically adjust task priorities to ensure timely task completion and efficient resource utilization. In this paper, we propose a novel Computing Offloading and Resource Allocation strategy under Relay Collaboration (CORA-RC). CORA-RC exploits the dual roles of RSU as computing and relay nodes to extend the offloading range and improve flexibility, while jointly optimizing task offloading, dynamic priority adjustment, and resource allocation. Simulation results demonstrate that CORA-RC ensures the timely completion of security-related tasks while effectively reducing the average task delay, thereby achieving the lowest average weighted cost among all compared strategies.

Internet of Vehicles (IoV) supported by terrestrial networks can satisfy the necessities of multiple computation-intensive applications. However, current terrestrial networks and resource management mechanisms may only partially guarantee vehicle and in-vehicle user equipment (VUE)’s quality of service due to the limited coverage of roadside units (RSU), especially in remote areas. This paper investigates vehicular edge computing (VEC) in satellite-terrestrial integrated networks with multiple low-earth orbit (LEO) satellites, ground RSUs, and VUEs. In remote areas without RSU coverage, VUEs can offload their partial tasks to satellites to save energy and guarantee latency. We aim to minimize VUEs’ weighted sum energy consumption by jointly optimizing VUEs’ association, data partition, computing resource allocation, power control, and bandwidth assignment under the constraints of maximum tolerant latency, maximum number of outage time slots, computation capacity at each satellite and each RSU, and maximum allowable transmission power at VUEs. Furthermore, we introduce an iterative algorithm by decomposing the original non-convex problem into several sub-problems. We efficiently solve each sub-problem by utilizing variable substitutions, the difference of convex functions algorithms, the Lagrangian dual method, and the Karush-Kuhn-Tucke conditions. Simulation results show that the introduced satellite-terrestrial integrated networks-enabled VEC scheme significantly reduces VUEs’ energy consumption compared to other schemes.

Intelligent Transportation Systems (ITS) leverage Integrated Sensing and Communications (ISAC) to enhance data exchange between vehicles and infrastructure in the Internet of Vehicles (IoV). This integration inevitably increases computing demands, risking real-time system stability. Vehicle Edge Computing (VEC) addresses this by offloading tasks to Road Side Unit (RSU), ensuring timely services. Our previous work FLSimCo algorithm, which uses local resources for Federated Self-Supervised Learning (SSL), though vehicles often can't complete all iterations task. Our improved algorithm offloads partial task to RSU and optimizes energy consumption by adjusting transmission power, CPU frequency, and task assignment ratios, balancing local and RSU-based training. Meanwhile, setting an offloading threshold further prevents inefficiencies. Simulation results show that the enhanced algorithm reduces energy consumption, improves offloading efficiency and the accuracy of Federated SSL.

In the Internet of Vehicles (IoV), vehicles need to process a large amount of perception data to support tasks such as road navigation and autonomous driving. However, their computational resources are limited. Therefore, it is necessary to explore the combination of vehicle–road cooperation with edge computing. Roadside units (RSUs) can provide data access services for vehicles, and deploying edge servers on RSUs can improve the data processing capability in IoV environments and ensure the sustainability of vehicle communications, thus supporting complex traffic scenarios more effectively. In this work, we study the deployment of RSUs in vehicle–road cooperative systems. To balance the deployment cost of RSUs and the quality of service (QoS) of vehicle users, we propose an RSU deployment optimization model with six objectives, including time delay, energy consumption and security when vehicles offload their tasks to RSUs, as well as load balancing and the number and communication coverage area of RSUs. In addition, we propose a Wasserstein generative adversarial network (WGAN)-based Two_Arch2 (WGTwo_Arch2) to solve this many-objective optimization problem to better ensure the diversity and convergence of the solutions. In addition, a polynomial variation strategy based on Lecy’s flight mechanism and a diversity archive selection strategy with an adaptive Lp-norm are also proposed to balance the exploratory and exploitative capabilities of the algorithm. The effectiveness of the proposed algorithm WGTwo_Arch2 for 6-objective RSU deployment optimization is verified by comparisons with five different algorithms.

In Vehicular Edge Computing (VEC), a reasonable and efficient user association and resource allocation approach is a worthwhile research issue. However, most studies in Internet of Vehicles (IoV) only consider vehicle mobility and IoV communication. Therefore, we propose a user association and resource allocation strategy in Integrated Sensing and Communication (ISAC)-aided VEC. Compared with existing solutions, we consider constraints such as sensing and communication interference, vehicle mobility, Road Side Unit (RSU) sensing performance, and vehicle user quality of service (QoS). By quantifying the sensing and communication performance of RSUs, we construct a user association and resource allocation model with the optimisation objective of maximising the average sensing performance and communication performance of the system. Then, combining double dueling deep Q-network (D3QN) algorithm and twin delayed deep deterministic policy gradient (TD3) algorithm, we propose a DRL algorithm based on D3QN-TD3. We represent the user association and resource allocation problem as a Markov Decision Process (MDP) and solve it using the proposed algorithm to obtain the optimal user association, channel allocation, and power allocation strategies. Experimental results show that the proposed algorithm has better performance in terms of downlink transmission rate, radar sensing mutual information, system utility, and task completion rate.

No abstract available

With the rapid advancement of intelligent transportation systems (ITS), vehicle edge caching (VEC) technology shows great potential in processing large-scale vehicle data and providing instant services. The high mobility of vehicles restricts the duration of their connection to a vehicle roadside unit (RSU) to a short interval within the unit's service region. It becomes a challenge to accurately predict and update the popular contents during this period and decide the caching location of the popular contents. The transfer of substantial data volumes concurrently presents a significant danger of privacy breaches. To address these difficulties effectively, we proposed a collaborative caching scheme for vehicular edges that considers mobility of vehicles, employing federated distillation (FD) and deep reinforcement learning (DRL). First, we establish a vehicle mobility model using the real-world vehicle movement dataset T-Drive. After that, we propose a framework that integrates recommender systems with federated distillation instead of the traditional autoencoder (AE)-based federated learning scheme to obtain more accurate global models to predict content popularity. Due to the limited storage capacity of a single RSU, in order to maximise the use of edge cache resources, we propose a cooperative caching algorithm based on deep reinforcement learning, which caches these popular contents in multiple RSUs close to the vehicle. Finally, we evaluate the performance of our proposed algorithm on the vehicle mobility model, and the experimental results show that our proposed scheme improves the caching efficiency by 6.7% and reduces the content delivery latency by 11.3% relative to the five baseline schemes on the MovieLens-1 M dataset.

Federated learning (FL) has emerged as a powerful approach for privacy-preserving model training in autonomous vehicle networks, where real-world deployments rely on multiple roadside units (RSUs) serving heterogeneous clients with intermittent connectivity. While most research focuses on single-server or hierarchical cloud-based FL, multi-server FL can alleviate the communication bottlenecks of traditional setups. To this end, we propose an edge-based, multi-server FL (MS-FL) framework that combines performance-driven aggregation at each server—including statistical weighting of peer updates and outlier mitigation—with an application layer handover protocol that preserves model updates when vehicles move between RSU coverage areas. We evaluate MS-FL on both MNIST and GTSRB benchmarks under shard- and Dirichlet-based non-IID splits, comparing it against single-server FL and a two-layer edge-plus-cloud baseline. Over multiple communication rounds, MS-FL with the Statistical Performance-Aware Aggregation method and Dynamic Weighted Averaging Aggregation achieved up to a 20-percentage-point improvement in accuracy and consistent gains in precision, recall, and F1-score (95% confidence), while matching the low latency of edge-only schemes and avoiding the extra model transfer delays of cloud-based aggregation. These results demonstrate that coordinated cooperation among servers based on model quality and seamless handovers can accelerate convergence, mitigate data heterogeneity, and deliver robust, privacy-aware learning in connected vehicle environments.

Meeting strict deadlines for computation tasks in dynamic Vehicular Edge Computing (VEC) networks remains a critical challenge due to rapidly changing network conditions and resource constraints. This paper presents Meta-PDNN, a novel meta-learning framework that optimizes task offloading to minimize latency while maximizing the number of tasks completed within deadlines. By integrating Model-Agnostic Meta-Learning (MAML) with a probabilistic deep neural network (PDNN), Meta-PDNN rapidly adapts offloading policies to new vehicular environments with minimal fine-tuning. The proposed system meta-trains on diverse edge computing scenarios (e.g., varying vehicle speeds, RSU loads) to learn a generalized policy initialization, enabling real-time adaptation for unseen tasks. Experiments demonstrate that Meta-PDNN significantly outperforms baseline methods in terms of adaptation speed, offloading efficiency, and reliability under varying vehicular conditions.

The Vehicular Edge Computing (VEC) paradigm significantly reduces task processing latency in Internet of Vehicles (IoV) and Intelligent Transportation Systems (ITS) by deploying computational resources at Roadside Units (RSUs). However, the high mobility of vehicles and dynamic task arrivals lead to uneven load distribution among RSUs, severely impacting system performance. Actually, load balancing as an important evaluation metric for VEC system greatly affects the performance of individual edge servers in terms of latency, energy consumption, and task completion rates. In view of this, we propose a Proximal Policy Optimization (PPO) based deep reinforcement learning (DRL) approach to determine the task offloading and migration decisions and incorporate the fairness into the constraint, aiming to achieve efficient load-balanced task offloading in VEC. Particularly, we introduce a metric named Load Balancing Metric (LBM) to optimize RSU resource allocation and employ dynamical task migration strategies to optimize the metric. Simulation results demonstrate that this approach significantly enhances load balancing performance, reduces average latency and energy consumption, and provides an efficient resource scheduling solution for VEC systems.

Nowadays, the convergence of Mobile Edge Computing (MEC) and vehicular networks has emerged as a vital facilitator for the ever-increasing intelligent onboard applications. This paper proposes a multi-tier task offloading mechanism for MEC-enabled vehicular networks leveraging vehicle-to-everything (V2X) communications. The study focuses on applications with sequential subtasks and explores the collaboration of two tiers. In the vehicle tier, we design a needing vehicle (NV)-helping vehicle (HV) matching scheme and inter-vehicle collaborative computation is studied, with joint optimization of task offloading decision, communication, and computation resource allocation to minimize energy consumption and meet delay requirements. In the roadside unit (RSU) tier, collaboration among RSUs is investigated to further address multi-access issues of subchannel and computation resources for multiple vehicles. A two-step method is designed to first obtain optimal continuous solutions of multifaceted variables, and then derive the solution for discrete uplink subchannel allocation with low complexity. Detailed experiments are conducted to demonstrate the proposed method reduces average energy consumption by at least 15% compared with benchmarks under varying task delay requirements and numbers of vehicles and assess the impact of various parameters on system energy consumption.

Vehicular Edge Computing (VEC) technology holds great promise, but also poses significant challenges to the limited computing power of in-vehicle devices and the capacity of Roadside Units (RSUs). At the same time, the highly mobile nature of vehicles and the frequent changes in the content of requests from vehicles make it critical to offload applications to edge servers and to effectively predict and cache the most popular content, so that the most popular content can be cached in advance in the RSU. And also considering protecting the privacy of vehicle user vehicular users (VUs), traditional data-sharing methods may not be suitable for this work, so we use an asynchronous Federated learning (FL) approach to update the global model in time and at the same time can protect the personal privacy of VUs. Unlike the traditional synchronous FL, asynchronous FL no longer needs to wait for all vehicles to finish training and uploading local models before updating the global model, which avoids the problem of long training time. In this paper, we propose an in-vehicle edge computing caching scheme based on asynchronous federated learning and deep reinforcement learning(AFLR), which prefetches possible popular contents in advance and caches them in the edge nodes or vehicle nodes according to the vehicle’s location and moving direction while reducing the latency of the content requests. After extensive experimental comparisons, the AFLR scheme outperforms other benchmark caching schemes.

The rapid development of vehicular networks creates diverse ultra-low latency constrained and computation-intensive applications, which bring challenges to both communication and computation capabilities of the vehicles and their transmission. By offloading tasks to the edge servers or vehicles in the neighbourhood, vehicular edge computing (VEC) provides a cost-efficient solution to this problem. However, the channel state information and network structure in the vehicular network varies fast because of the inherent mobility of vehicle nodes, which brings an extra challenge to task offloading. To address this challenge, we formulate the task offloading in vehicular network as a multi-armed bandit (MAB) problem and propose a novel road side unit (RSU)-assisted learning-based partial task offloading (RALPTO) algorithm. The algorithm enables vehicle nodes to learn the delay performance of the service provider while offloading tasks. Specifically, the RSU could assist the learning process by sharing the learning information among vehicle nodes, which improves the adaptability of the algorithm to the time-varying networks. Simulation results demonstrate that the proposed algorithm achieves lower delay and better learning performance compared with the benchmark algorithms.

Vehicular Edge Computing and Networks (VECoNs) have gained popularity for its enhanced Internet of Vehicles (IoV) capabilities. To satisfy the needs of delay-sensitive and computation-intensive in-vehicle applications, VECoNs need to provide low-latency task offloading services. However, existing offloading frameworks generally overlook the spatially and temporally heterogeneous computation task arrival patterns. The former causes overloading and underloading of RSU computational resources and thus hinders further reduction of offloading latency on the macro-scale, while the latter emphasizes the importance of long-term system performance, especially energy constraints, posing challenges to the design of offloading framework and optimization strategies. This paper introduces a novel distributed two-stage task offloading architecture based on Lyapunov and multi-agent deep deterministic policy gradient (MADDPG). On one hand, it jointly optimizes the initial offloading stage within VEC subsystems and the RSU peer offloading stage to minimize offloading delays for each VEC subsystem. On the other hand, it incorporates RSU energy consumption within long-term constraints to formulate the offloading optimization problem. After decoupling the energy coupling between RSU time slots using the Lyapunov algorithm, a Lyapunov and MADDPG-based distributed task offloading (LAMETO) algorithm is presented to solve the optimal problem in a distributed manner. Simulation results show that the proposed framework and algorithm can reduce the system delay, energy consumption, and energy deficit while stabilizing convergence.

