异构卫星任务规划

No abstract available

The increasing demand for satellite communication necessitates efficient resource management, especially as next-generation high-throughput satellites (HTSs) face challenges in optimizing user connections. This paper presents a novel method that integrates a discretely improved Particle Swarm Optimization (PSO) algorithm with a dynamic task management framework to enhance satellite resource allocation efficiency. The PSO algorithm is adapted for discrete selection problems to maximize a fitness function based on user priority, user preference, and capacity satisfaction, thus ensuring accurate user–satellite matching. A main loop function iteratively updates user data and connection statuses, thus achieving continuous optimization of satellite connections. Through simulation studies, we validated the effectiveness of this method under dynamically changing user demands, demonstrating that the Discrete PSO algorithm significantly outperforms Simulated Annealing (SA) and the Genetic Algorithm (GA) in user satisfaction, maintaining levels above 0.996 even under high demand. Additionally, it effectively prioritizes high-demand users, ensuring their satisfaction remains above 0.95. Overall, our method enhances the management of daily communication tasks, significantly improving service quality and user satisfaction.

Complex heterogeneous scenarios with multiple mission requirement relationships, poor model scalability, resource conflicts during mission planning are the serious challenges currently facing the field of multi-satellite imaging mission planning (MSIMP). To solve this difficult problem, this paper proposes an Objective task-matching strategy and Improved adaptive differential evolution algorithm (OTMS-IADE). Firstly, the target task matching strategy for MSIMP in complex heterogeneous scenarios is constructed for multi-user, multi-satellite and multi-task situations, which overcomes the problem of poor scalability of the planning model in complex heterogeneous scenarios, and reduces the loss of resources caused by inappropriate task allocation; Secondly, to address the problem of low execution efficiency and long planning time due to large MSIMP solution space and complex constraints in complex heterogeneous scenarios, an improved adaptive differential evolution algorithm is proposed to reasonably trade-off the spatial search performance and the spatial exploitation performance to enhance the algorithm solution efficiency. Simulation experiments show that the OTMS-IADE algorithm for processing complex heterogeneous scenarios MSIMP has obvious advantages regarding task importance optimization and timeliness.

Earth Observation Satellites (EOS) are critical for acquiring space-based information, supporting diverse applications from environmental monitoring to urban planning. The increasing demand for satellite imaging services necessitates efficient planning to handle complex and heterogeneous observation tasks. Traditional mission planning approaches, often relying on “single objective, single model, single algorithm” paradigms with meta-tasks and time windows, struggle to integrate these diverse requirements and face scalability issues with increasing task complexity. To address these limitations, this paper proposes a novel Grid Heat-driven Imaging Satellite Complex Task Planning Method. We introduce a Grid Heat-based Requirements Observation Model (GHROM) that unifies point, area, and moving targets into a single, time-variant geospatial heatmap, where grid cell “heat” represents observation priority. For this gridbased representation, we design a Satellite Discrete State Task Planning Model that transforms the problem into finding the highest heat gain path within a Directed Acyclic Graph (DAG) of discrete satellite states, where attitude transition constraints are embedded as graph edges. Based on this model, we develop a Grid Heat-Driven Dynamic Programming Algorithm (GMDPA). Extensive simulation experiments demonstrate the effectiveness and significant advantages of GMDPA, particularly in largescale scenarios, showcasing its superior performance in terms of solution quality and computational efficiency compared to traditional heuristic, metaheuristic, and genetic algorithms.

As aerospace technology continues to develop, more and more companies and institutions are launching their own satellite constellations, forming heterogeneous satellite networks in space controlled by different users. Due to the satellites being heterogeneous and controlled by different entities, it is difficult to establish a mechanism for collaborative mission planning, leading to a waste of satellite resources. Addressing this issue, this paper begins with the interests of different controlling users as a starting point. Firstly, a cooperative game theory model with controlling users as participants is constructed. Taking into account the characteristics and constraints of the multiple satellite collaborative mission planning problem, a characteristic function for the cooperative game model is designed. Subsequently, a solution framework for the cooperative game model based on the core and a hybrid particle swarm optimization algorithm is proposed. A hybrid particle swarm algorithm that integrates a genetic algorithm crossover operator and hill-climbing strategy is designed. Moreover, to address the issue that coalitions may not always have a core, an algorithm for finding coalitions that do have a core is devised. Finally, a series of experiments are designed to verify the effectiveness of the model and algorithm. Simulation results show that the cooperative game model can allocate tasks and profits effectively, thereby establishing a robust collaborative mission planning mechanism among multiple heterogeneous agile satellites. Furthermore, by comparing with the traditional particle swarm optimization algorithm, the effectiveness of the proposed hybrid particle swarm algorithm is demonstrated.

This paper proposes a risk-aware framework for Safe Multi-Agent Planning (SafeMAP) that unifies heterogeneous models for multi-agent systems in a Markovian process that allows for simultaneous system health monitoring, decision making under uncertainty, and multi-agent system collaboration. As operations beyond low earth orbit mature, there is an increased need for autonomous cyber-physical systems with onboard decision making capabilities. Multi-agent cyber-physical systems in particular offer the potential of increased efficiency, resiliency, and mission capabilities for future applications such as multi-rover terrain operations, distributed satellite operations, and management of smart lunar habitats. SafeMAP utilizes physics-based models of each agent and the relevant components, probability models of the environment and component operational states, and reward models for mission-specific objectives such as scientific task completion or resource consumption. The output of SafeMAP is a set of mission plans that satisfy the mission objective under specified risk/reward constraints. A readable interpretation of each of these generated mission plans is provided as an additional output. SafeMAP has been demonstrated on a simulated case study involving a four-rover system performing surface mapping operations and science tasks. The results of this paper demonstrate SafeMAP’s ability to generate explainable mission plans that satisfy the mission objective while minimizing risk under nominal and off-nominal conditions. 

Efficient and adaptive mission planning for Earth Observation Satellites (EOSs) remains a challenging task due to the growing complexity of user demands, task constraints, and limited satellite resources. Traditional heuristic and metaheuristic approaches often struggle with scalability and adaptability in dynamic environments. To overcome these limitations, we introduce AEM-D3QN, a novel intelligent task scheduling framework that integrates Graph Neural Networks (GNNs) with an Adaptive Exploration Mechanism-enabled Double Dueling Deep Q-Network (D3QN). This framework constructs a Directed Acyclic Graph (DAG) atlas to represent task dependencies and constraints, leveraging GNNs to extract spatial–temporal task features. These features are then encoded into a reinforcement learning model that dynamically optimizes scheduling policies under multiple resource constraints. The adaptive exploration mechanism improves learning efficiency by balancing exploration and exploitation based on task urgency and satellite status. Extensive experiments conducted under both periodic and emergency planning scenarios demonstrate that AEM-D3QN outperforms state-of-the-art algorithms in scheduling efficiency, response time, and task completion rate. The proposed framework offers a scalable and robust solution for real-time satellite mission planning in complex and dynamic operational environments.

We present a novel mission-planning strategy for heterogeneous multi-robot teams, taking into account the specific constraints and capabilities of each robot. Our approach employs hierarchical trees to systematically break down complex missions into manageable sub-tasks. We develop specialized APIs and tools, which are utilized by Large Language Models (LLMs) to efficiently construct these hierarchical trees. Once the hierarchical tree is generated, it is further decomposed to create optimized schedules for each robot, ensuring adherence to their individual constraints and capabilities. We demonstrate the effectiveness of our framework through detailed examples covering a wide range of missions, showcasing its flexibility and scalability.

Imaging satellite mission planning has received more and more attention as one of the core problems in the field of imaging satellite applications. In this paper, a hybrid discrete artificial bee colony (HDABC) algorithm is proposed to address this problem. The HDABC algorithm improves the three search phases of the basic artificial bee colony (ABC) algorithm to make them applicable to the discrete satellite mission planning problem. In the employed bee search phase, the population is divided and a multi-strategy search equation mechanism is used to balance the exploration and development of the algorithm. In the following bee search phase, two kinds of neighborhood search operators are designed based on the problem characteristics to further improve the fitness values of the better solutions. In the scout bee search phase, a migration operator and an immigration operator are introduced to improve the fitness values of the worse solutions and promote the exchange of different subpopulations to achieve co-evolution. In the experimental part, orthogonal experimental design is used to determine the appropriate algorithm parameters. Simulation experiments are carried out to test problems of different sizes. The experimental results show that the proposed HDABC algorithm shows good performance.

As the core technology in the field of imaging satellite application, imaging satellite mission planning has received more and more attention. Aiming at this problem, this paper proposes a distributed imaging satellite mission planning method (DISMPA) based on multi-agent system. Firstly, a distributed imaging satellite mission planning model with variable collaborative division of labor is constructed based on multi-agent system theory. The model defines the intelligence level of satellites in the satellite cluster and the interaction mode between satellites. Secondly, the cooperation mechanism between satellite agents is established based on blackboard model. Two negotiation strategies of worker serial activation and worker parallel activation are proposed. In order to improve the efficiency of negotiation, the targets are preassigned before negotiation. Finally, a hybrid discrete multi-verse optimization algorithm is proposed to solve the mission planning problem of worker agents in the model. Simulation experiments show that the average fitness values obtained by DISMPM using serial activation strategy and DISMPM using parallel activation strategy are up to 20% higher than those obtained by the centralized algorithm with the best solution effect in this paper, and the average negotiation times of the two DISMPM methods are both reduced by more than 80% compared with the contract network algorithm, effectively reducing the system traffic and reducing the communication burden. It is proved that DISMPA is suitable for distributed imaging satellite mission planning.