The integration of vehicle edge computing (VEC) and air-ground integrated network is considered as a key technology to achieve autonomous driving. It exploits the ubiquitous service coverage and enables tasks to be offloaded to various components, such as high-altitude platform (HAP), unmanned aerial vehicle (UAV), and roadside unit (RSU). In this article, we address the challenge of minimizing the overall task offloading delay in the air-ground integrated VEC network through a joint multicomputation equipment selection and multidimensional resource allocation (JCESRA) problem. Considering the nonconvexity inherent in the problem, we employ the fundamental idea of the block coordinate descent (BCD) method to tackle it. Initially, we exclude the HAP and decompose the primal problem into three subproblems: 1) low-altitude computation equipment selection; 2) joint bandwidth and computation resource allocation; and 3) UAV trajectory design. The first subproblem, which involves integer programming, is solved by using the many-to-one matching method. Meanwhile, we utilize the CVX and successive convex approximation (SCA) method to solve the last two subproblems, respectively. Considering the matching externality, we utilize the coalition game method to deal with it. Based on the solutions of the three subproblems, the JCESRA algorithm without considering the HAP has been proposed. Subsequently, we consider the HAP into the problem. Because the task offloading decision and computation resource allocation of the HAP problem can be viewed as a knapsack problem, we utilize the dynamic programming method to solve it. Because some tasks are offloaded to the HAP, there are some redundant computation resources in UAVs and RSU. We reallocate the computation resources of UAVs and RSU to further reduce the task offloading delay. At last, we present the complete JCESRA algorithm. The simulation results unequivocally indicate that the proposed JCESRA algorithm outperforms other algorithms by significantly reducing the task offloading delay.

Vehicular edge computing network emerges as a key technique to offload vehicles’ tasks to the nearby roadside unit (RSU). Considering that the RSU may not always meet computation requirements of task vehicles (TVs), utilizing idle computing resources of the surrounding resource vehicles (RVs) becomes a feasible solution to enhance the TVs’ experience. Due to the self-interested property and limited resource of RVs, an efficient incentive mechanism should be designed to encourage RVs to participate in task offloading and resource cooperation. In this work, a deep reinforcement learning-assisted contract incentive mechanism is investigated by considering TVs’ task offloading requirements, RVs’ computation resources, and the RSU’s transmission time incentive. To truthfully reveal TV-RV task-resource coordination types, a contract is designed with computation task-transmission time contract items. The joint task offloading and resource cooperation optimization issue is formulated to maximize the RSU’s utility with incentive compatible (IR), individual rationality (IC), and task offloading constraints. The optimal transmission time strategy is first derived from IR and IC constraints. To obtain the task offloading and task data size strategies, a Markov decision process is formulated. A multiagent parametrized deep Q-network scheme is developed to handle the discrete-continuous hybrid action space problem. Numerical simulations show the feasibility and effectiveness of our proposed incentive method to solve the joint task offloading and resource cooperation problem.

Computation offloading is widely used in vehicular edge computing (VEC) networks to satisfy the computational intensity and latency sensitivity requirements. However, many existing offloading algorithms do not comprehensively consider the dynamically changing characteristics of heterogeneous tasks within a roadside unit (RSU), resulting in tasks timeout and being dropped. In this paper, we design a competitive and cooperative computation offloading (C3O) model to reduce task execution latency. Specifically, when intensive heterogeneous tasks are generated, these vehicles jointly compete for the computing resource of a RSU, or they can also offload tasks to the task vehicle (TaV) with free computing resource. Meanwhile, We analyze the latency model of local execution and offloading to RSU or TaV execution and formulate a sequential task offloading decision problem, NP-hard. To address it, we propose a multi-agent reinforcement learning algorithm based on C3O (MARC3O) to intelligently determine the computation offloading policy for each vehicle according to the state of VEC networks. Simulation results demonstrate that the proposed algorithm can significantly reduce task execution latency and improve task completion rates compared with baseline schemes.

With the rapid development of intelligent vehicles and Intelligent Transport Systems (ITS), the sensors such as cameras and LiDAR installed on intelligent vehicles provides higher capacity of executing computation-intensive and delay-sensitive tasks, thereby raising deployment costs. To address this issue, Vehicular Edge Computing (VEC) has been proposed to process data through Road Side Units (RSUs) to support real-time applications. This paper focuses on the Age of Information (AoI) as a key metric for data freshness and explores task offloading issues for vehicles under RSU communication resource constraints. We adopt a Multi-agent Deep Reinforcement Learning (MADRL) approach, allowing vehicles to autonomously make optimal data offloading decisions. However, MADRL poses risks of vehicle information leakage during communication learning and centralized training. To mitigate this, we employ a Federated Learning (FL) framework that shares model parameters instead of raw data to protect the privacy of vehicle users. Building on this, we propose an innovative distributed federated learning framework combining Graph Neural Networks (GNN), named Federated Graph Neural Network Multi-Agent Reinforcement Learning (FGNN-MADRL), to optimize AoI across the system. For the first time, road scenarios are constructed as graph data structures, and a GNN-based federated learning framework is proposed, effectively combining distributed and centralized federated aggregation. Furthermore, we propose a new MADRL algorithm that simplifies decision making and enhances offloading efficiency, further reducing the decision complexity. Simulation results demonstrate the superiority of our proposed approach to other methods through simulations.

The introduction of edge computing provides a broad application scenario for the Internet of Vehicles. Service programs that were not able to be handled timely by On-Board Unit (OBU) can now be placed on Road Side Unit (RSU) to meet users' requirements of End-to-End (E2E) latency and reliability. However, based on the microservice architecture, services are decomposed of multiple microservices, and the complex dependencies between microservices pose new challenges to their placement. To tackle this problem, we first model the dependencies as a directed acyclic graph (DAG), and the long-term interference-aware placement model is then established to depict the load balance between RSUs and network. After that, we formulate it as an integer linear programming (ILP) problem with the aim to achieve a tradeoff between node load cost and transmission cost while reducing the E2E latencies. Considering the local features of DAG topology, an iterative two-phase heuristic microservice placement algorithm is then proposed. Finally, a simulation environment based on real-world electric taxis trajectory data is constructed, and intensive experiments with several baseline algorithms are conducted to verify the superiority of our proposed algorithm.

In vehicular edge computing environment, massive computation-intensive tasks would be produced from diverse vehicular applications. Data scheduling among vehicles and roadside units(RSUs) is a fundamental issue in timely processing those tasks. However, the task heterogeneity with different computation resource requirements and delay constraints, the distinct capacities of vehicles and RSUs, and the stochastic task arrival, pose significant challenges in realizing efficient data scheduling. The existing literature ignores the multi-core feature of both vehicles and RSUs in data scheduling, which may lead to an inefficient resource usage. To cope with these challenges, in this paper, we first construct a multi-queue multi-block model for heterogeneous task oriented data caching on both vehicle and RSU sides. By fully utilizing the multi-core features of both vehicles and RSUs, a fine-grained offloading model is then developed, involving the association between data blocks and computing cores, and the allocation of computation and communication resources. After that, a long-term loss minimization problem is formulated to facilitate data processing. We leverage the Markov decision process (MDP) to model the optimization problem, which is then solved by our proposed deep deterministic policy gradient (DDPG) based association mapping and resource allocation algorithm (D-AMRA). In D-AMRA, an action transformation method is proposed to map the outputs of DDPG to the form of optimization variables. Eventually, extensive simulations with comparative benchmarks are conducted to evaluate the effectiveness of our proposed D-AMRA.

No abstract available

This paper explores task offloading within load-imbalance vehicular multi-access edge computing (MEC) sys-tems. Addressing the uneven distribution of mobile vehicles causing road side unit (RSU) load imbalances, we leverage vehicle mobility and service pricing to redistribute task loads. RSUs strategically set service prices to alleviate congestion and enhance profitability. Meanwhile, vehicles assess these prices to determine task offloading to different RSU s while in motion, to maximize their individual utility. To achieve this, the Karush- Kuhn- Tucker (KKT) condition is applied to determine the optimal RSU ser-vice pricing. Furthermore, a multi-agent reinforcement learning algorithm, Nash Q-Iearning, is utilized to manage the vehicles' offloading decisions. Simulation results substantiate the efficacy of the Nash Q-Iearning-based task offloading scheme, enhancing the utility of mobile vehicles within competitive environments.

The rise of artificial intelligence and the Internet of Things (AIoT) has paved the way for the resource utilization in Vehicular Edge Computing (VEC) networks. However, vehicles willingness to participate in collaboration still needs to be investigated due to the high speed dynamics of the network. In this paper, we discuss the case of multiple Task Vehicles (TaVs) competing for resources on multiple Service Vehicles (SeVs) proxied by an RSU. To maximize the utility of both parties, reasonable resource prices and task offloading volume are necessary. Firstly, we formulate the interactions between SeVs and TaVs as a two-stage Stackelberg game. Then, we propose an Optimal Differentiated Pricing Approach (ODPA) to find the optimal solution. It consists of two parts. The first part determines the optimal resource price and task offloading volume by taking into account the energy consumption of SeVs and the delay-energy savings of TaVs. The second part matches SeVs with suitable TaVs to maximize the utility of TaVs and SeVs. Simulation results demonstrate that ODPA increases the overall utility of SeV and TaV by at least 6% and 10%, respectively, reduces the energy consumption and task latency of TaV compared to other benchmark approaches.

Vehicular Edge Computing (VEC) emerges as a rem-edy to achieve flexible and fine-grained air pollution monitoring, where vehicles equipped with onboard sensors can sense, process, calibrate and store air pollutants on the drive, and roadside units (RSUs) can be deployed for vehicles to offload data via low-cost vehicle-to-RSU (V2R) communication. However, existing VEC-based air pollution monitoring solution either suffers from high deployment cost, limited V2R communication distance, or degraded data collection latency. To address these challenges, we propose a novel cost-efficient VEC deployment solution for mobile air pollution monitoring, where a set of buses are used to monitor the air pollutants, and selected bus stations are equipped with RSU s for offloading the collected data, considering the effective communication distance and power consumption of V2R. To jointly minimize the VEC deployment cost and data collection latency, we build a multi-objective problem formulation under the constraints of resource, latency, etc. Then we propose a Two-stage Cost-efficient VEC Deployment (TCVD) algorithm based on two heuristic strategies, i.e., the near-equivalence point deployment strategy and the conditioned RSU deployment strategy, with a theoretically-proved worst-case bound. Through extensive evaluations on an open data set of Dublin bus, we verify that TCVD not only reduces the data collection latency by 25.04%, but also reduces the total VEC deployment cost by 30.81 % as compared with existing schemes.

Edge computing has been proved an efficient approach to provisioning computation offloading service to vehicles on road through Road‐Side Units (RSUs). However, the traffic volume on road is highly dynamic, while RSU‐based edge servers are static in terms of geographical location and computation capacity. To address this problem, this paper proposes a mobile edge server placement strategy using cruising UAVs along the roads based on the genetic algorithm. A mathematical model is first built to characterize the deployment cost of these UAV‐mounted servers and their routes. Next, a heuristic UAV‐mounted edge server deployment scheme based on K‐medoid clustering and genetic algorithms is designed. Experimental results verify that the proposed UAV deployment scheme satisfies the offloading demand of IoV nodes while reducing the total deployment cost by 17.05–48.94% compared with existing popular approaches.

Autonomous Vehicles (AV s) require substantial computational resources to perform operations that safely navigate vehicles in urban road networks. Resource-intensive operations are offloaded to roadside units (RSUs), acting as edge servers, to improve the responsiveness and reduce the energy consumed in execution. In this context, a cooperative execution involving the vehicular on-board units (OBUs) and the RSUs can act as a game changer. However, partial offloading is non-trivial and demands addressing the following research challenges. Firstly, the RSU's resources are limited, necessitating regulated resource assignments. Secondly, capturing distinctive vehicle parameters using a unified ranking scheme is imperative. Thirdly, an efficient partition strategy must consider the energy expended and adhere to the real-time operations' deadline needs. This paper proposes a partial offloading scheme, MOVE, catering to the above-mentioned challenges. A deferred acceptance algorithm (DAA) with preferences is proposed to address the first two challenges, whereas a novel energy-aware partitioning strategy resolves the final challenge. The performance of the proposed scheme is evaluated against baseline algorithms, and we observed a 54.04 % and 52.17 % reduction in offloading latency and energy.

With the appearance of more and more devices connected to the Internet, the world has witnessed an ever-growing number of data to be processed. Among those, many tasks require swift execution time, while the storage and computation capability of Internet of Things (IoT) devices are limited. To address the demands of delay-sensitive tasks, we present a vehicular edge–cloud computing (VECC) network that leverages powerful computation capabilities through the deployment of servers in proximity to task-generated devices, as well as the utilization of idle resources from smart vehicles to share the workload. Because these limited resources are vulnerable to sudden data arising, it is imperative to incorporate cloud servers to prevent system overload. The challenge now is to find a task offloading strategy that collaborates both edges and cloud resources to minimize the total time surpassing the quality baseline of each task (tolerance time) and make all tasks meet their soft deadlines of quality. To reach this goal, we first model the task offloading problem in VECC as a Markov decision process (MDP). Then, we propose advantage-oriented task offloading with a dueling actor-insulator network scheme to solve the problem. This value-based reinforcement learning (RL) method helps the agent find an effective policy when not knowing all the state attributes changes. The effectiveness of our method is demonstrated by performance evaluations based on real-world bus traces in Rio de Janeiro (Brazil). The experimental results show that our proposal reduces the tolerance time by at least 8.81% compared to other RL algorithms and 75% compared to greedy approaches.

Vehicular edge computing (VEC) is a promising technology for networked autonomous driving. It not only enhances task offloading and computation for autonomous vehicles, but also expands their perception range and significantly reduces computational costs. Current research primarily focuses on resource allocation for single-vehicle perception task offloading, while cooperative perception tasks have the potential to resolve challenges such as occlusion and sensing range limitations. The phased and composite nature of these tasks introduces specific, yet unmet, design requirements for resource allocation schemes. To address these challenges, this paper develops a novel multi-stage task offloading and resource allocation framework that accounts for the phased structure of collaborative perception tasks and the heterogeneous computational capabilities of vehicles. We introduce a dynamic task-service model for tasks within regions of interest (RoI), and formulate the joint offloading and resource allocation problem as a mixed-integer nonlinear program (MINLP) to minimize task completion latency. An optimal scheme is then proposed, covering task execution, node selection, and resource allocation to enable adaptive and fine-grained collaboration. Numerical results, when compared with existing benchmark algorithms, indicate that our method achieves substantial improvements, reducing latency by an average of 10.5%.