Hundreds of satellites equipped with cameras orbit the Earth to capture images from locations for various purposes. Since the field of view of the cameras is usually very narrow, the optics have to be adjusted and rotated between single shots of different locations. This is even further complicated by the fixed speed-determined by the satellite's altitude-such that the decision what locations to select for imaging becomes even more complex. Therefore, classical algorithms for this Satellite Mission Planning Problem (SMPP) have already been proposed decades ago. However, corresponding classical solutions have only seen evolutionary enhancements since then. Quantum computing and its promises, on the other hand, provide the potential for revolutionary improvement. Therefore, in this work, we propose a hybrid classical quantum computing approach to solve the SMPP combining the advantages of quantum hardware with decades of classical optimizer development. Using the Variational Quantum Eigensolver (VQE), Quantum Approximate Optimization Algorithm (QAOA), and its warm-start variant (W-QAOA), we demonstrate the applicability of solving the SMPP for up to 21 locations to choose from. This proof-of-concept-which is avail-able on GitHub (https://github.com/cda-tum/mqt-problemsolver) as part of the Munich Quantum Toolkit (MQT)-showcases the potential of quantum computing in this application domain and represents a first step toward competing with classical algorithms in the future.

Agile-satellite mission planning is a crucial issue in the construction of satellite constellations. The large scale of remote sensing missions and the high complexity of constraints in agile-satellite mission planning pose challenges in the search for an optimal solution. To tackle the issue, a dynamic destroy deep-reinforcement learning (D3RL) model is designed to facilitate subsequent optimization operations via adaptive destruction to the existing solutions. Specifically, we first perform a clustering and embedding operation to reconstruct tasks into a clustering graph, thereby improving data utilization. Secondly, the D3RL model is established based on graph attention networks (GATs) to enhance the search efficiency for optimal solutions. Moreover, we present two applications of the D3RL model for intensive scenes: the deep-reinforcement learning (DRL) method and the D3RL-based large-neighborhood search method (DRL-LNS). Experimental simulation results illustrate that the D3RL-based approaches outperform the competition in terms of solutions’ quality and computational efficiency, particularly in more challenging large-scale scenarios. DRL-LNS outperforms ALNS with an average scheduling rate improvement of approximately 11% in Area instances. In contrast, the DRL approach performs better in World scenarios, with an average scheduling rate that is around 8% higher than that of ALNS.

This paper presents a novel concept for a context-aware framework for mission planning and decentralized execution that integrates heterogeneous teammates such as autonomous drones, agents, and human responders into a unified, resilient system for dynamic operational scenarios. At the operational level, the system utilizes a Mission Planning function that receives high-level objectives, interprets mission intent, and decomposes complex objectives into a series of clear, actionable Mission-Essential tasks and sub-tasks. The process leverages contextual information from various sources such as sensor feeds, human reports and operational databases to dynamically assess resource availability and execution timing, while also considering operational constraints, including ethical constraints, and cost considerations. The system orchestrates a two-tier command structure, where the first level ensures that subordinates possess a robust understanding of mission objectives and the autonomy to adapt to changes in its operational environment. The second level comprises of diverse agents with varying levels of autonomy and capabilities, enabling iterative adaptation and collaboration. Experiments will be conducted in dynamic scenarios, such as deploying diverse drone platforms for Intelligence, Surveillance and Reconnaissance, and Search and Rescue missions to validate the framework’s feasibility. This research aims to develop context-awareness for system adaptation to its operational environment with a safety-monitoring function to enhance the safety, adaptability, and efficiency of coordinated drone-human teams in operational applications.

Aiming at modeling the giant heterogeneous sensing constellation and massive task planning problem, we proposed a modular planning algorithm framework that can be embedded with multiple algorithmic components, adopt multiple heuristic optimization algorithms combined with the strategy of predefined rules, and designed a lightweight and fast evaluation function, which can greatly improve the algorithm's performance, and validate the performance of the algorithm in solving the large-scale mission planning problem in a test dataset.

Artificial intelligence is transforming precision agriculture, offering farmers new tools to streamline their daily operations. While these technological advances promise increased efficiency, they often introduce additional complexity and steep learning curves that are particularly challenging for non-technical users who must balance tech adoption with existing workloads. In this paper, we present a natural language (NL) robotic mission planner that enables non-specialists to control heterogeneous robots through a common interface. By leveraging large language models (LLMs) and predefined primitives, our architecture seamlessly translates human language into intermediate descriptions that can be executed by different robotic platforms. With this system, users can formulate complex agricultural missions without writing any code. In the work presented in this paper, we extend our previous system tailored for wheeled robot mission planning through a new class of experiments involving robotic manipulation and computer vision tasks. Our results demonstrate that the architecture is both general enough to support a diverse set of robots and powerful enough to execute complex mission requests. This work represents a significant step toward making robotic automation in precision agriculture more accessible to non-technical users.

Remote sensing satellite mission planning is one of the hot issues in the space engineering research field, and a large number of mission planning approaches have been proposed in related research work. Numerous mission planning schemes were constructed for different mission requirements. How to evaluate the merits of the schemes is of great significance to improve the quality and effectiveness of remote sensing satellite missions. Based on the analysis of the basic problems of remote sensing satellite mission planning, a technology framework of mission scheme evaluation is proposed, and an evaluation index system for remote sensing mission planning schemes is constructed, including mission completion rate, planning timeliness and resources occupancy. A TOPSIS-based evaluation model is proposed to calculate the valuation of mission scheme according to the index system. The case study shows that the mission planning scheme evaluation approach proposed in this paper is feasible and effective.

Earth observation satellites, particularly agile Earth observation satellites (AEOSs) with enhanced attitude maneuverability, have become increasingly crucial in emergency response and disaster monitoring operations. Efficient mission planning for densely distributed ground targets with diverse priorities poses significant challenges, especially when considering strict attitude maneuver constraints and time-sensitive requirements. To address these challenges, this paper proposes a target clusters and dual-timeline optimization (TCDO) framework that integrates priority-based geographical clustering with temporal–spatial coordination mechanisms for efficient mission planning. The proposed approach effectively maintains satellite maneuver constraints while achieving significant improvements in priority-based target acquisition and computational efficiency. Experimental results demonstrate the framework’s superior performance, achieving a 94% coverage rate and a 99.5% reduction in computation time compared to traditional scheduling methods, such as linear programming and genetic algorithms.

To meet the demands of regional mapping with synthetic aperture radar (SAR) satellites, this study proposes a multistrip stitching imaging mission planning (MSIMP) method that considers energy consumption. Accounting for both regional coverage rate and satellite energy consumption, an MSIMP model was constructed. The model maximizes the region coverage benefit (RCB), minimizes the total imaging time (TIT) as the optimization objectives, and sets the indices of orbit selection, side-swing angle coefficients, and determination coefficients of the start time and end times of each orbit as the decision variables to realize the overall optimal scheduling of available orbital resources (ORs), side-swing angles, and imaging time of satellites. To address the problem of the large-scale and diverse types of decision variables in the MSIMP model, an improved particle swarm optimization (PSO) algorithm with a hybrid local search and differential evolution (LSDE) strategy (LSDE-PSO) that implements differential operations and local searches on two types of particles at various stages of the evolution process to enhance the model’s solution efficiency was proposed. The proposed method was validated using three simulation scenarios with different regional sizes. The experimental results showed that the proposed method demonstrated consistent adaptability to different-scale tasks, achieving the synchronized optimization of coverage revenue and on-orbit energy consumption. Compared with existing algorithms, the proposed LSDE-PSO algorithm can obtain the optimal imaging scheme with a higher RCB with lower TIT at almost the same computational cost, which provides significant technical support for mission planning for regional SAR satellite imaging in practical applications.

This paper addresses the mission planning and scheduling problem for satellite constellations by proposing an Adaptive Multi-Satellite Mission Allocation Algorithm (AMMAA). With the advancement of satellite imaging technology, satellite constellations are facing complex mission planning challenges involving multiple satellites, payloads, and constraints. This paper establishes a mathematical optimization model for mission planning with multiple constraints, including attitude maneuver time, power energy, and storage capacity, and designs a rule-based AMMAA. Through simulation experiments, the effectiveness of the algorithm in maximizing the target coverage rate and minimizing average revisit time is verified. The experimental results show that, in the simulation scenario, the algorithm achieves 100%target coverage, with high-priority targets being observed on average 15.7 times and an average revisit time of 19.15 minutes for all targets. The algorithm not only provides a feasible solution for multi-satellite mission planning problems but also offers a basis for quantitative analysis for the optimization of satellite constellation system schemes.

No abstract available

Aiming at the mission planning problem of the fully connected intelligent satellite system, this paper designs a hybrid architecture and a two-layer mission allocation method, which effectively improves the mission allocation efficiency and shortens the mission allocation time. The hybrid architecture combines the advantages of centralized and distributed architectures, and the two-layer mission allocation method improves the mission allocation efficiency through the collaborative work between and within the orbital plane. The experimental results show that the two-layer allocation takes the least time in the mission allocation process of the fully connected intelligent satellite system, regardless of the number of satellites or the size of the mission.

With the development of space technology, the application of agile earth observation satellites in complex observation missions has become increasingly important. This paper conducts research on collaborative mission planning for satellite constellations, focusing on the three-axis maneuvering characteristics of agile satellites. Firstly, considering various constraints faced by spacecraft during the observation of ground point targets, such as time constraints, attitude maneuvering constraints, and orbital dynamics constraints, the corresponding mission observation model and fitness function were established. The fitness function proposed in this paper can balance high observation benefits and low observation energy consumption, reflecting the observation requirements of practical engineering problems. Subsequently, based on integer coding, efficient planning for multi-point target missions was realized. To address the problem that the late-stage crossover process of traditional genetic algorithms causes significant disturbance to existing excellent solutions, a PMX (Partially Mapped Crossover) method based on an adaptive crossover operator was proposed. This crossover method introduces an adaptive crossover probability during the individual crossover optimization process, which improves the probability of the algorithm searching for the global optimal solution and accelerates the convergence speed of the algorithm. Simulation experiments show that the improved genetic algorithm outperforms traditional methods in both total observation benefits and planning efficiency, providing an effective solution for satellite mission planning.

No abstract available

The scheduling of Earth observation satellites presents a formidable multi-objective optimization challenge, characterized by inherent trade-offs among task completion rate, execution timeliness, and the temporal uniformity of revisits. To address this, we introduce the Multi-Satellite Observation Task Scheduling (MSOTS) framework, a novel end-to-end approach based on Multi-Agent Reinforcement Learning (MARL). This framework formulates the scheduling process as a Markov game, employing the Multi-Agent Proximal Policy Optimization (MAPPO) algorithm within a Centralized Training, Decentralized Execution (CTDE) paradigm to effectively navigate these competing objectives. Furthermore, to ensure a balanced evaluation, we propose a Composite Multi-Objective Performance Score grounded in a weighted harmonic mean. Comprehensive empirical evaluations conducted on large-scale, simulated orbital scenarios demonstrate that MSOTS significantly outperforms both traditional heuristics and existing deep reinforcement learning methods in comprehensive performance and robust efficiency. This research provides a highly effective and intelligent approach to modern satellite task scheduling.

With the rapid development of large-scale satellite constellations, efficient resource scheduling in dynamic and complex electromagnetic environments has emerged as a critical challenge. Traditional centralized optimization algorithms suffer from high computational complexity, poor adaptability to nonconvex mixed-integer constraints, and limited robustness in nonstationary environments. In this paper, we propose a novel multiagent reinforcement learning (MARL) framework based on Partially Observable Markov Decision Process (POMDP) modeling and a Modified Multi-Task Multi-Agent Twin Delayed Deep Deterministic Policy Gradient (Mod-MTMATD3) algorithm. Through a hierarchical state space design, joint action space optimization, and a composite reward mechanism, this framework achieves joint optimization of power allocation, task scheduling, and topology coordination in multi-satellite collaborative communication scenarios. Simulation results demonstrate that the proposed algorithm achieves over 40% improvement in convergence speed and a 63.6% increase in final reward value compared to traditional methods. Under extreme conditions where the inter-satellite distance is extended to 400 kilometers, it maintains a broadcast success rate of 96.2%, representing an 11.2% enhancement over benchmark algorithms. Additionally, the algorithm reduces energy consumption by 30%, significantly improving system energy efficiency. This study provides a scalable intelligent decision-making framework for large-scale satellite networks, offering crucial technical support for the development of space-air-ground integrated networks.

No abstract available

With the rise of sixth-generation (6 G) and the development of non-terrestrial networks (NTNs), low Earth orbit (LEO) satellites play a key role in global connectivity, especially in remote areas. However, due to the limited communication and computing resources of satellites, efficiently utilizing these capabilities has become an urgent problem. To address this, we propose a multi-satellite collaborative computation offloading framework for LEO satellite networks. We divide the satellites into communication satellites and computation satellites according to their resource status, and construct a cooperative processing model that supports cross-satellite task offloading and distributed execution. Meanwhile, we introduce the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm to achieve the access selection of ground terminals and the scheduling of computing tasks. Simulation results show that the method can effectively shorten the task completion time, improve the task success rate, and significantly enhance the resource utilization efficiency.

Assignment problems are a classic combinatorial optimization problem in which a group of agents must be assigned to a group of tasks such that maximum utility is achieved while satisfying assignment constraints. Given the utility of each agent completing each task, polynomial-time algorithms exist to solve a single assignment problem in its simplest form. However, in many modern-day applications such as satellite constellations, power grids, and mobile robot scheduling, assignment problems unfold over time, with the utility for a given assignment depending heavily on the state of the system. We apply multi-agent reinforcement learning to this problem, learning the value of assignments by bootstrapping from the known polynomial-time greedy solver and then learning from further experience. We then choose assignments using a distributed optimal assignment mechanism rather than by selecting them directly. We demonstrate that this algorithm is theoretically justified and avoids pitfalls experienced by other RL algorithms in this setting. Finally, we show that our algorithm significantly outperforms other methods in the literature, even while scaling to realistic scenarios with hundreds of agents and tasks.

The purpose of satellite tracking telemetry and control (TT&C) resource scheduling is to scientifically allocate limited ground TT&C resources, and to maximize the scheduled tasks of satellite TT&C through a certain scheduling method. Due to the sharp increase in satellite numbers, Traditional scheduling methods face challenges with more TT&C tasks, more resource competition and more emergencies. As the agent is independent, flexible, intelligent and distributed, we introduced agent technology to solve the problem. Combining characteristics of multi-satellite TT&C resource schedule with description method of attribute, interaction, rule and state on agent technology, we established four agent models: manager agent, station agent, equipment agent and task agent. Then, by analysing the requirements of satellite TT&C engineering, we proposed a comprehensive weighted priority determination method to determine TT&C resource allocation. Last, we simulated real application scenarios to validate this method. And it proves to be effective.

In extreme environments where ground communication infrastructure is often unavailable, UAV-assisted low Earth orbit (LEO) satellite edge computing networks provide a feasible solution for communication and computation services to terrestrial users. This paper focuses on the task offloading problem in such networks and proposes a unified “LEO-UAV-Ground” collaborative computing framework. By deploying edge servers on both LEO and UAVs, the framework enables efficient offloading and processing of tasks generated by ground user devices. To optimize task scheduling under constraints such as limited satellite resources and coverage duration, we design a multi-agent Double DQN-based reinforcement learning algorithm aimed at minimizing both task execution delay and energy consumption. The proposed algorithm leverages the policy-learning capability of deep reinforcement learning in complex scheduling scenarios to achieve coordinated allocation of heterogeneous resources. Experimental results on a real-world dataset demonstrate that our approach significantly outperforms traditional scheduling algorithms in terms of task completion delay and energy efficiency, validating its robustness and practicality in dynamic and resource-constrained environments.

No abstract available

No abstract available

With the rapid advancement of space technology, satellites are playing an increasingly vital role in fields such as Earth observation, communication and navigation, space exploration, and military applications. Efficiently deploying satellite missions under multi-objective, multi-constraint, and dynamic environments has become a critical challenge in the current aerospace domain. This paper integrates the concepts of game theory and proposes a distributed collaborative task model suitable for on-orbit satellite mission planning. A two-player impulsive maneuver game model is constructed using differential game theory. Based on the ideas of Nash equilibrium and distributed collaboration, multi-agent technology is applied to the distributed collaborative task planning, achieving collaborative allocation and countermeasure strategies for multi-objective and multi-satellite scenarios. Experimental results demonstrate that the method proposed in this paper exhibits good adaptability and robustness in multiple impulse scheduling, maneuver strategy iteration, and heterogeneous resource utilization, providing a feasible technical approach for mission planning and game confrontation in satellite clusters.

Space-air-ground integrated networks (SAGIN) combine the wide-area coverage of satellites with airborne platforms acting as relays, enabling efficient data delivery for large-scale Internet of Things (IoT) sensing applications. To further enhance transmission efficiency, this paper incorporates airborne onboard processing to reduce communication workloads and proposes an intelligent beam hopping strategy tailored to spatially uneven traffic demands. Specifically, we design a multi-agent deep reinforcement learning (MADRL)-based beam hopping framework, where satellite agents coordinate beam scheduling while considering spatial service heterogeneity and the diverse computational capacities of airborne platforms. To reduce the complexity introduced by individual task requirements, we integrate a summarized statistical profile of the computation tasks into the agent’s observation space, including average and maximum computation efficiencies across tasks. Simulation results demonstrate that the proposed scheme significantly accelerates convergence and reduces the data backlog by up to 80%, especially under scenarios with considerable heterogeneity in service demands and airborne platform computational capabilities.

The multi-agile satellite task scheduling problem (MSTSP) is a challenging scheduling optimization problem in the aerospace field, which directly influences the total task execution profit and load balance among satellites. The MSTSP has two issues to be handled sequentially, the initial task allocation and then the observation window allocation. Therefore, the quality of task allocation will heavily influence not only the results of observation window allocation but also the final profit and load balance of the multi-agile satellites system. However, existing researches mainly study the observation window allocation subproblem but fail to obtain satisfactory results in the initial task allocation, leading to poor performance when solving the entire MSTSP. To this end, this paper proposes a multi-population multi-objective genetic algorithm with an allocation knowledge-based crossover strategy (MPMOGA-AKCS). In order to optimize the two objectives of total profit and load balance more effectively, two strategies are proposed and integrated in MPMOGA-AKCS. First, a heuristic crossover strategy based on the multiple populations for multiple objectives (MPMO) framework is proposed to quickly generate high-quality offspring through incorporating the allocation knowledge from successful solutions for different obj ectives. Second, an adaptive archive update strategy is proposed to realize the dynamic balance between the diversity and convergence of the solution set through phased adjustments. Extensive experimental results show that the performance of the MPMOGA-AKCS is superior to some state-of-the-art algorithms.