To better provide fast computing services, vehicular edge computing can improve the quality of service and quality of experience for intelligent transportation in 6G by reducing task transmission delay. However, vehicular edge networks face network capability limitations and privacy issues in practice. High‐speed vehicles and the time‐varying environment make them unpredictable. In the meantime, smart vehicles with distinct computation capabilities need to process various tasks with different resource requirements, which will inevitably cause untimely task offloading and massive energy consumption. This paper proposes to use the space‐air‐ground integrated network with blockchain to enhance the network capability and the privacy protection of vehicular edge networks. The digital twin is taken to better capture the dynamic characteristics of vehicles and the entire environment. The urgency level is introduced to meet the delay requirements of different tasks, while considering the impact of digital twin deviation on task offloading. Moreover, the selection algorithm and the task distribution algorithm based on the improved genetic algorithmare are proposed to obtain the optimal offloading strategy. Simulation results demonstrate that, compared with the existing algorithms, the proposed scheme can maximize the system utility while diminishing the total time for task processing.

Vehicular Edge Computing (VEC) has emerged as a promising paradigm for supporting real-time applications by deploying communication and computation resources at the network edge. However, existing task offloading strategies often suffer from performance degradation within VEC systems, where offloading outcomes are difficult to predict accurately due to fluctuating environmental parameters that are challenging to obtain in real-time. Moreover, current offloading methods rely on learning-based approaches, necessitating extensive training efforts to adapt to diverse vehicular applications. Accordingly, we investigate the Cooperative Task Offloading (CTO) problem by considering dynamic nature of vehicular environment, which aims to minimize overall task completion time. We reformulate CTO as a cooperative contextual multi-armed bandit problem and propose a Cooperative Kernel-based Server Selection (CK-SS) algorithm, which facilitates offloading decisions by enabling online reward estimation through information sharing among vehicles. Specifically, we develop a composite kernel function that captures both task characteristics and temporal correlations among historical context-action pairs and an efficient rule for online parameter update. The reward under a given context is estimated based on Gaussian Process and the action is determined using Upper Confidence Bound (UCB) policy. Finally, we implement a simulation model and comprehensive simulation results demonstrate the effectiveness of the CK-SS across various scenarios.

With the rapid development of intelligent transportation and vehicular edge computing (VEC), efficient, fair, and interpretable task offloading has become a key challenge in dynamic and resource-constrained environments. Nonorthogonal multiple access (NOMA) can enhance connectivity and spectrum efficiency. However, conventional resource allocation strategies typically rely solely on channel gain ordering while overlooking spatial factors and fairness. In addition, the lack of transparency in multiagent deep reinforcement learning (MADRL) decision-making raises concerns regarding transparency and trustworthiness. To address these challenges, we propose a NOMA-based task offloading framework that integrates distance and channel-aware (DACA) resource allocation, and we design a distributed multiagent decision-making algorithm based on potential games (DACA-MAD4PG), further incorporating Shapley additive explanations (SHAP) to improve interpretability. The proposed framework is significantly different from existing NOMA-based task offloading approaches in the following three aspects. First, it introduces a DACA joint resource allocation mechanism to enhance both efficiency and fairness in VEC. Second, an exact potential game is incorporated to guarantee system stability and the existence of Nash equilibria. Last, SHAP is integrated to provide posthoc interpretability, thereby improving transparency in multiagent decision-making. Experiments based on real-world DiDi trajectory data demonstrate that the proposed approach significantly reduces task latency, improves service success rate, cumulative reward (CR), and fairness, and outperforms several baselines, thereby providing a stable and interpretable solution for VEC task offloading.

Efficient task offloading for delay-sensitive applications, such as autonomous driving, presents a significant challenge in multi-hop Vehicular Edge Computing (VEC) networks, primarily due to high vehicle mobility, dynamic network topologies, and complex end-to-end congestion problems. To address these issues, this paper proposes a graph attention-based reinforcement learning algorithm, named GAPO. The algorithm models the dynamic VEC network as an attributed graph and utilizes a graph neural network (GNN) to learn a network state representation that captures the global topological structure and node contextual information. Building on this foundation, an attention-based Actor–Critic framework makes joint offloading decisions by intelligently selecting the optimal destination and collaboratively determining the ratios for offloading and resource allocation. A multi-objective reward function, designed to minimize task latency and to alleviate link congestion, guides the entire learning process. Comprehensive simulation experiments and ablation studies show that, compared to traditional heuristic algorithms and standard deep reinforcement learning methods, GAPO significantly reduces average task completion latency and substantially decreases backbone link congestion. In conclusion, by deeply integrating the state-aware capabilities of GNNs with the decision-making abilities of DRL, GAPO provides an efficient, adaptive, and congestion-aware solution to the resource management problems in dynamic VEC environments.

Vehicular edge computing (VEC) systems face critical challenges in balancing computational efficiency, task delay, and data privacy. This paper presents a Federated Multi-Agent Deep Reinforcement Learning (FMADRL) framework to achieve privacy-preserving task offloading in dynamic vehicular networks. The proposed method leverages federated learning to collaboratively train task offloading policies across vehicles, Mobile Edge Computing (MEC) servers, and the cloud, ensuring that sensitive data remains localized while enabling global optimization. A novel reward function is designed to balance task completion, delay, energy consumption, and privacy constraints, while a federated actor-critic model ensures robust decision-making under dynamic network conditions. Simulation results demonstrate that the FMADRL framework significantly reduces average task delay and energy consumption by 30% and 25%, respectively, compared to traditional methods, while maintaining data privacy. These findings underscore the potential of FMADRL to enhance the scalability, efficiency, and security of VEC systems in intelligent transportation networks.

: With the development of vehicle networks and the construction of roadside units, Vehicular Ad Hoc Networks (VANETs) are increasingly promoting cooperative computing patterns among vehicles. Vehicular edge computing (VEC) offers an effective solution to mitigate resource constraints by enabling task offloading to edge cloud infrastructure, thereby reducing the computational burden on connected vehicles. However, this sharing-based and distributed computing paradigm necessitates ensuring the credibility and reliability of various computation nodes. Existing vehicular edge computing platforms have not adequately considered the misbehavior of vehicles. We propose a practical task offloading algorithm based on reputation assessment to address the task offloading problem in vehicular edge computing under an unreliable environment. This approach integrates deep reinforcement learning and reputation management to address task offloading challenges. Simulation experiments conducted using Veins demonstrate the feasibility and effectiveness of the proposed method.

Vehicular Edge Computing enables the utilization of idle resources on vehicles. Task offloading schemes have been widely investigated to satisfy the need for high computational power. In this paper, we propose a novel V2V connectivity prediction and independent task offloading framework for vehicular edge computing, called vConnect. There are three necessary modules for supporting the overall task offloading process: the mobility prediction module, deadline-constraint task partitioning module, and independent task offloading module. This framework aims to maximize the success rate of task offloading within the deadline in urban scenarios. For practicality in the various tasks, the deadline-constraint task partitioning was introduced. Owing to the challenging characteristics of vehicular networks, we introduced a new metric, the so-called V2V connectivity time, which can efficiently represent the duration of uninterrupted communication between vehicles. We introduced the independent task offloading module to select and offload subtasks to multiple optimal candidate vehicles in a one-to-one relationship. We evaluated the performance of our vConnect in both the realistic movement of vehicles in an urban scenario and the real trace file covering the Central Business District (CBD) in Bangkok, Thailand, using OMNeT++, Veins, SUMO, and Python programming. The results show that our vConnect outperforms existing baselines in terms of the success ratio, especially when the required offloading time exceeds 10 s in urban scenarios and 30 s in real-world scenarios. Specifically, our vConnect achieves a 100% success rate in most urban scenarios and up to 94% in the real-world cases, demonstrating its effectiveness in improving task offloading performance.

No abstract available

Modern vehicles require a significant processing power along with storage space. To address this need, offloading the vehicular tasks to adjacent edge servers has been proposed. However, even the considerable processing power of edge servers may not be sufficient to cope with the increasing demand of vehicular computational tasks, leading to potential failure to complete tasks within their deadlines. Vehicular Edge Computing (VEC) has emerged as a promising approach to provide these necessary computational and storage needs through powerful edge servers together with the resources of the nearby vehicles, whereby improving latency for delay-sensitive tasks. However, this heterogeneous processing environment requires a suitable task offloading and scheduling mechanism. This paper introduces a lightweight greedy task offloading algorithm, addressing this requirement by considering the possibility of collaboration between edge servers and utilizing the idle resources of parked vehicles within the edge servers' coverage area. A system solution for efficient use of resources in vehicles and edge servers will be presented. The task offloading problem is modeled as an optimization problem with the objective of maximizing the task completion ratio (TCR), which will be solved using a two-layer greedy approach. Simulation results demonstrate that, on average, the proposed task offloading method can improve the TCR by a factor of 1.29.

Mobile-edge computing (MEC) is a promising computing scheme to support computation-intensive AI applications in vehicular networks, by enabling vehicles to offload computation tasks to edge computing servers deployed on road side units (RSUs) that approximate to them. In this work, we consider an MEC-enabled vehicular edge network (VEN), where each vehicle can offload tasks to edge/cloud computing servers via vehicle-to-infrastructure (V2I) links or to other end-vehicles via vehicle-to-vehicle (V2V) links. In such a cloud-edge–end collaborative offloading scenario, we focus on the joint task offloading, scheduling, and resource allocation problem for vehicles, which is challenging due to the online and asynchronous decision-making requirement for each task. To solve the problem, we propose a Multilayer deep reinforcement learning (DRL)-based approach, where each vehicle constructs and trains three modules to make different layers’ decisions: 1) Offloading Module (first layer), determining whether to offload each task, by using the dueling and double deep Q-network (D3QN) framework; 2) Scheduling Module (second layer), determining where and how to offload each task in the offloading queues, together with the transmission power, by using the parameterized deep Q-network (PDQN) framework; and 3) Computing Module (third layer), determining how much computing resource to be allocated for each task in the computation queues, by using classic optimization techniques. We provide the detailed algorithm design and perform extensive simulations to evaluate its performance. Simulation results show that our proposed algorithm outperforms the existing algorithms in the literature, and can reduce the average cost by 25.86%–72.51% and increase the average satisfaction rate by 3.48%–90.53%.

Vehicular edge computing (VEC) is an effective paradigm in Internet of Vehicles (IoV), which allows vehicles to offload delay-sensitive tasks to nearby road side units (RSUs) for processing, thereby enhancing the Quality of Service (QoS). However, the software defined networking (SDN) controller that manages RSUs often have individual rationality and selfishness, and thus is unwilling to provide free computation resources to vehicles. Meanwhile, the dependency relationships among vehicular subtasks are not well investigated, resulting in unsatisfactory task latency and energy consumption. In order to effectively motivate the selfish SDN controller to participate in computation offloading and comprehensively consider all dependency situations among multiple subtasks, this article proposes a Stackelberg game-based dependency-aware task offloading and resource pricing framework (SDOP). Specifically, we first model the interaction between the SDN controller and vehicles as a Stackelberg game, where both parties wish to maximize their utility. Then, we employ the backward induction approach to analyze the investigated problem, and prove the existence and uniqueness of Nash and Stackelberg equilibrium. Next, we propose a gradient ascent plus genetic algorithm (GAPG) to solve the considered problem. Finally, extensive simulation results show that the proposed GAPG can significantly improve the utility of both the SDN controller and vehicles under various scenarios, when compared with other baseline schemes.

Smart vehicles are increasingly equipped with advanced sensors and computational resources which enable them to detect surroundings and enhance driving safety. VEoTC (Vehicular Edge of Things Computing) solutions aim to exploit these embedded sensors and resources to provide computational services to other users. VEoTC can enhance the Quality of Experience (QoE) of vehicle and mobile users requesting computational tasks by providing context-aware services closer to the users that are otherwise not easily accessible in real time. Additionally, such solutions can extend the computational coverage to areas lacking Roadside Unit (RSU) infrastructure. However, VEoTC frameworks face several challenges in effectively localizing and allocating the distributed resources and offloading tasks successfully due to the high mobility of vehicles and fluctuating user densities. The paper proposes a distributed Machine Learning (ML)-based solution which optimizes task scheduling to smart vehicles and/or RSUs through joint resource allocation and task offloading. We formulate a belief-based optimization problem to maximize the QoE of vehicular users while providing performance guarantees that account for geospatial uncertainty associated with the availability of embedded resources. We propose a Deep Reinforcement Learning (DRL)-based solution to solve the formulated problem in real-time adapting to the dynamic network conditions. We analyze the performance of the proposed approach as compared to benchmark optimization and other ML-based techniques. Furthermore, we conduct hardware-based field test experiments to verify the effectiveness of our proposed algorithm to satisfy the stringent real-time latency requirements for various vehicular applications. According to our extensive simulation and experimental results, the proposed solution has the potential to satisfy the stringent QoE guarantees required for critical road safety applications.

With the emergence of compute-intensive and delay-sensitive applications in vehicular networks, unmanned aerial vehicles (UAVs) have emerged as a promising complement for vehicular edge computing due to the high mobility and flexible deployment. However, the existing UAV-assisted offloading strategies are insufficient in coordinating heterogeneous computing resources and adapting to dynamic network conditions. Hence, this paper proposes a dual-layer UAV-assisted edge computing architecture based on partial offloading, composed of the relay capability of high-altitude UAVs and the computing support of low-altitude UAVs. The proposed architecture enables efficient integration and coordination of heterogeneous resources. A joint optimization problem is formulated to minimize the system delay and energy consumption while ensuring the task completion rate. To solve the high-dimensional decision problem, we reformulate the problem as a Markov decision process and propose a hierarchical offloading scheme based on the soft actor-critic algorithm. The method decouples global and local decisions, where the global decisions integrate offloading ratios and trajectory planning into continuous actions, while the local scheduling is handled via designing a priority-based mechanism. Simulations are conducted and demonstrate that the proposed approach outperforms several baselines in task completion rate, system efficiency, and convergence speed, showing strong robustness and applicability in dynamic vehicular environments.

No abstract available

This paper proposes a Vehicle Trajectory-aware Offloading Multi-Objective Optimization Algorithm (VT-MOOA), a multi-objective optimization algorithm that employs energy consumption, communication and computation delays, vehicle trajectory prediction, task division into sub-tasks, and SDN-based load balancing to optimize task offloading from vehicles to suitable edge servers in vehicular edge networks. The main aim of this work is to design an offloading framework that is robust to high vehicle mobility while ensuring energy efficiency, reduced delays, and balanced resource utilization. The proposed VT-MOOA enhances the S-Metric Selection Evolutionary Multi-Objective Algorithm (SMS-EMOA) by integrating hypervolume-based selection for faster convergence and improves solution quality by minimizing computation delay, minimizing transmission energy, and minimizing physical distance of the vehicle from the RSU while satisfying load balancing constraints, thereby efficiently managing resources in highly dynamic vehicular environments. Existing approaches are often slow, provide sub-optimal solutions due to single objective, false positive prediction or crowding distance reliance, and ignore critical parameters such as real-time vehicle mobility and trajectory prediction. The proposed VT-MOOA approach addresses these gaps by considering these important parameters along with energy efficiency, task deadlines, and load balancing, enabling more effective offloading decisions. Extensive simulations with real-world vehicular mobility datasets demonstrate that VT-MOOA achieves 14% lower energy consumption, 11% faster task completion time, and 13% reduction in computation delay, while also improving load distribution by about 17% compared to existing solutions, outperforming them.

Vehicular edge computing (VEC) has emerged as a promising paradigm to support delay-sensitive and computationintensive applications in intelligent transportation systems. However, effective task offloading in VEC is challenged by heterogeneous vehicle resources, dynamic network conditions, and insufficient incentives for resource participation. To address these issues, this paper proposes a Stackelberg game-based incentive mechanism that encourages resource vehicles to contribute their computational power while ensuring reliable task execution. A fuzzy inference system is introduced to evaluate candidate vehicles by jointly considering computational capacity, mobility stability, and link quality. Based on this evaluation, task vehicles determine incentive levels, and resource vehicles respond with optimal strategies to maximize their utilities. Extensive simulations demonstrate that the proposed scheme significantly outperforms baseline approaches, achieving higher task completion ratio, lower average task delay, and better fairness. These results highlight the effectiveness of combining game-theoretic incentive design with multi-dimensional quality assessment for cooperative task offloading in vehicular networks.

No abstract available

This work considers a parallel task execution strategy in vehicular edge computing (VEC) networks, where edge servers are deployed along the roadside to process offloaded computational tasks of vehicular users. To minimize the overall waiting delay among vehicular users, a novel task offloading solution is implemented based on the network cooperation balancing resource under-utilization and load congestion. Dual evaluation through theoretical and numerical ways shows that the developed solution achieves a globally optimal delay reduction performance compared to existing methods, which is also validated by the feasibility test over a real-map virtual environment. The in-depth analysis reveals that predicting the instantaneous processing power of edge servers facilitates the identification of overloaded servers, which is critical for determining network delay. By considering discrete variables of the queue, the proposed technique's precise estimation can effectively address these combinatorial challenges to achieve optimal performance.

Vehicular edge computing (VEC) leverages roadside units (RSUs) to provide low-latency and energy-efficient computation for vehicular networks. However, existing offloading schemes still face challenges under high vehicle mobility, dynamic network conditions, uneven RSUs loads, and different Quality of Service (QoS) requirements. To address this, we propose a mobility-aware partial task offloading (MAPRO) framework that partitions tasks for parallel processing across multiple RSUs while considering mobility, network dynamics, and QoS demands. A hybrid framework combining Proximal Policy Optimization (PPO) and numerical optimization efficiently solves offloading, resource allocation, and power control. Experiments on real trajectories show that MAPRO outperforms state-of-theart methods.

With the development of 6G communication technology, a large number of computation-intensive and latency-sensitive applications have emerged, compelling vehicles to achieve the expected Quality of Service (QoS) through task offloading. Meanwhile, against the backdrop of the growth of the low-altitude economy, Unmanned Aerial Vehicle (UAV)-assisted Vehicular Edge Computing (VEC) has come into being. In this article, we study the task offloading problem in UAV-assisted VEC, where vehicles can choose four modes to process their tasks: local computing, Vehicle-to-Vehicle (V2V) offloading, Vehicle-to-Infrastructure (V2I) offloading and Vehicle-to-UAV (V2U) offloading. We develop a blockchain-based Asynchronous Multi-Agent Reinforcement Learning (AMARL) algorithm to solve the task offloading problem. Asynchronous updates can alleviate the problem of slow aggregation speed in traditional federated learning. The participation of the consortium chain makes the model parameter transmission more secure. Simulation results show that our proposal has achieved an up to 8.6% improvement in offloading efficiency compared to the state-of-the-art schemes.

Vehicular Edge Computing (VEC) plays a crucial role in enabling low-latency services in intelligent transportation systems. However, high vehicle density in urban areas often leads to dynamic fluctuations in wireless signal quality and edge energy consumption, which degrades system performance. This paper proposes a lightweight, priority-aware task offloading scheme that leverages real-time network indicators Received Signal Strength Indicator (RSSI) and Packet Loss Ratio (PLR) to make intelligent offloading decisions. Our proposed approach prioritizes safety-critical and latency-sensitive tasks during offloading, improving energy efficiency and reducing latency without requiring deep learning or complex optimization. Results show improved performance across key metrics including edge energy consumption, packet delivery, and network latency under varying vehicular loads.

: The rapid advance of Connected-Automated Vehicles (CAVs) has led to the emergence of diverse delay-sensitive and energy-constrained vehicular applications. Given the high dynamics of vehicular networks, unmanned aerial vehicles-assisted mobile edge computing (UAV-MEC) has gained attention in providing computing resources to vehicles and optimizing system costs. We model the computing offloading problem as a multi-objective optimization challenge aimed at minimizing both task processing delay and energy consumption. We propose a three-stage hybrid offloading scheme called Dynamic Vehicle Clustering Game-based Multi-objective Whale Optimization Algorithm (DVCG-MWOA) to address this problem. A novel dynamic clustering algorithm is designed based on vehicle mobility and task offloading efficiency requirements, where each UAV independently serves as the cluster head for a vehicle cluster and adjusts its position at the end of each time slot in response to vehicle movement. Within each UAV-led cluster, cooperative game theory is applied to allocate computing resources while respecting delay constraints, ensuring efficient resource utilization. To enhance offloading efficiency, we improve the multi-objective whale optimization algorithm (MOWOA), resulting in the MWOA. This enhanced algorithm determines the optimal allocation of pending tasks to different edge computing devices and the resource utilization ratio of each device, ultimately achieving a Pareto-optimal solution set for delay and energy consumption. Experimental results demonstrate that the proposed joint offloading scheme significantly reduces both delay and energy consumption compared to existing approaches, offering superior performance for vehicular networks.

The concept of vehicular edge computing (VEC) has recently been envisioned as a promising paradigm to satisfy the quality of service (QoS) requirement of delay-sensitive intelligent applications in future networks. However, the limited radio frequency communication range necessitates the dense deployment of communication infrastructures to accommodate the increasing number of connected vehicles and data traffic. This could lead to growing equipment and energy costs, hindering the full realization of the VEC system. To address such limitations, we introduce a double roadside reconfigurable intelligent surface (RIS) assisted VEC network in this paper, where RISs are deployed inside the coverage gaps between two RSUs to extend the service range. The research goal is to maximize the sum offloading efficiency by jointly optimizing offloading decisions, computation resource allocation, and phase shift vectors of RISs. Since the original problem is a challenging mixed-integer non-linear problem (MINLP), we decompose it into a top-problem for optimizing offloading destinations and phase shift vectors, and a sub-problem for optimizing offloading ratios and computation resources. We propose a deep reinforcement learning (DRL)-based algorithm for quickly obtaining near-optimal offloading decisions and phase shift vectors, alongside a Dinkelbach-based method for obtaining optimal offloading ratios and computation resource allocation. Simulation results demonstrate that our proposed algorithm achieves near-optimal performance compared to other benchmarks and enables real-time decision-making.

With the advances in artificial intelligence and communication technologies, vehicular edge computing (VEC), as a newly developed computing paradigm, is gaining more and more attention from both academia and industry. Complex demands and on-board applications need to be offloaded to edge servers for Quality of Experience (QoE). Nevertheless, the offloading process increases the risk of user privacy leakage, and the effectiveness of resource allocation algorithms is urgently desired in latency-sensitive tasks. To this end, we employ quantum key distribution (QKD) and blockchain to secure communication and computation, where key generation rate (KGR) associated with transmission and computation-aware is investigated for resource allocation problem. In consideration of the number of existing qubits and technical bottlenecks, we propose a tensor network preprocessing-based quantum deep reinforcement learning algorithm (TN-QDRL), which exploits amplitude encoding and the unique properties of quantum superposition and entanglement states to tackle the complex Markov decision process in a multi-dimensional state space. Additionally, we provide a search strategy for quantum state probabilistic transformations integrated with an improved Grover's algorithm. Simulation results indicate that our algorithm achieves a convergence speed that is 62.11% faster in high-dimensional real-world VEC scenarios and consumes 58.19% fewer quantum resources compared to other benchmarks.

In Unmanned Aerial Vehicle (UAV)-assisted Vehicular Edge Computing (VEC), Federated Learning (FL) offers a means to protect user privacy during the training of models using multiple vehicle datasets. However, involving numerous vehicles in the training process can lead to significant communication overhead, thereby increasing FL latency and energy consumption. To address this issue, we propose an energy-latency tradeoff scheme for the joint optimization of vehicle selection and resource allocation in UAV-assisted VEC. Our investigation focuses on maximizing long-term training rewards for vehicle selection and resource allocation in FL, while considering constraints such as UAV energy consumption, vehicular energy consumption, bandwidth, and vehicle mobility. This problem is formulated as a Mixed-Integer Nonlinear Programming (MINLP) problem and modeled as a Markov Decision Process (MDP). We proposed an algorithm based on AdamW and Butterfly Optimization Algorithm (BOA) for Double-Depth Q-networks (AB-DDQN) to determine the optimal decisions. To expedite algorithm convergence, we replace the stochastic gradient descent (SGD) algorithm with AdamW algorithm and employ BOA to select hyperparameters, enhancing algorithm performance. Experimental validation using the GTSDB dataset demonstrates that our algorithm effectively reduces latency and energy consumption in FL.

In Vehicular Edge Computing (VEC) environments, the increasingly complicated functional and non-functional requirements from vehicular applications such as MetaVehicles usually incur larger sizes of task-input data, which not only increase the transmission delay of task-input data via the front-haul links but also degrade the quality of experience for users, even if computation tasks can be offloaded and executed at the network edge. In this article, we put forward a caching-enabled task offloading strategy, by caching and reusing the universal context data at the edge server, to avoid duplicated data transmission in VEC systems. The goal is to minimize the overall response latency for all the tasks, by jointly optimizing task offloading, content caching, and resource allocation decisions in VEC. The optimization problem is formulated as a Mixed-Integer Nonlinear Programming (MINLP) problem. To efficiently solve this problem, we decompose this problem into two subproblems, namely, the computing Resource Allocation (RA) problem and the Joint Offloading and Caching (JOC) problem. The corresponding algorithms are put forward to solve the content caching and task offloading problems, respectively. Numeric evaluation reveals that our strategies and algorithms can achieve better performance in minimizing the overall response latency, in comparison with other approaches.

With the rapid development of vehicular edge computing (VEC) and artificial intelligence (AI), the emergence of vehicle edge intelligence meets the need for real-time vehicle intelligence applications. But the execution of deep neural networks (DNNs) requires a large amount of data input, which results in a large amount of computing resources required for the execution of DNN tasks. This also brings a certain burden to the deployment of DNN tasks and the resource allocation of edge servers. In addition, due to the high mobility of vehicles in the VEC, the backhaul delay of vehicle edge intelligent task results increases, affecting the vehicle’s quality of experience (QoE). We propose a joint optimization strategy for service migration and resource allocation aimed at minimizing the average task completion delay. This strategy comprehensively considers service migration actions and edge server resource allocation, which is proved to be a mixed integer nonlinear programming (MINLP) problem, and hence we formulate it as an Markov decision process (MDP). To solve this problem, we propose a service migration algorithm based on the self-attention mechanism-based double deep Q-network and deep deterministic policy gradient algorithm (SA-DDQN-DDPG) to solve it to obtain the optimal system service migration strategy. The experimental results show that the proposed SA-DDQN-DDPG algorithm has good performance in reducing latency. The average migration latency is reduced by 40.41%, 20.7%, and 14.50% compared with always, DQN and DDQN, respectively.

Intelligent connected vehicles (ICVs) face challenges in handling intensive onboard computational tasks due to limited computing capacity. Vehicular edge computing networks (VECNs) offer a promising solution by enabling ICVs to offload tasks to mobile edge computing (MEC), alleviating computational load. As transportation systems are dynamic, vehicular tasks and MEC capacities vary over time, making efficient task offloading and resource allocation crucial. We explored a vehicle–road collaborative edge computing network and formulated the task offloading scheduling and resource allocation problem to minimize the sum of time and energy costs. To address the mixed nature of discrete and continuous decision variables and reduce computational complexity, we propose a hybrid hierarchical deep reinforcement learning (HHDRL) algorithm, structured in two layers. The upper layer of HHDRL enhances the double deep Q-network (DDQN) with a self-attention mechanism to improve feature correlation learning and generates discrete actions (communication decisions), while the lower layer employs deep deterministic policy gradient (DDPG) to produce continuous actions (power control, task offloading, and resource allocation decision). This hybrid design enables efficient decomposition of complex action spaces and improves adaptability in dynamic environments. Results from numerical simulations reveal that HHDRL achieves a significant reduction in total computational cost relative to current benchmark algorithms. Furthermore, the robustness of HHDRL to varying environmental conditions was confirmed by uniformly designing random numbers within a specified range for certain simulation parameters.

The increasing adoption of mobile edge computing (MEC) and non-orthogonal multiple access (NOMA) in vehicular networks is to reduce the execution delay of computation-intensive tasks and improve the spectrum efficiency. In this paper, we introduce a NOMA-assisted vehicular edge computing network, where vehicular users (VUs) form NOMA groups to share the radio resource with cellular users (CUs) for offloading their computing tasks to the MEC server in a highway scenario. Under the VU's execution delay constraints, we jointly optimize the computation resource allocation of the MEC server, the data transmission time, and the offloading decision to minimize the long-term energy consumption of the system. However, the long-term stochastic optimization problem is intricate due to the VU's mobility and the time-varying of the wireless channel. We thus propose a Lyapunov optimization based algorithm to transform the original problem into a single time slot optimization problem. Specifically, we decouple this problem into the computation resource allocation sub-problem solved at the MEC server and the offloading decision sub-problem solved at each VU. The optimal computation resource allocation is obtained by solving the knapsack problem, while the cross-entropy based algorithm is used to determine the optimal offloading decision for VUs. After that, our numerical simulations are conducted to demonstrate the effectiveness of the proposed algorithms.