For the task scheduling problem in multi-agent systems, this paper proposes a collaborative optimization method based on Graph Neural Network and Reinforcement Learning. Firstly, a heterogeneous graph structure is constructed to uniformly model the temporal dependencies, resource competition, and agent capability differences among tasks, and multi-dimensional node features are designed to fully describe the scheduling state. Secondly, the Proximal Policy Optimization algorithm is adopted to achieve efficient training and stable convergence of the policy network based on graph embedding, supporting rapid decision-making for large-scale instances. To verify its effectiveness, 30 test cases are generated for each of the three scales, totaling 90 cases. It is compared with Genetic Algorithm and Gurobi's exact solver. Through a large number of simulation experiments, the effectiveness and advantages of this method in solving the studied problem have been verified.

The Space-Air-Ground Integrated Network (SAGIN) framework is a crucial foundation for future networks, where satellites and aerial nodes assist in computational task offloading. The low-altitude economy, leveraging the flexibility and multifunctionality of Unmanned Aerial Vehicles (UAVs) in SAGIN, holds significant potential for development in areas such as communication and sensing. However, effective coordination is needed to streamline information exchange and enable efficient system resource allocation. In this paper, we propose a Clustering-based Multi-agent Deep Deterministic Policy Gradient (CMADDPG) algorithm to address the multi-UAV cooperative task scheduling challenges in SAGIN. The CMADDPG algorithm leverages dynamic UAV clustering to partition UAVs into clusters, each managed by a Cluster Head (CH) UAV, facilitating a distributed-centralized control approach. Within each cluster, UAVs delegate offloading decisions to the CH UAV, reducing intra-cluster communication costs and decision conflicts, thereby enhancing task scheduling efficiency. Additionally, by employing a multi-agent reinforcement learning framework, the algorithm leverages the extensive coverage of satellites to achieve centralized training and distributed execution of multi-agent tasks, while maximizing overall system profit through optimized task offloading decision-making. Simulation results reveal that the CMADDPG algorithm effectively optimizes resource allocation, minimizes queue delays, maintains balanced load distribution, and surpasses existing methods by achieving at least a 25% improvement in system profit, showcasing its robustness and adaptability across diverse scenarios.

Non-geostationary orbit (NGSO) constellations enabled by beam hopping (BH) technology are characterized by wide coverage and high spectrum efficiency. However, how to efficiently schedule multi-satellite beam resources to satisfy the heterogeneous and uneven terrestrial traffic demands remains a huge challenge for satellite operators. This paper proposes a satellite-terrestrial coordinated multi-satellite BH scheduling framework, where the complex multi-satellite BH problem is formulated into a long-term and a short-term subproblems. The long-term subproblem is cell-satellite association problem, which is solved by a low-complexity iterative algorithm executed in network operation control center (NOCC) to minimize the traffic load gap among satellites while considering the interference avoidance. The short-term subproblem is multi-satellite traffic-driven BH problem and we propose a multi-agent deep reinforcement learning (MADRL) architecture where each satellite can cooperatively make real-time BH decisions using the well-trained model by QMIX algorithm to adapt to time-varying and heterogeneous traffic. Simulation results demonstrate that the traffic load gap and network delay have been reduced by 70% and 50% respectively compared with non-load-balancing scheme. Besides, the proposed algorithm outperforms other benchmarks in terms of the network throughput under various traffic load cases and the average network delay is kept within 4 ms. Furthermore, the proposed QMIX-BH can be applied to real-time scheduling since the execution time is less than 1 ms.

With the advancement of low earth orbit (LEO) satellite communications, direct-to-satellite technology has attracted widespread attention. Due to the constrained onboard resources of LEO satellite, an efficient resource allocation strategy is essential to meet diverse user traffic demands. Furthermore, the spectrum sharing among multi-layer LEO satellites requires effective interference avoidance mechanism to enhance the overall service quality of the satellite system. To achieve efficient resource management in a dual-layer LEO satellite system, we propose a resource allocation strategy based on multi-agent proximal policy optimization (MAPPO). This strategy consists of three steps. Firstly, beam hopping time plan (BHTP) is designed based on a low complexity strategy. Next, the users served in each time slot are selected according to the buffer status. Finally, the MAPPO framework is employed to enable the flexible allocation of bandwidth and power resources. Simulation results demonstrate that the proposed strategy not only satisfies the diverse traffic demands of terminal users but also effectively reduces satellite power consumption.

Laser inter-satellite links (LISLs) have greatly extended communication distance between satellites, allowing for establishment of dynamic links to reduce communication delay. However, a closed-loop control is required for LISL, which causes high energy consumption. Proper scheduling of dynamic LISLs can effectively reduce energy consumption and communication delay. In this study, a satellite link mode with three fixed LISLs and one dynamic LISL is designed, and its feasibility is analyzed. The optimization problem is formulated and transformed into a Markov decision process (MDP) by modeling it as a sequential decision problem. By decomposing states, actions, and reward functions, the MDP is divided into the proposed multi-agent deep reinforcement learning (MADRL). Moreover, compressed sensing is utilized to cut down state information to reduce communication, storage, and computation overhead. Furthermore, network parameters and experience sharing, and prioritized experience replay have been adopted to improve stability and convergence speed of network training with a large number of agents. Experimental results show that under different routing strategies, the proposed MADRL can reduce energy consumption by over 15% and delay by approximately two hops compared to fixed LISLs scenario within several iterations.

In the face of rapidly evolving communication technologies and increasing user demands, traditional terrestrial networks are challenged by the need for high-quality, high-speed, and reliable communication. This paper explores the integration of heterogeneous satellite networks (HSN) with emerging technologies such as Mobile Edge Computing (MEC), in-network caching, and Software-Defined Networking (SDN) to enhance service efficiency. By leveraging dual-layer satellite networks combining Low Earth Orbit (LEO) and Geostationary Earth Orbit (GEO) satellites, the study addresses resource allocation and interference coordination challenges. This paper proposes a novel resource allocation and interference coordination strategy for dual-layer satellite networks integrating LEO and GEO satellites. We formulate a mathematical optimization problem to optimize resource allocation while minimizing co-channel interference and develop an ADMM-based distributed algorithm for efficient problem-solving. The proposed scheme enhances service efficiency by incorporating MEC, in-network caching, and SDN technologies into the satellite network. Simulation results demonstrate that our proposed algorithm significantly improves network performance by effectively managing resources and reducing interference.

When users with different preferences and entitlements compete for limited access to a shared resource, maximizing total utility can lead to significant disparities in how users are served. We examine this tension in the context of allocating satellite observation windows, where users differ in their willingness to pay or their contribution to the system. The goal is to schedule observations efficiently while promoting balanced, fair access among users. This challenge is amplified in settings where coordination is decentralized and users negotiate outcomes without a central authority. We propose a hybrid algorithm designed to balance fairness and efficiency in distributed scheduling. Our method produces allocations that retain high efficiency while reducing inequality. Although developed for satellite scheduling, the algorithm applies more broadly to decentralized systems where users with heterogeneous preferences share limited resources.

Data distribution across different facilities offers benefits such as enhanced resource utilization, increased resilience through replication, and improved performance by processing data near its source. However, managing such data is challenging due to heterogeneous access protocols, disparate authentication models, and the lack of a unified coordination framework. This paper presents DynoStore, a system that manages data across heterogeneous storage systems. At the core of DynoStore are data containers, an abstraction that provides standardized interfaces for seamless data management, irrespective of the underlying storage systems. Multiple data container connections create a cohesive wide-area storage network, ensuring resilience using erasure coding policies. Furthermore, a load-balancing algorithm ensures equitable and efficient utilization of storage resources. We evaluate DynoStore using benchmarks and realworld case studies, including the management of medical and satellite data across geographically distributed environments. Our results demonstrate a 10 % performance improvement compared to centralized cloud-hosted systems while maintaining competitive performance with state-of-the-art solutions such as Redis and IPFS. DynoStore also exhibits superior fault tolerance, withstanding more failures than traditional systems.

No abstract available

Future wireless networks will be artificial intelligence (AI) native and highly heterogeneous, where 5G/6G, next generation Wi-Fi, low-altitude drone networks, and satellite networks, have to coexist in the crowded gigahertz and tens of gigahertz frequency bands. This introduces grave challenges in radio resource sharing due to fundamental technical specification discrepancies and complex time-varying interference patterns. As a remedy, we propose a learn-to-share (L2S) framework—a distributed multi-agent paradigm for spectrum sharing among multiple radio access technologies (RATs) that will coexist in the 6G era. In this framework, user terminals (UTs) that are mutually interfering with each other, regardless of which RAT they use, can operate as intelligent agents capable of autonomously making decisions on spectrum access through local sensing and learning, without centralized coordination. We formulate this spectrum sharing problem as a decentralized partially observable Markov decision process (Dec-POMDP) and design a spectral-temporal attention and recurrence (STAR) algorithm as the cognitive engine. STAR decomposes individual agent decisions into three tightly coupled subtasks: channel selection via Transformer networks that capture spatial interference correlations, threshold adaptation through bidirectional long short-term memory (LSTM) networks that model temporal channel dynamics, and waiting time scheduling using predictive LSTM planners that intelligently avoid collisions. Through extensive OMNeT++-based simulations, we demonstrate that STAR significantly outperforms traditional approaches: achieving 73.5% improvement in access success rate, 30.6% enhancement in fairness (Jain’s index), and 88.1% increase in system throughput compared with the conventional listen-before-talk (LBT) mechanism, while requiring 50.6% fewer convergence episodes than the deep recurrent Q-network (DRQN) baseline. Moreover, compared with the mixing Q-network (QMIX), a representative centralized training with decentralized execution (CTDE) baseline, our algorithm reduces the computing power needed per UT by 21.9% while achieving comparable access success rate and system throughput as a fully decentralized multi-agent framework.

Distributed space systems are increasingly valued in the space industry, as they enhance mission performance through collaborative efforts and resource sharing among multiple heterogeneous satellites. Additionally, enabling autonomous and real-time satellite-to-satellite communications through Inter-Satellite Links (ISLs) can further increase the overall performance by allowing cooperation without relying on ground links and extensive coordination efforts among diverse stakeholders. Given the constrained resources available onboard satellites, a crucial element of achieving cost-effective and autonomous cooperation involves minimizing energy wastage resulting from unsuccessful or unnecessary communication. To address this challenge, satellites must anticipate their ISL opportunities or encounters with minimal resource utilization. Building upon prior publications, this work presents further insights into the use of supervised learning to enable satellites to forecast their encounters without relying on orbit propagation. In particular, a more realistic definition of satellite encounters, along with a versatile solution applicable to all polar low-Earth orbit satellites is implemented. Results show that the trained model can anticipate encounters for realistic and unseen data from an available data source with a balance accuracy of around 90% and six times faster when compared with the well-known Simplified General Perturbation 4 orbital model.

Due to the uneven spatial distribution of ground communication demands and the multi-coverage characteristics of Low Earth Orbit satellites, efficient coordination of multiple satellite beams is essential to meet heterogeneous ground requirements. Traditional methods typically rely on centralized beam planning at ground control centers, which often leads to scheduling delays and increased communication overhead, thereby limiting system responsiveness and flexibility. This paper proposes a Neighboring Satellite Beam Cooperation algorithm based on Multi-Agent Deep Reinforcement Learning (NSBC-MDRL), enabling autonomous inter-satellite coordination without Ground Control Center. We model the beam cooperation problem under a full-frequency reuse scenario, aiming to maximize system throughput. To improve scheduling timeliness and avoid additional communication overhead, NSBC-MDRL adopts an architecture that combines centralized offline training with distributed onboard inference. During the inference phase, each agent independently makes decisions based on local observations and inferred behaviors of other agents, enabling implicit coordination without inter-satellite communication. Simulation results demonstrate that the proposed NSBC-MDRL algorithm consistently outperforms existing benchmark methods across all evaluated scenarios.

No abstract available

Task offloading in satellite networks, which involves distributing computational tasks among heterogeneous satellite nodes, is crucial for optimizing resource utilization and minimizing system latency. However, existing approaches such as static offloading strategies and heuristic-based offloading methods neglect dynamic topologies and uncertain conditions that hinder adaptability to sudden changes. Furthermore, current collaborative computing strategies inadequately address satellite platform heterogeneity and often overlook resource fluctuations, resulting in inefficient resource sharing and inflexible task scheduling. To address these issues, we propose a dynamic gradient descent-based task offloading method. This method proposes a collaborative optimization framework based on dynamic programming. By constructing delay optimization and resource efficiency models and integrating dynamic programming with value iteration techniques, the framework achieves real-time updates of system states and decision variables. Then, a distributed gradient descent algorithm combined with Gradient Surgery techniques is employed to optimize task offloading decisions and resource allocation schemes, ensuring a precise balance between delay minimization and resource utilization maximization in dynamic network environments. Experimental results demonstrate that the proposed method enhances the global optimizing result by at least 1.97%, enhances resource utilization rates by at least 3.91%, and also reduces the solution time by at least 191.91% in large-scale networks.

Satellite networks play important roles in fields, such as communication, meteorology, and the Internet of Things. However, in highly dynamic multilayer heterogeneous satellite networks, the diverse satellite transmission demands pose challenges for network management. In large-scale satellite networks, conventional centralized network management methods have various adverse effects, such as high latency, high communication overhead, and high computational complexity. In this article, to overcome these challenges, we propose a distributed scheme for allocating intersatellite links between satellites at different orbital heights considering their different transmission demands. Specifically, we model the distributed link allocation framework as a Stackelberg game. Low-Earth orbit (LEO) satellites, as leaders, apply for access to medium-Earth orbit (MEO) satellites using a proposed link selection algorithm based on a stochastic best response strategy. The MEO satellites, as followers, allocate link resources using a proposed heuristic-based time slot resource allocation algorithm in accordance with the accessing applications. Simulation results show that the proposed algorithm outperforms benchmark algorithms in terms of the degree of matching between the transmission capacity and transmission requirements.

This article investigates the distributed optimal coordination problem and distributed constrained optimal coordination problem for multiple heterogeneous linear systems over a directed communication network. The objective is to develop distributed optimization algorithms for a set of heterogeneous linear agents over a directed communication network, which will drive the decision variables of the agents to the optimal state with optimal system performance. First, a new distributed optimal coordination algorithm is developed, comprising an output regulation tracking controller and a surplus-based optimal policy generator. This is achieved by utilizing the output regulation technique and surplus-based consensus method. Second, a projected distributed optimal coordination algorithm is designed to address the distributed constrained optimal coordination problem, where the decision variables of the agents are constrained by local feasible set constraints and network resource constraints. In comparison to previous distributed optimal coordination studies, the agent dynamics and communication framework have been significantly expanded, and this represents the first investigation of the distributed constrained optimal coordination problem for heterogeneous linear systems over weight-unbalanced directed communication networks. Finally, the practical applications of multiple rotary-wing air vehicle systems and battery energy storage systems are presented, thereby demonstrating the efficacy of the proposed distributed optimal coordination and distributed constrained optimal coordination algorithms in a real-world context.

To achieve global coverage and ubiquitous connectivity, the non-terrestrial network (NTN) has been regarded as a key enabler in the sixth generation (6G) network, which includes uncrewed aerial vehicles (UAVs), high-altitude platforms (HAPs), and satellites. Since the unique characteristics of various NTN platforms strongly affect their implementation and lead to a highly dynamic and heterogeneous NTN scenario, achieving distributed coordination remains an important research direction. However, the explicit and systematic analysis of the individual layers'challenges and corresponding distributed coordination solutions in heterogeneous NTNs has not been proposed yet. Therefore, in this paper, we summarize the unique characteristics of each NTN platform, identify communication challenges within individual layers, and propose potential delay-tolerant or delay-sensitive coordinated solutions accordingly. We further analyse the feasibility of leveraging multi-agent deep reinforcement learning (MADRL) algorithms to achieve the proposed coordinated solutions. Finally, we present a case study of the joint scheduling and trajectory optimization problem in heterogeneous NTN, where a two-timescale multi-agent deep deterministic policy gradient (TTS-MADDPG) algorithm is developed to validate the effectiveness of distributed coordination.

In this article, the robust distributed optimal coordination (DOC) problem over unbalanced directed communication networks is investigated for heterogeneous multiagent systems subjected to external disturbances. First, by introducing an external disturbance observer, a balanced compensation variable and using the output regulation technique, a new robust continuous-time DOC algorithm is developed, ensuring that the sum of the local convex objective functions is optimal and the behaviors of all agents are coordinated. Then, a robust event-triggered DOC algorithm is designed, which consists of an event-triggered closed-loop controller and an event-triggered balanced compensator. Compared with the existing distributed optimization algorithms, the algorithms of this article solve the DOC problem over unbalanced directed communication network. The heterogeneous agent dynamics with external disturbances considered in this article are in a general form. Further, the introduction of the event-triggering mechanism can avoid continuous information transmission and improve the efficiency of system resources utilization. Finally, numerical simulations are shown to demonstrate the effectiveness of the theoretical results.

This study explores the distributed optimal coordination problem for Linear Time-Invariant Multi-Agent Systems (MAS). Each agent has its cost function. The goal is to control the agents to minimise the combined cost function, which sums their individual costs. Unlike previous studies that rely on known agent dynamics for control design, our research assumes the dynamics are almost entirely unknown. Additionally, the communication topology is assumed to change over time in a switching manner. We propose a novel distributed control framework with two layers to tackle this challenge. The top layer searches for the minimiser and provides a reference signal to the bottom layer. The bottom layer uses adaptive controllers to enable agents to track the reference signal and approach the minimiser. It is not enough for the reference trajectories to converge to the minimiser; they must also be feasible for the agents to follow. The unknown dynamics and switching communication graph make this task more complex. Our framework addresses these challenges by applying Model Reference Adaptive Control (MRAC) principles. We validate the theoretical results through numerical simulations and provide practical examples to show the approach’s applicability in real-world scenarios.