In contrast to the traditional vehicular edge computing (VEC) network, the air-ground integrated vehicular edge computing (AGI-VEC) network presents significant benefits characterized by continuous coverage, lower delay, and higher data rates for the Internet of vehicles (IoV). In this paper, considering the vehicles can be connected to the roadside units (RSUs) and unmanned aerial vehicles (UAVs) by using the cellular vehicle-toeverything (C-V2X) links, we investigate a joint task offloading and resource allocation problem in the AGI-VEC network with dual Uu/PC5 interface to minimize the task offloading delay. Due to the non-convexity of the problem, it is difficult to solve by utilizing the traditional methods. We transform it into a Markov decision process (MDP), and then propose a joint task offloading, power allocation, computation resource allocation, and UAV trajectory design (JTPCU) algorithm based on the multi-agent soft actor-critic (MASAC) method. Simulation results demonstrate that, compared to the other benchmark algorithms, the proposed algorithm achieves better performance in reducing task offloading delay.

With the rise of the Internet of Vehicles (IoV), a growing number of in-vehicle applications have been developed, significantly enhancing the driving experience, while simultaneously imposing excessive higher demands on computing resources. Vehicle Edge Computing (VEC) emerges as a promising solution by offloading computational tasks to edge servers positioned near vehicles. However, the limited computing capacity of these edge servers necessitates the efficient allocation of resources to adequately meet the demands of all vehicles. In this paper, we introduce a multi-vehicle VEC offloading framework that considers both the execution time of tasks and the costs incurred when utilizing edge server resources and transmitting tasks between Roadside Units (RSUs). Building on this framework, we design an optimization problem to minimize the average weighted cost, modeled as a Markov Decision Process (MDP). To address this, we propose a DDPG-based Resource Allocation and Offloading Decision Algorithm (DRAODA). This algorithm enables the control center to generate resource allocation strategies while allowing individual vehicles to make task offloading decisions independently, based solely on their own data and the edge server's status, without relying on information from other vehicles. Additionally, we propose an Optimal Task Offloading Destination Selection Algorithm (OTODSA) to further minimize the average weighted cost, enhancing the overall efficiency and effectiveness of the resource allocation process.

The increasing adoption of smart mobility and connected vehicles necessitates significant improvements in underlying infrastructure, particularly in real-time data processing and decision-making. Vehicular Edge Computing (VEC) has emerged as a vital solution by enabling computation closer to data sources, thereby reducing latency and reliance on centralized cloud systems. However, efficient allocation of edge resources (processing power, bandwidth, and storage) remains a critical challenge due to the highly dynamic, decentralized nature of vehicular networks. Traditional optimization techniques often fall short under these conditions. This study explores a quantum-inspired optimization framework designed to enhance resource management in VEC environments by leveraging principles of quantum computing such as superposition and probabilistic state selection within classical hardware. Extensive simulations involving 10 vehicles and 3 edge servers were conducted to evaluate the framework's performance. The dynamic resource demand fluctuated between 7 and 18 units, and server utilization ranged from 0.2% to 1.4%, illustrating diverse operational conditions. The proposed quantum-inspired model showed superior efficiency, achieving up to 35% improvement in fitness gain compared to traditional algorithms, with convergence to optimal fitness in just 45 iterations. The solution space was explored effectively using quantum state amplitude representations, which improved solution diversity and robustness in decision-making. Furthermore, fairness in resource distribution was evaluated using Jain’s Fairness Index, yielding a high score of 0.914, demonstrating equitable allocation among vehicles. Additional results revealed that task completion times ranged from 1.5 to 3.5 seconds, with processing delays being the major contributor. The system exhibited sublinear scalability, performing well up to 50 vehicles but declining as the vehicle count increased to 200, indicating a need for further optimization strategies. Although the model operates in a classical environment without quantum hardware, it offers substantial performance benefits. This research highlights the potential of quantum-inspired optimization for real-time, fair, and scalable resource management in vehicular networks. Future work should incorporate real-world vehicular trace data, expand scalability tests, and explore integration with 5G and energy harvesting mechanisms. These advancements will further support intelligent, secure, and sustainable transportation systems driven by edge computing technologies.

Mobile Edge Computing (MEC) effectively alleviates the pressure on limited in-vehicle computing resources and energy supply caused by computation-intensive vehicular applications. However, the uneven spatial distribution of users leads to load imbalance among adjacent MEC servers, significantly increase the latency and energy consumption costs for vehicles. Therefore, achieving optimal configuration of available computing resources in MEC servers to accomplish the goal of low-latency and low-energy task offloading has become a critical issue to address. To tackle this problem, this study proposes a Multi-RSU Load Balancing (MRLB) strategy based on multi-hop network technology. This strategy dynamically allocates computing tasks to neighboring RSU server clusters with available computing resources through task segmentation and computation offloading mechanisms. Meanwhile, adaptive resource allocation strategies are implemented based on task quantity and task scale characteristics. Specifically, this study designs a multi-RSU collaborative offloading algorithm based on Deep Deterministic Policy Gradient (DDPG) to solve the optimal offloading decision. Additionally, by integrating the Lagrange multiplier method and Sequential Quadratic Programming (SQP) algorithm, the joint optimization of imbalanced task segmentation decisions and optimal CPU frequency allocation decisions for RSU servers is achieved. Experimental results demonstrate that the proposed method can achieve efficient multi-RSU resource allocation and ensure coordinated optimization of both system latency and energy consumption costs across diverse device conditions and varying network scenarios, particularly in load-imbalanced situations.

To leverage the vast amounts of onboard data while ensuring privacy and security, federated learning (FL) is emerging as a promising technology for supporting a wide range of vehicular applications. Although FL has great potential to improve the vehicular edge intelligence(VEI), challenges arise due to vehicle mobility, wireless channel instability, and data heterogeneity. To mitigate the issue of heterogeneous data across vehicles in FL, artificial intelligence-generated content (AIGC) can be employed as an innovative data synthesis technique to enhance FL model performance. In this paper, we propose AIGC-assisted Federated Learning for Vehicular Edge Intelligence (GenFV). We further propose a weighted policy using the Earth Mover’s Distance (EMD) to measure data distribution heterogeneity and introduce a convergence analysis for GenFV. Subsequently, we analyze system delay and formulate a mixed-integer nonlinear programming (MINLP) problem to minimize system delay. To solve this MINLP NP-hard problem, we propose a two-scale algorithm. At large communication scale, we implement label sharing and vehicle selection based on mobility and data heterogeneity. At the small computation scale, we optimally allocate bandwidth, transmission power and amount of generated data. Extensive experiments show that GenFV significantly improves the performance and robustness of FL in dynamic, resource-constrained environments, outperforming other schemes and confirming the effectiveness of our approach.

Vehicle edge computing (VEC) leverages compact cloud computing at the mobile network edge to meet the processing and latency needs of vehicles. By bringing computation closer to the vehicles, VEC reduces data transmission, minimizes latency, and boosts performance for compute-intensive applications. However, during peak hours of urban road traffic, the scarce computational resources available at edge servers could pose challenges in fulfilling the processing needs of vehicles. Introducing remote aerial vehicles (RAVs) as supplementary edge computing nodes could significantly mitigate the aforementioned issue. In this article, we propose a flexible edge computing framework in which a fleet of RAVs function as mobile computational service providers, offering computation offloading services to multiple vehicles. We design and optimize a computation offloading model for the RAV-enabled VEC environment. The proposed model tackles the task offloading challenge, aiming to optimize RAV revenue and task processing efficiency while considering the constraints of RAVs’ restricted computational power and energy resources. Toward this end, our model jointly considers two key factors: 1) task partitioning and 2) computational resource allocation. To tackle the challenges posed by the aforementioned nonconvex optimization problem, we construct a Markov decision process (MDP) model for the multi-RAV-enabled mobile edge computing system and introduce an innovative multiagent deep reinforcement learning (MADRL) framework addressing the decision-making challenge represented by MDP model. Comprehensive simulation outcomes illustrate that our devised task offloading technique outperforms other optimization methods.

Recently, vehicular edge computing (VEC) has become one of the hottest research fields in Internet of Vehicles (IoV). It provides certain computing, storage, and caching resources at the edge of radio network to execute different kinds of vehicular applications, which can significantly reduce the latency of network operation and service delivery. Also edge caching is an effective way to reduce execution delay and backhaul workload. Undeniably, due to the lack of global information and the time-variety of IoVs, it is a challenge to design a comprehensive execution and resource allocation scheme, including whether to offload and cache, how to offload and cache, and so on. So in this article, we mainly propose a multiuser computation offloading and wireless-caching resource allocation problem with linearly related requests in a VEC system. A multivariable, nonlinear and coupled problem is formulated to minimize the average execution delay, including local execution, interaction among consecutive requests, and mobile edge computing (MEC) execution. Then the deep deterministic policy gradient (DDPG) algorithm is adopted to solve the proposed problem, as it is a strategy learning method for continuous behavior. And simulation results show that our proposed method outperforms other methods in many aspects.

Computer vision plays a crucial role in enabling connected autonomous vehicles (CAVs) to observe and comprehend their surroundings. The computer vision tasks are typically based on convolutional neural networks (CNNs). However, CNNs often require significant processing power. Techniques like early exiting and split computing enhance CNN task execution latency and adaptability to varying environmental conditions. Since the split computing introduces additional overhead for offloading of the task from the CAV to an edge servers, we incorporate multiple autoencoders within each split point to enhance the adaptability of splitting under varying environmental conditions. However, the autoencoders introduce an additional layer of complexity related to the selection of the optimal compression strategy alongside the splitting and exiting decisions. To tackle this challenge, we introduce a novel approach based on the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm. This algorithm dynamically and jointly determines the most suitable exit point, split point, and autoencoder. Furthermore, the MADDPG-based approach considers other CAVs when selecting action, promoting cooperation among CAVs. Our results demonstrate that the proposed approach reduces latency up to 44.4% while maintaining at least comparable or even higher accuracy of the computed vision outcome compared to the state-of-the-art solutions.

No abstract available

Vehicular Edge Computing (VEC) has emerged as a promising paradigm to enhance the Quality of Service (QoS) in the Internet of Vehicles (IoV). However, the limited computing capabilities of Task Vehicles (TVs), coupled with increasing computational demands during peak periods, pose significant challenges. Meanwhile, the unused computing resources of nearby Parked Vehicles (PVs) are often underutilized. To address these challenges, we propose a parked vehicle-assisted VEC architecture that integrates multiple TVs, Service Providers (SPs), and PVs. Specifically, we first design a Game-based Preoffloading Assignment (GPA) algorithm to determine the optimal SP for each TV. After SP selection, the interaction between each TV and its selected SP is modeled as a Stackelberg game, where the SP acts as the leader by setting the pricing strategy, and the TV acts as the follower by determining its offloading strategy. Moreover, we demonstrate the existence of a unique Nash equilibrium and propose a Gradient-based Stackelberg Pricing (GSP) algorithm to derive the equilibrium strategies. Additionally, to enhance the utilization of PV resources, we introduce an Optimized Recruitment Allocation (ORA) algorithm, enabling each PV to select the SP that maximizes its utility. Finally, extensive simulations demonstrate that the proposed scheme significantly improves offloading efficiency.

Vehicular Edge of Things Computing (VEoTC) schemes have emerged to enhance the Quality of Experience (QoE) of users requesting computational tasks. Such schemes exploit various computing resources embedded in Service Vehicles (SVs) to provide on demand computation. Typically, Vehicular Cloud Computing (VCC) and Vehicular Edge Computing (VEC) facilitate Internet of Vehicles (IoV) applications. However, the stringent latency guarantees required by such applications are challenging to satisfy given the high mobility and fluctuating densities of users in vehicular environments. Alternatively, VEoTC schemes have the potential to extend the computational coverage to areas with no or limited Roadside Unit (RSU) infrastructure and minimize the task latency as services are provided closer to the user. The SVs serve as mobile edge computing units providing computational services to vehicular users. In this paper, we propose a distributed task scheduling scheme to maximize the QoE of users in vehicular environments. The proposed technique facilitates task offloading to the SVs and/or RSUs through joint computing resource and channel allocation. Comparative experiments demonstrate that the proposed distributed approach enhances the QoE of users under different operational conditions.

Emerging intelligent transportation services are latency-sensitive with heavy demand for computing resources, which can be supported by a multi-tier computing system composed of vehicular edge computing (VEC) servers along the roads and micro servers on vehicles. In this work, we investigate the dual Uu/PC5 interface offloading and resource allocation strategy in Cellular Vehicle-to-Everything (C-V2X) enabled multi-tier VEC system. The successful transmission probability is characterized to obtain the normalized transmission rate of PC5 interface. We aim to minimize the system latency of task processing while satisfying the resource requirements of Uu and PC5 interfaces. Due to the non-convex and variables coupling, we decompose the original problem into two subproblems, i.e., resource allocation and offloading strategy subproblems. Specifically, we derive the closed-form expressions of packet transmit frequency of PC5 interface, transmission power of Uu interface, and CPU computation frequency in the resource allocation subproblem. Moreover, for the offloading strategy subproblem, the offloading ratio matrix is obtained by proposing the PC5 interface based greedy offloading (PC5-GO) algorithm, which concludes offloading decision and ratio. Simulation results are provided that the proposed PC5-GO algorithm can significantly improve the system performance compared with other baseline schemes by 13.7% at least.