With the rapid development of low Earth orbit (LEO) satellite networks, the issues of spectrum resource scarcity and intersatellite interference have become increasingly prominent. Traditional fixed and centralized multibeam satellite systems suffer from a lack of flexibility and single-point failure vulnerability, making them inadequate for handling dynamically changing traffic demands. Beam hopping (BH) technology enables flexible resource allocation; however, due to the uneven distribution of ground traffic demand and its temporal variations, efficiently allocating beams remains a critical challenge. In this article, we propose a distributed load balancing and interference avoidance BH (DLBIA-BH) protocol within a distributed heterogeneous satellite network (DHSN). The core concept of DLBIA-BH is to first implement a load balancing mechanism that distributes network traffic evenly among LEO satellites. Then, each LEO satellite employs a real-time (RT) traffic-supporting BH mechanism. Finally, medium Earth orbit satellites coordinate the BH resource scheduling among LEO satellites to mitigate intersatellite interference and enhance overall network performance. Compared with benchmark methods, DLBIA-BH maximizes the overall network throughput and RT traffic throughput in DHSN while also improving the total traffic and RT traffic satisfaction rate across the network.

No abstract available

Given the crucial role of the earth observation satellites in numerous key applications, using Low Earth Orbit (LEO) satellite internet as intermediaries via Laser Inter-Satellite Links (LISLs) has emerged as a promising solution to help transmit substantial amounts of observation data to ground stations. For Very Low Earth Orbit (VLEO) observation satellites, optimizing the cross-layer topology between themselves and LEO communication satellites has become paramount. To mitigate existing centralized algorithms’ reliance on global data transfer requirement information, a Novel Distributed Interactive Mechanism (NDIM) for cross-layer LISL establishment is proposed. Here, the VLEO observation satellite decides its own access strategy based on local network information gleaned from three information exchanges with the LEO communication satellite. Within this mechanism, the cross-layer link optimization is performed via the formulation of a multi-objective topology optimization model, which considers the transmission requirements of observation satellites, load balancing amongst the communication satellite layer, and the transmission delay of emergency tasks. Based on this framework, we propose a Distributed Multi-objective cross-layer Topology Optimization (DMTO) algorithm. Our algorithm is novel in considering the remaining load of the intermediary communication satellites, and it allows observation satellites to decide on access plans on demand, given incoming data. Additionally, we used real data from Typhoon LEKIMA to establish a multi-task scenario and conducted packet-level simulations based on the Starlink and Dove constellations. The results indicate that, compared to the existing baseline, the DMTO algorithm increased the observation data throughput by 1.37% (326.95 GB) and reduced the average transmission delay of emergency task data by 4.20% (20.5 seconds).

This article investigates the continuous-time optimal distributed coordination problem with resource allocation constraints for general linear multiagent systems. The study is conducted over a connected undirected graph. By integrating the tracking controller design with global resource allocation optimization, fully distributed state-feedback controllers are proposed to solve optimization problems with output-based local objective functions. The dynamics of the entire multiagent system are well studied at the equilibrium point, which solves the optimal resource allocation and keeps system stable at the same time. The characteristics of the stable states are extracted as additional optimization constraints. By transferring the output-based optimization problem of general linear dynamics systems to a combination of a state-based optimization problem of single-integrator dynamics systems and a state tracking problem, the state equation of the system dynamics can be simplified through appropriate transformation, thereby decreasing the difficulty that brings to the performance index optimization, eliminating assumptions about the structure of the system state matrices, and achieving output stability and performance optimality simultaneously. The team performance, formed by a sum of privately known convex local objective functions and a demand for the total resource, is optimized in a fully distributed fashion. Sufficient conditions are given to ensure that the multiagent system with the proposed algorithms can reach the optimal resource allocation. Numerical simulations are provided to verify the feasibility of the controllers.

This note considers the distributed optimal coordination (DOC) problem for heterogeneous linear multiagent systems. The local gradients are locally Lipschitz and the local convexity constants are unknown. A control law is proposed to drive the states of all agents to the optimal coordination that minimizes a global objective function. By exploring certain features of the invariant projection of the Laplacian matrix, the global asymptotic convergence is guaranteed utilizing only local interaction. The proposed control law is then extended with event-triggered communication schemes, which removes the requirement for continuous communications. Under the event-triggered control law, it is proved that no Zeno behavior is exhibited and the global asymptotic convergence is preserved. The proposed control laws are fully distributed, in the sense that the control design only uses the information in the connected neighborhood. Furthermore, to achieve the DOC for linear multiagent systems with unmeasurable states, an observer-based event-triggered control law is proposed. A simulation example is given to validate the proposed control laws.

The evolution of software-defined networking (SDN) has transitioned from centralized control plane architectures to distributed control plane models, enhancing scalability, reliability, and fault tolerance. While central controllers often face challenges in managing extensive networks, distributed control architectures provide superior performance and flexibility. Initially, homogeneous distributed architectures were prevalent; however, the demand for heterogeneous distributed control planes where controllers interact with diverse systems and components has increased, necessitating advanced coordination and interoperability. This paper compares two widely adopted open-source distributed controllers, the Open Network Operating System (ONOS) and OpenDaylight (ODL), focusing on their methodologies for distributed clustering. It further explores the roles of Atomix and Akka (now Apache Pekko) distributed frameworks in ONOS and ODL respectively, discussing their impact on performance and scalability. Furthermore, the paper examines the data consistency models used by ONOS and ODL to manage network state and synchronization. Finally, it offers recommendation for enhancing the interoperability and performance of heterogeneous distributed control planes in large-scale and diverse network deployments.

The consensus problem under a single leader forms the foundation of coordination control in multi-agent systems (MASs). Nonetheless, the issue of containment control, which involves multiple leaders, represents a critical aspect that demands further investigation. This paper addresses the containment control of nonlinear heterogeneous partial difference multi-agent systems (HPDMASs) characterized by state time delays and governed by hyperbolic or parabolic partial difference equations. This study proposes a D-type distributed iterative containment control protocol incorporating initial state learning, which leverages local state information from neighboring agents. Using the contraction mapping principle, this research examines the convergence conditions for the containment error within a finite time interval. The results demonstrate that, under the proposed containment control protocol, the containment error in the HPDMASs can gradually converge to zero as the number of iterations increases. Simulation results are provided to validate the effectiveness of the proposed control protocol.

This paper deals with the distributed estimation and the attitude coordination problem for a network of heterogeneous satellites with flexible appendages. For the estimation task, the agents employ a distributed observer in a minimal-order formulation to obtain the angular rates of a non-collaborative target, starting from partial measurements of its angular positions. It is proven that, for an undirected topology, the estimation error asymptotically converges to zero if the consensus gain is greater than a certain threshold. The main feature of such an observer is the reduction of its order compared to other approaches in the literature when the observed system can be expressed in an integral state-space form. For the control task, a consensus protocol is shown to achieve attitude coordination among the satellites even though they share only their partial target information. Finally, the separation principle assures that the coupled implementation of the estimation and control procedures gives rise to a stable system. Numerical examples illustrate the results for a mixed network of sensing and non-sensing agents.

The rapidly increasing proliferation of independently operated distributed energy resources (DERs) has imposed significant management costs on power systems, necessitating the active coordination of widely dispersed and small-scale DER clusters by intermediate aggregators, such as virtual power plants (VPPs). Under the context of two-way free choice between VPPs and DERs, VPPs need to fairly and reasonably quantify the aggregation value of heterogeneous DERs and establish differentiated dynamic incentive mechanisms to enhance the willingness of DERs to participate in unified management. Therefore, this study focuses on multiple grid-supervised VPPs (GVPPs) of techno-market integrated type as the research objects. This paper designs an evaluation method based on a multi-factor normalized improved Shapley value to quantify the aggregation value of DERs, and, on this basis, constructs an internal dynamic differentiated pricing mechanism for GVPPs. Furthermore, a nonlinear satisfaction evaluation function based on opportunity cost is developed, which is combined with actual revenues to quantify the participation value of GVPPs. In addition, this paper develops a multi-participant dynamic hybrid game model using a scenario-based two-stage stochastic optimization, thereby constructing a dual dynamic model of GVPP dynamic pricing and DER dynamic aggregation. Ultimately, the proposed model's effectiveness is validated through case studies, sensitivity analyses, and scalability analysis, demonstrating the successful integration of all DER categories within the GVPPs' collective management.

This paper presents an optimisation model for the coordinated operation of a virtual power plant (VPP) that integrates a range of VPP-owned distributed energy resources (DERs), including wind turbines, photovoltaic systems, diesel generators, battery energy storage systems (BESSs), flexible loads, and behind-the-meter prosumers. A mixed-integer linear programming (MILP) framework is developed to determine optimal dispatch schedules and grid interactions under time-of-use (ToU) pricing and feed-in tariffs (FiT), with the objective of maximising VPP profit. The model is evaluated across three operational scenarios: (1) No VPP coordination, (2) VPP without BESS optimisation, and (3) VPP with BESS optimisation. The optimisation is implemented in Python using the Gurobi Optimiser's branch-and-cut algorithm. Simulation results demonstrate a substantial $\text{5 6 \%}$ increase in daily profit under the optimised VPP with BESS scenario, highlighting the economic advantages of coordinated DER utilisation and strategic grid exports. The proposed approach provides a practical decision-support tool for VPP operators, with potential for future enhancement to incorporate uncertainty in renewable energy generation and associated market conditions.