High-definition (HD) map caching at roadside units (RSUs) is an important component of localization for self-driving vehicles, HD map content delivery services must be efficient for various self-driving vehicles. Nevertheless, HD maps are dynamic files that must be updated and replenished in real time. Developing an effective content delivery strategy for different types of self-driving vehicles to require HD maps, while ensuring safe driving and minimizing bandwidth consumption, is challenging. To maximize the monetary utility of the vehicle system, in this article, we jointly optimize the HD map update strategy and the wireless bandwidth resource allocation strategy, considering service delay constraints and overall system risk. However, the optimization problem described above is an NP-hard mixed-integer nonlinear programming (MINLP) problem. Additionally, in a self-driving vehicle scenario, the highly dynamic character of HD maps, the diversity of self-driving vehicle types, and the randomness of vehicle trajectories are unknown to the vehicle system in advance. The intractable optimization problem and the highly uncertain nature of the driving environment make it difficult to find an existing method that allows vehicles to obtain HD maps that meet their localization requirements in a timely manner. To address the above issues, we propose DRL-MURA, which can learn HD map updates and implement a wireless bandwidth resource allocation strategy by constantly interacting with environment based on a deep reinforcement learning (DRL) algorithm. Finally, we prove the accuracy and effectiveness of our method through simulation experiments.

No abstract available

Vehicular edge computing (VEC) is a promising technology to support real-time vehicular applications, where vehicles offload intensive computation tasks to the nearby VEC server for processing. However, the traditional VEC that relies on single communication technology cannot well meet the communication requirement for task offloading, thus the heterogeneous VEC integrating the advantages of dedicated short-range communications (DSRC), millimeter-wave (mmWave) and cellular-based vehicle to infrastructure (C-V2I) is introduced to enhance the communication capacity. The communication resource allocation and computation resource allocation may significantly impact on the ultra-reliable low-latency communication (URLLC) performance and the VEC system utility, in this case, how to do the resource allocations is becoming necessary. In this paper, we consider a heterogeneous VEC with multiple communication technologies and various types of tasks, and propose an effective resource allocation policy to minimize the system utility while satisfying the URLLC requirement. We first formulate an optimization problem to minimize the system utility under the URLLC constraint which is modeled by the moment generating function (MGF)-based stochastic network calculus (SNC), then we present a Lyapunov-guided deep reinforcement learning (DRL) method to convert and solve the optimization problem. Extensive simulation experiments illustrate that the proposed resource allocation approach is effective.

Vehicular edge computing (VEC) is emerging as a promising architecture of vehicular networks (VNs) by deploying the cloud computing resources at the edge of the VNs. However, efficient resource management and task offloading in the VEC network is challenging. In this work, we first present a hierarchical framework that coordinates the heterogeneity among tasks and servers to improve the resource utilization for servers and service satisfaction for vehicles. Moreover, we formulate a joint resource allocation and task offloading problem (JRATOP), aiming to jointly optimize the intra-VEC server resource allocation and inter-VEC server load-balanced offloading by stimulating the horizontal and vertical collaboration among vehicles, VEC servers, and cloud server. Since the formulated JRATOP is NP-hard, we propose a cooperative resource allocation and task offloading algorithm named BARGAIN-MATCH, which consists of a bargaining-based incentive approach for intra-server resource allocation and a matching method-based horizontal-vertical collaboration approach for inter-server task offloading. Besides, BARGAIN-MATCH is proved to be stable, weak Pareto optimal, and polynomial complex. Simulation results demonstrate that the proposed approach achieves superior system utility and efficiency compared to the other methods, especially when the system workload is heavy.

No abstract available

The efficient operation and interaction of autonomous robots play crucial roles in various fields, e.g., security, environmental monitoring, and disaster response. For these purposes, processing large volumes of sensor data and sharing data between robots is essential; however, processing such large data in an on-device environment for robots results in substantial computational resource demands, causing high battery consumption and heat issues. Thus, this study addresses challenges related to processing large volumes of sensor data and the lack of dynamic object information sharing among autonomous robots and other mobility systems. To this end, we propose an Edge-Driving Robotics Platform (EDRP) and a Local Dynamic Map Platform (LDMP) based on 5G mobile edge computing and Kubernetes. The proposed EDRP implements the functions of autonomous robots based on a microservice architecture and offloads these functions to an edge cloud computing environment. The LDMP collects and shares information about dynamic objects based on the ETSI TR 103 324 standard, ensuring cooperation among robots in a cluster and compatibility with various Cooperative-Intelligent Transport System (C-ITS) components. The feasibility of operating a large-scale autonomous robot offloading system was verified in experimental scenarios involving robot autonomy, dynamic object collection, and distribution by integrating real-world robots with an edge computing–based offloading platform. Experimental results confirmed the potential of dynamic object collection and dynamic object information sharing with C-ITS environment components based on LDMP.

The communication system is under a paradigm transformation that shifts from traditional bit-level transmission to semantic-level transmission. This transition lays the foundation for complex autonomous driving, necessitating instantaneous processing of substantial data within the constraints of computing capacity and communication bandwidth. In this paper, we propose a novel Task-oriented Source-Channel Coding (TSCC) framework that jointly optimizes source coding and channel coding in a task-oriented manner. Specifically, to reduce communication overhead and guarantee autonomous driving performance, we leverage an autonomous driving agent to guide source-channel coding based on a modified Conditional Variational Autoencoder (CVAE). We test the proposed framework on a well-known autonomous driving platform with different communication channel conditions. The results show that compared to traditional communication and state-of-the-art deep Joint Source-Channel Coding (JSCC), our proposed framework achieves superior performance by saving 98.36% communication overhead and maintains an 83.24% driving score even at 0 dB Signal-to-Noise Ratios (SNR).

With the rapid advancement of devices requiring intensive computation, such as Internet of Things (IoT) devices, smart sensors, and wearable technology, the computational demands on individual platforms with limited resources have escalated, necessitating the offloading of the generated tasks by the devices to edge. These tasks are often real-time with strict response time requirements. Among these devices, autonomous vehicles present unique challenges due to their critical need for timely and accurate processing to ensure passenger safety. Selecting suitable servers in a heterogeneous mobile edge computing (MEC) architecture is vital to optimizing real-time task processing rates for such applications. To address this, we present an algorithmic solution to improve the allocation of heterogeneous servers to real-time tasks, aiming to maximize the number of processed tasks. By analyzing task and server characteristics in the MEC architecture, we develop the suitability-based adaptive resource selection (SARS) algorithm, which evaluates server suitability based on factors like time constraints and server capabilities. Additionally, we introduce the proactive on-demand resource allocation (PORA) algorithm, which strategically reserves computational resources to ensure availability for critical real-time tasks. We compare the proposed algorithms with several classical and state-of-the-art algorithms. Computational results demonstrate that our approach outperforms existing algorithms, processes more tasks, and effectively prioritizes urgent tasks, particularly in autonomous driving applications.

The scientific community is focusing on developing 6G solutions beyond the 5G communications network, which will use Artificial Intelligence (AI) solutions at all levels of the communications architecture. The potential benefits of AI will automate and optimize network operation, boost cognitive capability and use a fully AI-enabled architecture. This paper proposes a framework for solving the current problem of orchestrating cooperative control of autonomous vehicles using 6G, which identifies new challenges, requirements, and improvements to deploy future cooperative Connected Autonomous Vehicles (CAV) successfully. The proposed architecture would enable edge-decentralized cooperative control of CAV using AI techniques and specialized AI hardware at different levels of the 6G communication architecture. The framework proposes vehicle management and control tasks to be performed in nodes called Vehicle Edge Nodes (VEN), which would be trained in a fine-grained way the diverse vehicle control systems and infer them, keeping latency to a minimum and providing high reliability for cooperative vehicular management and control using a set of nodes called PicoCells.

As the utilization of complex and heavy applications increases in autonomous driving, research on using mobile edge computing and task offloading for autonomous driving is being actively conducted. Recently, researchers have been studying task offloading algorithms using artificial intelligence, such as reinforcement learning or partial offloading. However, these methods require a lot of training data and critical deadlines and are weakly adaptive to complex and dynamically changing environments. To overcome this weakness, in this paper, we propose a novel task offloading algorithm based on Lyapunov optimization to maintain the system stability and minimize task processing delay. First, a real-time monitoring system is built to utilize distributed computing resources in an autonomous driving environment efficiently. Second, the computational complexity and memory access rate are analyzed to reflect the characteristics of the deep learning applications to the task offloading algorithm. Third, Lyapunov and Lagrange optimization solves the trade-off issues between system stability and user requirements. The experimental results show that the system queue backlog remains stable, and the tasks are completed within an average of 0.4231 s, 0.7095 s, and 0.9017 s for object detection, driver profiling, and image recognition, respectively. Therefore, we ensure that the proposed task offloading algorithm enables the deep learning application to be processed within the deadline and keeps the system stable.

No abstract available

Artificial intelligence (AI)-powered autonomous vehicles (AVs) can integrate different machine learning (ML) techniques to build up a complex autonomous driving system. However, single AV intelligence is not enough to cope with ever-changing driving environments. The underlying reason is that, with current neural network design and training algorithms, it is challenging for the driving model to generalize to diverse driving environments all at once due to sample inefficiency and the curse of dimensionality. Powerful computing resources and massive amount of data can be used to train a good driving model offline. However, the driving model obtained offline might fail in corner case scenarios. In this paper, we propose an intelligence networking framework among AVs assisted by multi-access edge computing (MEC) with end-to-end learning for demonstration. In this framework, driving road is divided into segments and data is collected for each road segment separately. Assisted by MEC networks, a continuously updated driving model is produced in near real-time for each road segment when the environment changes. By dividing the road into segments, we aim to reduce the burden of generalization since a single model only needs to adapt to a specific road segment. Simulation results show that our solutions can produce updated driving model for each road segment to adapt to environmental changes better than the existing scheme. Upcoming AVs can then adapt to changing environments by downloading the updated driving model.

Edge Computing and Network Function Virtualization (NFV) concepts can improve network processing and multi-resources allocation when intelligent optimization algorithms are deployed. Multiservice offloading and allocation approaches pose interesting challenges in the current and next-generation vehicle networks. The state-of-the-art optimization approaches still formulate exact algorithms, and tune approximation methods to get sufficient solutions. These approaches are data-centric that aim to use heterogeneous data inputs to find the near optimal solutions. In the context of connected and autonomous vehicles (CAVs), these techniques show an exponential computational time and deal only with small and medium scale networks. Therefore, we are motivated by using recent Deep Reinforcement Learning (DRL) techniques to learn the behavior of exact optimization algorithms while enhancing the Quality of Service (QoS) of network operators and satisfying the requirements of the next-generation Autonomous Vehicles (AVs). DRL algorithms can improve AVs service offloading and optimize edge resources. An Optimal Virtual Edge Autopilot Placement (OVEAP) algorithm is proposed using Integer Linear Programming (ILP). Moreover, an autopilot placement protocol is presented to support the algorithm. Optimal allocation and Virtual Network Function (VNF) placement and chaining of the autopilot, based on several new constraints such as computing and networking loads, network edge infrastructure, and placement cost, are designed. Further, a DRL approach is formulated to deal with dense Internet of Autonomous Vehicle (IoAV) networks. Extensive simulations and evaluations are carried out. Results show that the proposed allocation strategies outperform the state-of-the-art solutions and give better performance in terms of Total Edge Servers Utilization, Total Edge Servers Allocation Time, and Successfully Allocated autopilots.

As the level of autonomous driving(AD) increases, the number of sensors in a single vehicle increases, and the required computing power also increases rapidly. The intelligence level of AD depends on algorithms and hardware performance. However, due to the problems such as limited production cost or computing resources, some vehicles may not have enough computation power for high level of AD. To overcome such obstacles, the distributed AD system architecture based on edge computing has become a trend. We build an edge computing-based distributed AD system on top of Robot Operating System(ROS2) and measure the performance with different settings. Our test results show the feasibility of AD systems centered on edge computing units, which open new research directions for exploiting edge computing in a more diverse way.

The emergence of new vehicles generation such as connected and autonomous vehicles led to new challenges in the vehicular networking and computing managements to provide efficient services and guarantee the quality of service. The edge computing facility allows the decentralization of processing from the cloud to the edge of the network. In this paper, we design and propose an end-to-end, reliable and low latency communication architecture that allows the allocation of compute-intensive autonomous driving services, in particular autopilot, to shared resources on edge computing servers and improve the level of performance for autonomous vehicles. The reference architecture is used to design an Advanced Autonomous Driving (AAD) communication protocol between autonomous vehicles, edge computing servers, and the centralized cloud. Then, a mathematical programming approach using Integer Linear Programming (ILP) is formulated to model the autopilot chain resources Offloading at the network edge. Further, a deep reinforcement learning (DRL) approach is proposed to deal with dense Internet of Autonomous Vehicle (IoAV) networks. Moreover, several scenarios are considered to quantify the behavior of the optimization approaches. We compare their efficiency in terms of Total Edge Servers Utilization, Total Edge Servers Allocation Time, and Successfully Allocated Edge Autopilots.

To simultaneously enable multiple autonomous driving services on affordable embedded systems, we designed and implemented LoPECS, a Low-Power Edge Computing System for real-time autonomous robots and vehicles services. The contributions of this paper are three-fold: first, we developed a Heterogeneity-Aware Runtime Layer to fully utilize vehicle’s heterogeneous computing resources to fulfill the real-time requirement of autonomous driving applications; second, we developed a vehicle-edge Coordinator to dynamically offload vehicle tasks to edge cloudlet to further optimize user experience in the way of prolonged battery life; third, we successfully integrated these components into LoPECS system and implemented it on Nvidia Jetson TX1. To the best of our knowledge, this is the first complete edge computing system in a production autonomous vehicle. Our implementation on Nvidia Jetson demonstrated that it could successfully support multiple autonomous driving services with only 11 W of power consumption, and hence proves the effectiveness of the proposed LoPECS system.

A key challenge for autonomous driving is to process a massive amount of sensor data and make safe and reliable decisions in real time. However, autonomous vehicles often have insufficient onboard resources to provide the required computation capacity. To address this problem, this article advocates a novel approach to offload computation-intensive autonomous driving services to roadside units and cloud for swift executions. Our approach combines an integer linear programming (ILP) formulation for offline optimization of the scheduling strategy and a fast heuristics algorithm for online adaptation. We verify our technique with both synthetic task graphs and real-world deployment. The experimental results show that our approach can improve system performance effectively.

Abstract Autonomous driving has received widespread attention in recent years, while the limited battery life and computing capability of autonomous vehicles cannot support some necessary computation-intensive and urgent tasks with strict response time requirements. The results of the tasks would be useless and may cause serious safety hazards if the given time constraints are exceeded. On the other side, mobile edge computing (MEC) offers the possibility of autonomous vehicles to complete these time-constraint tasks due to its proximity and strong computing capabilities, with the faster 5G wireless networks to enable vehicles and MEC servers to exchange data in milliseconds. Then, it is a key issue to make the MEC servers to execute and complete these time-constraint autonomous-driving tasks as many as possible. So, we propose a task scheduling algorithm that can consider characteristics of autonomous-driving tasks and select more suitable MEC servers with task migration, based on an improved earliest deadline first algorithm through the replacement and recombination of tasks. From the experimental results, it can be concluded that the algorithm can schedule more tasks and benefit the urgent tasks effectively with the increase of the task amounts.

Autonomous driving has become the future direction of intelligent transportation system. To enable autonomous driving, mobile edge computing is proposed for the task uploaded from the cars. Using lane line detection as a typical scenario, we build a decision-making mathematical model, and design a working strategy for edge computing. The model will decide the information to be uploaded to the server or processed in the vehicle. We also carry out simulation comparison according to the existing communication conditions, so as to obtain a better decision-making strategy model to improve the accuracy and response speed of lane line detection.

This paper presents Edge-based Mixture of Experts (MoE) Collaborative Computing (EMC2), an optimal computing system designed for autonomous vehicles (AVs) that simultaneously achieves low-latency and high-accuracy 3D object detection. Unlike conventional approaches, EMC2 incorporates a scenario-aware MoE architecture specifically optimized for edge platforms. By effectively fusing LiDAR and camera data, the system leverages the complementary strengths of sparse 3D point clouds and dense 2D images to generate robust multimodal representations. To enable this, EMC2 employs an adaptive multimodal data bridge that performs multi-scale preprocessing on sensor inputs, followed by a scenario-aware routing mechanism that dynamically dispatches features to dedicated expert models based on object visibility and distance. In addition, EMC2 integrates joint hardware-software optimizations, including hardware resource utilization optimization and computational graph simplification, to ensure efficient and real-time inference on resource-constrained edge devices. Experiments on open-source benchmarks clearly show the EMC2 advancements as an end-to-end system. On the KITTI dataset, it achieves an average accuracy improvement of 3.58% and a 159.06% inference speedup compared to 15 baseline methods on Jetson platforms, with similar performance gains on the nuScenes dataset, highlighting its capability to advance reliable, real-time 3D object detection tasks for AVs. The official implementation is available at https://github.com/LinshenLiu622/EMC2.

Abstract Mobile edge computing for autonomous driving needs to manage heterogeneous resources and process large amounts of data or multi-purpose payload. There needs to be deploying, scheduling and migrating tasks on edge nodes to ensure the reliability of tasks or maximize the utilization of resources. However, applying autonomous learning methods on autonomous driving is exceptionally difficult, due to the complexity of multi-dimensional context and the sensitivity to hyperparameters. In this paper, we propose a learning approach to quality-of-service (QoS) prediction of services via multi-dimensional context, and develop a stable approach for service deployment that requires minimal hyperparameter tuning and a modest number of trials to learn multilayer neural network policies. This approach can automatically trades off exploration against exploitation by automatically tuning hyperparameter based on maximum entropy reinforcement learning. We then demonstrate that this approach achieves state-of-the-art performance on Autoware benchmark environments.

Abstract For autonomous driving area, the communication among autonomous vehicles is quite frequent and inevitable to perceive the surrounding environment. Massive data needs to be processed in time on the edge to control vehicles to make suitable strategies. For a group of vehicles, collaborative strategies are necessary to avoid unintentional congestions. However, less attention is paid to competition and cooperation between vehicle groups. In this paper, we explore the possibility of adopting game theory for decision making to obtain win-win between autonomous vehicles. The Lemke-Howson algorithm in game theory is the best known combinatorial algorithm that computes a Nash equilibrium of a bimatrix game. More importantly, we implement the Lemke-Howson algorithm on FPGA to accelerate the computation process. We explain the design challenges of solving the performance bottleneck and how to make optimizations. We implement the Lemke-Howson accelerator on a KCU116 board and obtain a speedup of about 2.4 times versus that running on a CPU.

Abstract Autonomous driving, which can free our hands and feet, is of increasing importance in our daily lives. However, the capacity of onboard computation and communication limits the rapid development of autonomous driving. To address this issue, this paper proposes a novel model named the edge computing-based lanes scheduling system (ECLSS) to study lane scheduling for each vehicle around crossroads with real-time edge devices. There are several edge computing devices (ECD) deployed around crossroads in ECLSS that can collect information from vehicles and road conditions with short-range wired/wireless transmissions. Based on the strong computing power of ECDs and their real-time transmission performance, we propose two centralized management lane scheduling methods: the searching for efficient switching algorithm (SESA) and special vehicles lane switching algorithm (SVLSA). These edge computing-based autonomous driving methods aim to achieve high speed passing through crossroads and guarantee that special vehicles can pass through crossroads within a certain time. Extensive simulations are conducted, and the simulation results demonstrate the superiority of the proposed approaches over competing schemes in typical lane switching scenarios.

To simultaneously enable multiple autonomous driving services on affordable embedded systems, we designed and implemented {\pi}-Edge, a complete edge computing framework for autonomous robots and vehicles. The contributions of this paper are three-folds: first, we developed a runtime layer to fully utilize the heterogeneous computing resources of low-power edge computing systems; second, we developed an extremely lightweight operating system to manage multiple autonomous driving services and their communications; third, we developed an edge-cloud coordinator to dynamically offload tasks to the cloud to optimize client system energy consumption. To the best of our knowledge, this is the first complete edge computing system of a production autonomous vehicle. In addition, we successfully implemented {\pi}-Edge on a Nvidia Jetson and demonstrated that we could successfully support multiple autonomous driving services with only 11 W of power consumption, and hence proving the effectiveness of the proposed {\pi}-Edge system.

Mobile edge computing has been proposed as a solution for solving the latency problem of traditional cloud computing. In particular, mobile edge computing is needed in areas such as autonomous driving, which requires large amounts of data to be processed without latency for safety. Indoor autonomous driving is attracting attention as one of the mobile edge computing services. Furthermore, it relies on its sensors for location recognition because indoor autonomous driving cannot use a GPS device, as is the case with outdoor driving. However, while the autonomous vehicle is being driven, the real-time processing of external events and the correction of errors are required for safety. Furthermore, an efficient autonomous driving system is required because it is a mobile environment with resource constraints. This study proposes neural network models as a machine-learning method for autonomous driving in an indoor environment. The neural network model predicts the most appropriate driving command for the current location based on the range data measured with the LiDAR sensor. We designed six neural network models to be evaluated according to the number of input data points. In addition, we made an autonomous vehicle based on the Raspberry Pi for driving and learning and an indoor circular driving track for collecting data and performance evaluation. Finally, we evaluated six neural network models in terms of confusion matrix, response time, battery consumption, and driving command accuracy. In addition, when neural network learning was applied, the effect of the number of inputs was confirmed in the usage of resources. The result will influence the choice of an appropriate neural network model for an indoor autonomous vehicle.

No abstract available

Autonomous vehicles and mobile edge computing’s confluence have raised an innovative model for immediate decision-making and improved computational abilities. But, enhancing vehicle management systems to guarantee effective enactment remains an important challenge. Present approaches regularly depend on intricate algorithms and multiple sensors, that result in improved computational overhead and potential latency. The current study resolves the present gap by offering a new hybrid framework, which synergistically mixes optimization algorithms and deep neural networks through the advantages of mobile edge computing. Precisely, this research presents a hybrid model for autonomous vehicle management by integrating a refined version of the RPO or Pelican Optimizer with deep neural networks attuned to mobile edge computing environments. The chief contributions of the present study have been threefold: (1) the improvement of a particular autonomous driving method optimized for mobile edge computing platforms; (2) the arrangement of an optimized MobileNet method employing the RPO algorithm that uses LiDAR sensor data for effective object recognition and path design; and (3) the construction of an indoor vehicle prototype by mean of a microcontroller and LiDAR sensors, after a comprehensive performance evaluation of inference models, and analyzing the trade-offs between input size and computational effectiveness. Experimental outcomes show the efficiency and reliability of the suggested hybrid model, through improving autonomous vehicle management and decision-making abilities within the mobile edge computing paradigm. The current study contributes to the enhancement of autonomous vehicle research and provides an innovative solution for effective and precise vehicle control within edge computing environments.

For intelligent connected vehicles, trajectory planning faces challenges such as high latency, heavy computational load, and delayed responses. This paper addresses these core issues by deploying edge computing nodes, collaboratively allocating tasks, and constructing a hierarchical architecture. By integrating edge computing collaboration mechanisms, layered design, and parallel computing technologies, the proposed method effectively resolves path planning and obstacle avoidance coordination challenges. Simulation results show that this approach significantly reduces trajectory generation latency, improves path planning response speed, and enhances obstacle avoidance success rates, demonstrating strong real-time performance and feasibility. This study provides reliable technical support for efficient decision-making in autonomous driving systems.

Object detection in real-time would play a critical role in safe navigation, evasion of obstacles, and decision making of autonomous cars. However, the latency and the bandwidth limitations imposed by the established cloud-based processing paradigm degrade the sensitivity and reliability required to perform autonomous driving. In this study, an advanced technique of real-time object detection that uses edge computing, applicable to driverless vehicles, is offered. This drastically reduces the time spent in inference spatially but maintains promising levels of detection accuracy by offloading compute-intensive work to edge nodes, which are deployed closer to the vehicle, for example, at roadside units or onboard processors. In order to guarantee effective processing under various network and environmental conditions, the suggested system incorporates lightweight deep learning models optimized for edge hardware, using adaptive model compression and intelligent workload distribution. Experimental tests show that, without sacrificing detection accuracy, our method can reduce latency by up to 40% and boost energy economy by up to 25% when contrasted to traditional cloud-dependent methods. The system is a good option for practical implementation since it also demonstrates strong scalability and flexibility in dynamic driving situations. This study highlights how edge computing, which strikes a mix amongst speed, accuracy, and resource efficiency, has the possibility to revolutionize next-generation autonomous car systems.

No abstract available

Connected and automated vehicles (CAVs) have emerged as an efficient solution to improve the driving experience in the intelligent transportation systems (ITSs), in which the targeted vehicle (TV) can switch between the human-driven (HD) and autonomous-driven (AD) modes to act as server or terminal in vehicular edge computing networks (VECNs). However, due to the dynamic nature of traffic networks and the moving of vehicles, distribution of computational resources is imbalanced and variable, it is a challenge to design the cooperative resource management scheme for the whole journey of vehicle users. In this article, we propose a joint driving model selection and resource management scheme for TV in each road segment, to maximize the vehicle users’ satisfaction of the whole journey. For the complex formulated joint optimization problem, we design a three-stage hierarchical optimization (3SHO) framework, using deep Q-network (DQN) for driving mode optimization in the first stage and deep deterministic policy gradient (DDPG) for optimizing resource management under different selected driving modes. And a terminal-server matching mechanism is introduced to enable dynamic service quality improvement for TV. Specially, we design a new user satisfaction function with the quality of service, traffic revenue, and the gap between expected and actual revenues of users are considered. Experimental results showcase the robust convergence of the 3SHO algorithm, the adeptness to dynamic traffic networks, and the capacity to enhance user satisfaction significantly.

This article explores the critical role of three foundational technologies that enable the safe and efficient operation of autonomous vehicles: security systems, observability solutions, and edge computing infrastructure. It examines how these technologies work in concert to address the unique challenges of self-driving cars, including threat detection and mitigation, performance monitoring, and real-time data processing. The article highlights how telecommunications providers are integrating these capabilities into their networks to create the necessary infrastructure for autonomous vehicle ecosystems. By detailing the technical aspects and performance metrics of these systems, the article demonstrates how advances in these areas are driving progress toward widespread adoption of self-driving technology while ensuring operational safety, reliability, and efficiency.

This paper introduces a CNN-RNN hybrid architecture of edge-optimized system to detect and track real-time objects in self-driving cars. The architecture combines convolutional neural networks with spatial feature detectors and recurrent layers with a temporal sequence detector, which allows it to detect effectively in dynamic settings. KITTI and nuScenes are preprocessed to guarantee that the input is of high quality, and edge deployment is more efficient in the context of limited hardware. The experimental findings show that the suggested model has 96.8% accuracy, has a shorter inference time of 95 ms, and ID switches are 20 percent lower than the current models. Comparison of performance between vehicles and vulnerable road users indicates that there are consistent increases in performance at increased levels of the IOU, which justifies its versatility in the handling of different traffic conditions. Also, model compression algorithm plays a very important role in minimizing deployment overheads, which makes computing edge viable. It is a hybrid solution, which is scalable and practical in terms of intelligent transportation systems, enhancing safety and efficiency in autonomous driving.

Integrating large language models (LLMs) into autonomous driving enhances personalization and adaptability in open-world scenarios. However, traditional edge computing paradigm still faces significant challenges in processing complex driving data, particularly regarding real-time performance and system efficiency. To address these challenges, this study introduces EC-Drive, a novel edge-cloud collaborative autonomous driving system with data drift detection capabilities. EC-Drive utilizes drift detection algorithms to selectively upload critical data, including new obstacles and traffic pattern changes, to the cloud for processing by GPT-4, while routine data is efficiently managed by smaller models on edge devices. This approach not only reduces inference latency but also improves system efficiency by optimizing communication resource usage. Experimental validation confirms the system’s robust processing capabilities and practical applicability in real-world driving conditions, demonstrating the effectiveness of this edge-cloud collaboration framework.

The growth of autonomous driving technology is accelerating. However, complete autonomous driving has not been implemented yet. This paper proposes a multi-camera interoperable emulation framework for developing autonomous vehicle driving. We implement two components of the Advanced Driver Assistance System (ADAS). The vehicle adjusts its speed based on the distance from the object and stays in its lane. Smart Cruise Control (SCC) and Lane Keeping Assist (LKA) are. These two systems are remotely controlled in our framework. As a result, developing, applying, and simulating algorithms will be more convenient, and this can protect drivers from accidents caused by incomplete algorithms during simulations. Moreover, these systems can relieve the drivers' burden and fatigue during real driving and prevent dangerous situations that can occur due to other vehicles or pedestrians.