No abstract available

The coordination of distributed energy resources (DERs) within virtual power plants (VPPs) is expected to generate significant economic benefits and enhance the operational stability of modern power systems. However, achieving massive coordination of heterogeneous and uncertain DERs remains a challenge in current research. To address this issue, this article proposes a novel bi-level optimization approach based on mean-field reinforcement learning (MFRL) to enable the coordination of massive DERs in VPPs. The problem is decomposed into multiple subproblems: the upper-level subproblem models power dispatch among integrated energy systems (IESs) in response to coordinated demand, while a series of lower-level subproblems determine the operational schemes of DERs within individual IESs. Considering the large decision space, an MFRL algorithm with fast Shapley credit allocation is developed to efficiently solve the upper-level optimization. Meanwhile, the lower-level subproblems are formulated as small-scale mixed-integer linear programming (MILP) problems, addressing the difficulties caused by IES heterogeneity in applying mean-field approximation. Simulation results show that the proposed approach significantly improves convergence speed and reduces the global cost of VPP operation, especially in massive-scale scenarios. In test scenarios ranging from 10 to 500 agents, the proposed bi-level optimization approach improves the objective by 4.8%–26.6%, compared to the advanced baseline method.

The emerging architecture in next-generation mobile networks leverages the coexistence of low Earth orbit (LEO) and geostationary Earth orbit (GEO) satellites within heterogeneous networks. This setup not only enables seamless coverage but also enhances user data rates. However, due to the scarcity of resources in such heterogeneous satellite coexistence networks, efficiently allocating onboard resources, particularly spectrum resources, poses a significant challenge. In the context of satellite communication systems under dynamic heterogeneous multi-system coexistence environments, we design a cross-system joint optimization framework and propose an online resource allocation scheme based on beam-hopping (BH) to coordinate inter-system transmissions. This work further considers the uncertainty of beam service durations and formulates a resource optimization problem. We decompose this problem into two sub-problems: inter-cell resource allocation and intra-cell resource allocation. For inter-cell resource allocation, we adopt the Deep Deterministic Policy Gradient (DDPG) algorithm to provide a solution. Intra-cell resource allocation is addressed using traditional convex optimization methods. Simulation results validate that the proposed algorithm ensures fairness in LEO satellite data processing while effectively improving overall system performance.

In order to ensure efficient data transmission from satellites to ground stations (GSs), resource allocation schemes must be designed. In cases where direct data transmission from satellites to GSs is not possible due to a dynamic network topology and limited contact time, efficient relay selection or route selection schemes should be employed. This paper considers a satellite communication network in which a number of source low earth orbit(SLEO) satellites are attempting to transmit their data flows to the designated GSs. To improve the transmission performance of the data flows, one geosynchronous earth orbit (GEO) satellite and a number of relay LEO (RLEO) satellites in the network are used as relays. To maximize energy efficiency, a joint route selection and resource allocation mechanism is proposed. The energy cost of the system, which is the sum of the energy cost of the SLEO satellites and the RLEO satellites, is used to formulate the joint route selection and resource allocation as a system energy cost minimization problem. Since the original optimization problem is NP hard, it is transformed into three subproblems: inter-satellite power allocation, satellite-ground power allocation, and route selection. These subproblems are solved using a greedy algorithm, the Lagrange dual method, and a matching-based heuristic algorithm, respectively. The numerical results demonstrate the effectiveness of the proposed scheme.

With the rapid development of heterogeneous satellite networks integrating geostationary earth orbit (GEO) and low earth orbit (LEO) satellite systems, along with the significant growth in the number of satellite users, it is essential to consider frequency compatibility and coexistence between GEO and LEO systems, as well as to design effective system resource allocation strategies to achieve efficient utilization of system resources. However, existing beam-hopping (BH) resource allocation algorithms in LEO systems primarily focus on beam scheduling within a single time slot, lacking unified beam management across the entire BH cycle, resulting in low beam-resource utilization. Moreover, existing algorithms often employ iterative optimization across multiple resource dimensions, leading to high computational complexity and imposing stringent requirements on satellite on-board processing capabilities. In this paper, we propose a BH-based beam scheduling and resource allocation framework. The proposed framework first employs geographic isolation to protect the GEO system from the interference of the LEO system and subsequently optimizes beam partitioning over the entire BH cycle, time-slot beam scheduling, and frequency and power resource allocation for users within the LEO system. The proposed scheme achieves frequency coexistence between the GEO and LEO satellite systems and performs joint optimization of system resources across four dimensions—time, space, frequency, and power—with reduced complexity and a progressive optimization framework. Simulation results demonstrate that the proposed framework achieves effective suppression of both intra-system and inter-system interference via geographic isolation, while enabling globally efficient and dynamic beam scheduling across the entire BH cycle. Furthermore, by integrating the user-level frequency and power allocation algorithm, the scheme significantly enhances the total system throughput. The proposed progressive optimization framework offers a promising direction for achieving globally optimal and computationally tractable resource management in future satellite networks.

With the rapid development of low Earth orbit (LEO) satellite networks, the deployment of edge computing nodes on LEO satellites has become a crucial technology for latency optimization. However, most existing task offloading strategies rely on binary decision mechanisms, failing to fully leverage crosslayer collaborative computing architectures or effectively utilize inter-satellite link technology for load balancing. We formulate a joint optimization framework for adaptive task partitioning and heterogeneous resource allocation in satellite-terrestrial integrated networks. Leveraging inter-satellite link technology, our approach simultaneously minimizes end-to-end task latency while maintaining computational load balancing across satellite nodes. We further develop GDPG (Gated Recurrent Unit Enhanced Deep Deterministic Policy Gradient), through the integration of sequence modeling and reinforcement learning, GDPG enhances decision-making accuracy. Experimental results show that the proposed algorithm outperforms benchmark methods in reducing task latency, verifying its adaptability and effectiveness in dynamic satellite edge network environments.

The emerging architecture in the next generation of mobile networks leverages the coexistence of Low Earth Orbit (LEO) and Geostationary Orbit (GEO) satellites in a heterogeneous network. This setup not only offers seamless coverage but also enhances user rates. Nevertheless, the efficient allocation of onboard resources, particularly spectrum resources, poses a significant challenge due to their scarcity in such heterogeneous satellite coexistence networks. A practical solution is found in the use of beam hopping (BH) technology. This technology enables multi-beam satellites to serve users using fewer beams than traditional spot-beam systems. This paper proposes a resource allocation strategy for the heterogeneous LEO-GEO coexistence satellite network. We formulate this resource allocation strategy as a joint optimization problem. Due to the complexity of the system arising from the coupling of multiple variables, we break down the original problem into two manageable sub-problems. The first addresses user association, subcarrier, and power allocation and employs a standard convex optimization algorithm for a solution. The second tackles the illuminated beam selection issue, with a genetic algorithm (GA) providing a solution. The effectiveness of our proposed scheme is established through simulation experiments, demonstrating clear performance gains.

With the continuous development of satellite technology, satellite edge computing is garnering unprecedented attention as a crucial component of 6G. However, the heterogeneous network architecture and resource distribution of future 6G networks present huge challenges for effective and reasonable resource allocation for diverse task requirements. This paper proposes a heterogeneous resource allocation strategy to address these needs in future LEO networks. By employing a multi-agent deep reinforcement learning method based on federated learning, the proposed strategy can achieve self-learning, self-updating and be adaptive to different resource requirements of various tasks. Simulation results demonstrate that the proposed method exhibits superior performance in task completion and overall energy consumption across different resource environments compared to benchmark algorithms.

This article proposes an innovative resource management framework for the next generation heterogeneous satellite networks (HSNs), which can achieve cooperation between independent satellite systems and maximizing resource utilization. The key points of the proposed design lie in the architecture that supports the intercommunication between different satellite systems, and the SDN/NFV-based management offering the matching between resources and services. Based on the framework, we then apply deep reinforcement learning (DRL) into the system due to its strong ability in optimal matching. The two problems of multiobjective reinforcement learning and multiagent reinforcement learning are studied to adapt the development of the HSN. The combination of the DRL and resource allocation achieves integrated resource management across different satellite systems and achieves resource allocation in the HSN which can be implemented more flexibly and efficiently.

This paper proposes a hierarchical multi-agent deep reinforcement learning (HDRL) framework for heterogeneous Low Earth Orbit (LEO) satellite networks. The framework addresses the timing mismatch between the execution and the decision by incorporating delay-aligned traffic predictions into the agent state, structuring the control so that an upper-level agent selects an active set of cells while lower-level agents allocate beams and channels under visibility constraints per-satellite. Using a shared reward composed of demand-weighted efficiency, a duplication penalty for identical cell-channel pairs, and a coverage bonus, the framework reduces early exploration loss and converges to higher rewards, while suppressing resource duplication and jointly improving efficiency and coverage—yielding a practical cooperative policy for heterogeneous LEO systems.

The integration of Non-Terrestrial Networks (NTNs) with Low Earth Orbit (LEO) satellite constellations into 5G and Beyond is essential to achieve truly global connectivity. A distinctive characteristic of LEO mega-constellations is that they constitute a global infrastructure with predictable dynamics, which enables the pre-planned allocation of radio resources. However, the different bands that can be used for ground-to-satellite communication are affected differently by atmospheric conditions such as precipitation, which introduces uncertainty on the attenuation of the communication links at high frequencies. Based on this, we present a compelling case for applying integrated sensing and communications (ISAC) in heterogeneous and multilayer LEO satellite constellations over wide areas. Specifically, we propose a sensing-assisted communications framework and frame structure that not only enables the accurate estimation of the atmospheric attenuation in the communication links through sensing but also leverages this information to determine the optimal serving satellites and allocate resources efficiently for downlink communication with users on the ground. The results show that, by dedicating an adequate amount of resources for sensing and solving the association and resource allocation problems jointly, it is feasible to increase the average throughput by 59% and the fairness by 700% when compared to solving these problems separately.