In the field of autonomous driving, common challenges include difficulties in detecting small vehicles and pedestrians on the road, high computational demands of algorithms, and low accuracy of detection algorithms. This paper proposes a YOLOv8n-FAWL object detection algorithm tailored for edge computing, incorporating the following three improvements: (1) The Faster-C2f-EMA module is created, designed through the synergy of the FasterNet architecture and the concept of EMA modules, effectively addressing the challenge of suboptimal feature extraction for small objects. (2) The WIOU loss function is adopted to resolve the issue of imbalanced training samples. (3) The LAMP pruning technique is applied to reduce the model parameters and complexity, thereby enhancing the overall model accuracy. The experimental results show that compared to the baseline model, the proposed algorithm achieves improvements of 6.2% and 4.5% in the mAP@0.5, and 3.8% and 2.7% in the mAP@0.5:0.95, on the Udacity and BDD100K-tiny datasets,respectively. In addition, the model parameters we’re reduced by 49.2% and 46%. The model achieved real-time performance at 54 FPS, thereby advancing the development of autonomous driving technology.

No abstract available

Abstract With the development of science and technology, the realization of intelligent vehicle autonomous driving has become a popular research. In order to realize the optimal autonomous driving behavior, this paper proposes a vehicle-splittable task offloading algorithm for edge offloading in the IoT environment based on the study of edge computing technology and the integration of 5G communication. On this basis, a task offloading model is built by combining Markov decision learning algorithm in order to improve the ability of real-time computation of intelligent vehicles in real driving environment and, at the same time, to realize the recording of dynamic driving behavior of the vehicle and the prediction of intelligent and safe paths in the dynamic IoV environment. Simulation experiments verify the performance of the VPEO algorithm. When there are varying numbers of vehicles and tasks, the VPEO algorithm performs better according to the experimental results. After 750 iterations, the VPEO algorithm achieves a 94% success rate when calculating tasks and stabilizes. In the intelligent warning experiments, the VPEO algorithm measured the average accuracy of lane offset distance, the average judgment correctness of vehicle class, and vulnerable traffic participant in relation to the position of the lane line were 84.66%, 92.26% and 94.69%, respectively. The VPEO algorithm can perform real-time calculations of intelligent driving information, and warnings about intelligent driving safety can be provided.

No abstract available

Smart grid delivers power with two-way flows of electricity and information with the support of information and communication technologies. Electric vehicles (EVs) with rechargeable batteries can be powered by external sources of electricity from the grid, and thus charging scheduling that guides low-battery EVs to charging services is significant for service quality improvement of EV drivers. The revolution of communications and data analytics driven by massive data in smart grid brings many challenges as well as chances for EV charging scheduling, and how to schedule EV charging in a smart and resilient way has inevitably become a crucial problem. Toward this end, we in this paper leverage the techniques of software defined networking and vehicular edge computing to investigate a joint problem of fast charging station selection and EV route planning. Our objective is to minimize the total overhead from users’ perspective, including time and charging fares in the whole process, considering charging availability and electricity price fluctuation. A deep reinforcement learning (DRL) based solution is proposed to determine an optimal charging scheduling policy for low-battery EVs. Besides, in response to dynamic EV charging, we further develop a resilient EV charging strategy based on incremental update, with EV drivers’ user experience being well considered. Extensive simulations demonstrate that our proposed DRL-based solution obtains near-optimal EV charging overhead with good adaptivity, and the solution with incremental update achieves much higher computation efficiency than conventional game-theoretical method in dynamic EV charging.

Vehicular edge computing (VEC) is considered to be a key technology to improve the processing efficiency of computing tasks for the Internet of Vehicles (IoV). Using roadside units (RSUs) distributed on both sides of a road as edge servers, computation-intensive and latency-sensitive in-vehicle tasks can be responded to quickly. However, some Quality of Service (QoS) is often difficult to ensure due to clogged dense urban buildings or lack of infrastructure in remote areas. In this article, we propose a software-defined network (SDN)-driven partial offloading model for unmanned aerial vehicle (UAV)-assisted VEC networks, where the RSUs and UAVs jointly provide computing services to the vehicles and collect global information through centralized control using an SDN controller. To guarantee these vehicles obtain computing results in time and rationally utilize computing resources, we develop an optimal offloading mechanism using Age of Information (AoI), together with energy consumption and rental price as a comprehensive weighted cost of our above optimization objective. The total system cost of the performing tasks is minimized by jointly optimizing the UAV trajectory, user association, and offloading decision. Considering the mobility of the vehicles and UAVs and the dynamic network environment, we design a deep reinforcement learning (DRL)-based joint trajectory control and offloading allocation algorithm (DRL-TCOA) to solve the proposed computation offloading problem. Experimental results show that the proposed DRL-TCOA algorithm maintains better information freshness and lower system cost than the other baseline offloading strategies.

Vehicular Edge Computing (VEC) allows vehicles to offload their delay-sensitive tasks to nearby Road Side Units (RSUs) for processing, which improves network quality of service (QoS). However, the self-interested SDN controller is unwilling to ask RSUs to provide free computing resources for vehicles. At the same time, complicated dependencies between vehicular subtasks may cause non-ideal task delay and energy consumption. In order to solve these problems, this paper proposes a Stackelberg game-based Dependency-aware task Offloading and resource Pricing framework (SDOP). Specifically, we first model a vehicular edge network that partially offloads dependency-aware tasks. Then, we depict the interaction between the SDN controller and vehicles as a Stackelberg game, with the goal of maximizing the utility of both parties. Next, we present a Gradient Ascent Plus Genetic algorithm (GAPG) to solve the problem. Finally, numerous simulations are performed, and the results show that compared with other baseline schemes, the proposed GAPG can significantly improve the utility of both the SDN controller and vehicles under various scenarios.

The development of the Internet of Vehicles (IoV) has made people’s lives and travels safer, more efficient, and more comfortable. The combination of edge computing and IoV can provide processing and storage capabilities close to vehicles, thus becoming a potential paradigm. At this time, the software-defined networking (SDN) architecture is extremely necessary to realize centralized control and convenient management for complex and dynamic vehicular edge networks. However, as the brain of the SDN architecture, little attention has been paid to the security of the SDN controller. Once the controller is threatened, severe global chaos may happen. Therefore, in this article, we study the attack against the SDN controller, which is the topology poisoning attack. We successfully implement this attack in four mainstream controllers and analyze its impact from multiple levels. To the best of our knowledge, we are the first to study this attack in the vehicular edge network. In addition, in view of the counter-attacks of the existing defence mechanisms, we propose an attack-tolerance scheme based on deep reinforcement learning (DRL) to enhance the vehicular edge network with a certain degree of self-recovery.

In vehicular edge computing (VEC), the execution of offloading task needs not only the task data uploaded by the requesting vehicle, but also the additional data to support the task to be executed successfully, and how to efficiently cache and access these supporting data becomes the key issue for task offloading in VEC. In this paper, we study the efficient caching mechanism to minimize the acquisition delay of the supporting data. Firstly, with the software defined network (SDN) based VEC framework, we analyze the acquisition ways of the supporting data and the caching collaboration between VEC servers. Then, according to the density of the requesting vehicles, we divide the VEC coverage into dense and ordinary areas. With the consideration of the similarity of the requested data and the distance between edge servers, the edge servers are clustered into multiple groups based on K-mean++ algorithm. Finally, each server's storage space is divided into three partitions, and the most beneficial data for itself, its group and the whole system are respectively stored in these partitions. Based on service area dividing, server grouping and storage space partitioning, we propose an efficient edge-cloud collaborative caching strategy, which can reduce the delay of data migration while task execution. Simulation results show that, compared with other schemes, the proposed caching strategy has better performance in terms of average data migration delay and application QoS.

This paper proposes a Stacklberg game-based Dependent task Offloading and resource Pricing framework (SDOP), where vehicles partially offload their dependent substaks to the SDN controller and pays corresponding fees. Firstly, we model the interaction between the SDN controller and vehicles as a Stackelberg game, where both parties wish to maximize their utility. Then, we employ the backward induction approach to analyze the investigated problem, and prove the existence and uniqueness of Nash and Stackelberg equilibrium. Next, we propose a Gradient Ascent Plus Genetic algorithm (GAPG) to solve the considered problem. Finally, extensive simulation results show that the proposed GAPG outperforms other baseline schemes under various scenarios.

With the emergence of computation-intensive vehicular applications, vehicular edge computing (VEC) offers a new paradigm to augment the capabilities of vehicles. In this article, we study the problem of dependency-aware task offloading and service caching in VEC, where each application can be divided into multiple tasks with task dependency, and vehicles can access the software-defined network (SDN) via roadside units (RSUs) to request edge servers to assist in processing tasks. Edge servers can selectively cache executed services for reuse by subsequent tasks to improve the offloading efficiency of the system. The offloading efficiency is defined as a weighted sum of the computation time of the task and the energy drained from the respective vehicle. To maximize the offloading efficiency, we formulate the tasks offloading and service caching problem as a mixed-integer non-linear programming (MINLP) problem and develop a semi-distributed algorithm based on dynamic programming to solve the optimization problem. Simulation results demonstrate that the proposed algorithm has higher offloading efficiency and a higher completion rate under application deadlines compared with similar offloading algorithms.

Recently, the rapid advance of vehicular networks has led to the emergence of diverse delay-sensitive vehicular applications such as automatic driving, auto navigation. Note that existing resource-constrained vehicles cannot adequately meet these demands on low / ultra-low latency. By offloading parts of the vehicles’ compute-intensive tasks to the edge servers in proximity, mobile edge computing is envisioned as a promising paradigm, giving rise to the vehicular edge computing networks (VECNs). However, most existing works on task offloading in VECNs did not take the load balancing of the computation resources at the edge servers into account. To address these issues and given the high dynamics of vehicular networks, we introduce fiber-wireless (FiWi) technology to enhance VECNs, due to its advantages on centralized network management and supporting multiple communication techniques. Aiming to minimize the processing delay of the vehicles’ computation tasks, we propose a software-defined networking (SDN) based load-balancing task offloading scheme in FiWi enhanced VECNs, where SDN is introduced to provide supports for the centralized network and vehicle information management. Extensive analysis and numerical results corroborate that our proposed load-balancing scheme can achieve superior performance on processing delay reduction by utilizing the edge servers’ computation resources more efficiently.

Federated edge learning (FEEL) emerges as a privacy-preserving paradigm to effectively integrate edge computing for the implementation of deep learning-based vehicular applications. Nevertheless, the incentive mechanism for vehicles participating in varied learning tasks, has not been well explored yet. In this paper, software-defined network (SDN) technology is adopted for the training control among vehicles, and a novel FEEL framework, namely SDN-assisted semi-decentralized FEEL (SSD-FEEL) is investigated, where multiple edge servers collectively coordinate a large number of vehicular models from different learning tasks. By exploiting the low-cost and similar learning tasks among vehicles and edge servers, SSD-FEEL incorporates more training samples, while enjoying the flexibility of edge server assisted model aggregation. Aiming at motivating vehicles to actively participate in training while improving the model accuracy of multiple learning tasks, a joint contract design and task reorganization problem, combined with the evaluation of model convergence and contract performance, is formulated. Then, we propose a two-stage optimization algorithm incorporating iterative reward allocation and task matching, where the model parameters in different tasks are reconstructed according to the matching results with the mobility constraints. Extensive experiments conducted on multiple data sets validate that the proposed algorithm can achieve higher cluster utility and outperform the conventional multi-task FEEL schemes in terms of learning performance.

Cloud computing has emerged as one of the popular technologies which provide on-demand services to the end users. Such services are hosted by massive geo-distributed data centers (DCs). Nowadays, connected vehicles in a smart city can also avail cloud services through Internet using cellular technologies. But, the advent of 5G technology has posed challenges for DCs such as-low latency and higher data rate requirements. To handle these challenges, edge-DCs (EDCs) can be deployed across a smart city to provide low latency services to the connected vehicles. In lieu of this, in this paper, EDCSuS: Sustainable EDC as a service framework in software defined vehicular environment is proposed. In EDCSuS, first, a software defined controller handles the incoming requests and suggest an optimal flow path. Second, a multi-leader multi-follower Stackelberg game is presented for resource allocation. Third, to improve the resource utilization, a cooperative resource sharing scheme is designed, thereby minimizing the energy consumption of servers in the EDCs. Lastly, a caching scheme is presented to avert excessive energy consumption for retracing the lost link due to vehicular mobility. The efficacy of the proposed scheme has been evaluated using extensive simulations with respect to various parameters. The results obtained prove the effectiveness of EDCSuS.

With the surge in the demand for online services and multimedia applications, the traffic on the underlying network infrastructure has escalated (multi-folded) in recent years. To meet the strict latency requirements, Software-defined Networking (SDN) provides flexible network control (and possible intelligence) that can act as an enabler for application-oriented service industry. However, the crippling gap between the business needs and the network delivery potential necessitates the underlying network to constantly (and consistently) adapt, protect, and inform across all strands of the service-oriented landscape. Intent-based network has emerged as a recent solution to the cover the above gap by capturing business intent and thereafter activating and assuring it networkwide. Motivated from these facts, in this article, an Intent-based network control framework has been designed over the SDN architecture for data dissemination in the vehicular edge computing ecosystem. In this framework, a tensor-based mechanism is used to reduce the dimensionality of the incoming elephant-like traffic and then classifying the specific-attribute data traffic according to the defined priority requirement of the underlying applications. Here, the network policies are configured using the intent-based controller according to the application requirement and then forwarded to the SDN controller to enable intelligent data dissemination (through an optimal route) at the data plane. Convolution Neural Network is used to train the flow table to allocate the route dynamically for the classified traffic queues. The proposed framework has been evaluated through extensive simulations and the results supports the claims in terms of the quality of service requirements.

Enabling HD-map-assisted cooperative driving among CAVs to improve navigation safety faces technical challenges due to increased communication traffic volume for data dissemination and an increased number of computing/storing tasks on CAVs. In this article, a new architecture that combines MEC and SDN is proposed to address these challenges. With MEC, the interworking of multiple wireless access technologies can be realized to exploit the diversity gain over a wide range of radio spectrum, and at the same time, computing/storing tasks of a CAV are collaboratively processed by servers and other CAVs. By enabling NFV in MEC, different functions can be programmed on the server to support diversified AV applications, thus enhancing the server's flexibility. Moreover, by using SDN concepts in MEC, a unified control plane interface and global information can be provided, and by subsequently using this information, intelligent traffic steering and efficient resource management can be achieved. A case study is presented to demonstrate the effectiveness of the proposed architecture.