Space information network (SIN) is difficult to fully utilize the limited on‐board resource due to its dynamic and heterogeneous nature, while the traditional resource management methods cannot adapt to the increasingly diverse task requirements. Space cloud network architecture is an effective technology to reduce the pressure on satellite resources. To effectively manage the space cloud network resources, we design a resources allocation strategy based on resource clustering. Firstly, we propose the space cloud network architecture. Then, we propose a genetic algorithm to cluster the space cloud resources. Finally, we propose a dynamic resource allocation method based on reinforcement learning for the dynamic characteristics of space cloud resources. The method improves the reinforcement learning algorithm through dynamic objective optimization to complete the optimization of multiple objectives in the process of space cloud resources allocation. The simulation results show that the algorithm proposed in this paper reduces the task execution delay by an average of 10.5% compared with the original DQN algorithm and increases the execution success rate by 2.17%.

The sixth generation (6G) satellite twin network is an important solution to achieve seamless global coverage of 6G. The deterministic geometric topology and the randomness of the communication behaviors of 6G networks limit the realism and transparency of cross-platform and cross-object communication, twin, and computing co-simulation networks. Meanwhile, the parallel-based serverless architecture has a high redundancy of computational resource allocation. Therefore, for the first time, we present a new hypergraph hierarchical nested kriging model, which provides theoretical analysis and modeling of integrated relationships for communication, twin, and computing. We explore the hierarchical unified characterization method which joins heterogeneous topologies. A basis function matrix for local flexible connectivity of the global network is designed for the connection of huge heterogeneous systems to decouple the resource mapping among heterogeneous networks. To improve the efficiency of resource allocation in communication, twin, and computing integrated network, a multi-constraint multi-objective genetic algorithm (MMGA) based on the common requirements of operations, storage, interaction, and multi-layer optimal solution conflict is proposed for the first time. The effectiveness of the algorithm and architecture is verified through simulation and testing.

As the spaceborne resources are heterogeneous and the satellite network topology changes constantly, various time-varying derivative graphs are designed to represent the data acquisition and delivery process in satellite-terrestrial integrated networks (STNs). However, the time complexity of resource allocation approaches based on assorted time-varying graphs is remaining obstinately exponential. Derived from the satellite vertical coverage feature, we design a more general relationship to involve a group of nodes in a hyperedge rather than the bilateral relationship between two nodes. In this paper, we propose a time-expanded hypergraph (TEH) to contract the adjacency matrix of the network topology. Based on the proposed TEH, the problem is formulated to minimize the consumption of the communication resource in the resource-limited STN while completing the same number of tasks. Since the problem is intractable by exhaustive search, we further propose a hybrid hyperedge and Lagrangian relaxation algorithm to perform optimal resource allocation through an oriented search for the feasible hyperedges to reduce the scale of searching. The simulation results validate that the proposed algorithm can effectively complete the tasks with lower time complexity.

No abstract available

As a supplement to terrestrial communication networks, satellite edge computing can break through geographical limitations and provide on-orbit computing services for people in some remote areas to achieve truly seamless global coverage. Considering time-varying channels, queue delays, and dynamic loads of edge computing satellites, we propose a multiagent task offloading and resource allocation (MATORA) algorithm with weighted latency as the optimization goal. It is a mixed integer nonlinear problem decoupled into task offloading and resource allocation subproblems. For the offloading subproblem, we propose a distributed multiagent deep reinforcement learning algorithm, and each agent generates its own offloading decision without knowing the prior knowledge of others. We show that the resource allocation problem is convex and can be solved using convex optimization methods. The experiment shows that the proposed algorithm can better adapt to the change of channel and the dynamic load of edge computing satellite, and it can effectively reduce task latency and task drop rate.

The multi-access edge computing (MEC)-enabled low Earth orbit (LEO) satellite network is a promising approach to meet the growing ubiquitous diverse computation demands around the world. In this paper, a joint task offloading and resource allocation strategy is proposed for hybrid MEC-enabled LEO satellite networks, where two types of MEC tasks, namely delay-sensitive edgy-cloud task and data-and computation-intensive cloudy-edge task, are considered simultaneously. Specifically, we first design the cost functions for the two types of tasks, which take the delay-sensitive feature of edgy-cloud task and data-and computation-intensive characteristics of cloudy-edge task into consideration. Then, an overall terminal cost minimization problem is formulated for task offloading and resource allocation under the communication and computation capability constraints and the service delay requirements. In practice, terminals usually only care about their own costs, but satellites pursue the overall cost minimization of all the served terminals. Thus, considering the individual and collective rationality simultaneously, a two-level hierarchical game is constructed to solve the formulated problem. In the upper level, a hedonic coalition formation game is established, which enables each terminal to make the coalition selection and task offloading decision based on the designed coalition switch rule. In the lower level, the joint channel and power allocation in each coalition is first formulated as a noncooperative game to represent the individual rationality of each terminal. Then, each satellite performs the optimal computation resource allocation to maximize the coalition value with collective rationality. We prove that the Nash equilibrium (NE) for the noncooperative game exists and the coalition partition converges to a Nash stable state. Simulation results are provided to demonstrate the superiority of the proposed strategy.

In scenarios like automated warehouses, assigning tasks to robots presents a heterogeneous multi-task and multi-agent task allocation problem. However, existing task allocation study ignores the integration of multi-task and multi-attribute agent task allocation with heterogeneous task allocation. In addition, current algorithms are limited by scenario constraints and can incur significant errors in specific contexts. Therefore, this study proposes a distributed heterogeneous multi-task and multi-agent task allocation algorithm with a time window, called group consensus-based heterogeneous auction (GCBHA). Firstly, this method decomposes tasks that exceed the capability of a single Agent into subtasks that can be completed by multiple independent agents. And then groups similar or adjacent tasks through a heuristic clustering method to reduce the time required to reach a consensus. Subsequently, the task groups are allocated to agents that meet the conditions through an auction process. Furthermore, the method evaluates the task path cost distance based on the scenario, which can calculate the task cost more accurately. The experimental results demonstrate that GCBHA performs well in terms of task allocation time and solution quality, with a significant reduction in the error rate between predicted task costs and actual costs.

As satellite constellations grow in size, there is an increasing need for autonomous, scalable, and real-time dynamic task assignment to address the unique operational challenges of such distributed systems. In particular, a time-varying task assignment (i.e., for observing various regions of Earth) often means that the corresponding satellite has to reorient itself or its sensors, costing time and energy. However, most assignment algorithms for area requests proposed in the literature do not account for the significant cost associated with task handover in satellite constellations. In this paper, we develop a framework for solving the seemingly non-deterministic polynomial-time (NP)-hard problem of optimal dynamic task allocation while minimizing task handover. In particular, we develop Handover-Aware Assignment with Lookahead (HAAL), an algorithm with centralized and distributed variants, and solutions that provably achieve 50% of the optimal value. We then proceed to show that HAAL significantly outperforms similar heuristic methods proposed in the literature for realistic constellation experiments with up to a thousand satellites. The algorithm scales polynomially in the number of satellites/tasks and offers a smooth tradeoff between computational efficiency and performance, allowing the designer to tune the algorithm based on available computing resources, communication bandwidth, and required performance.

Multi-agent systems (MAS) are increasingly applied to complex task allocation in two-sided markets, where agents such as companies and customers interact dynamically. Traditional company-led Stackelberg game models, where companies set service prices, and customers respond, struggle to accommodate diverse and personalised customer demands in emerging markets like crowdsourcing. This paper proposes a customer-led Stackelberg game model for cost-efficient task allocation, where customers initiate tasks as leaders, and companies create their strategies as followers to meet these demands. We prove the existence of Nash Equilibrium for the follower game and Stackelberg Equilibrium for the leader game while discussing their uniqueness under specific conditions, ensuring cost-efficient task allocation and improved market performance. Using the satellite constellation services market as a real-world case, experimental results show a 23% reduction in customer payments and a 6.7-fold increase in company revenues, demonstrating the model's effectiveness in emerging markets.

The Internet of Things (IoT) is one of the most promising applications in the field of computer networking. Edge computing is a computationally efficient method for processing user data in a terrestrial-satellite hybrid environment, where each device is connected exclusively through a low-elevation (LEO) satellite. This paper focuses on an IoT context, introducing methodologies to effectively manage the computation-communication trade-off by strategically distributing processing tasks across various satellites. In particular, an adaptive load balancing approach is considered for efficient utilization of satellite resources. The proposed method can be implemented in a distributed manner, enabling each satellite to evaluate its task handling capacity and forward tasks if it is beyond its capability. The numerical results demonstrate the effectiveness of the proposed method compared to conventional fixed allocation and cloud processing methodologies.

This paper presents an organizational framework for a distributed, self-adaptive avionics architecture targeting Very Low Earth Orbit (VLEO) virtual satellite constellations. The objective is to develop a self-organizing computing platform capable of autonomously allocating tasks while optimizing resource utilization under constraints related to availability, satellite lifespan, and spatial dynamics. To reduce human intervention and eliminate manual mission planning, a Mixed-Integer Linear Programming (MILP) approach is proposed for dynamic and non-aligned satellite constellations. Two different formulations of decision variables are implemented within the MILP model to evaluate their impact on computational performance. The methodology is validated in two stages: first, using a simplified scenario to examine model feasibility and constraint behavior; and second, through application to a realistic virtual satellite constellation. Experimental results demonstrate the effectiveness of the proposed approach in fulfilling operational requirements efficiently, with minimal need for human intervention.

The rapid evolution of satellite networks promises to expand global internet service. However, optimizing task allocation for the efficient and sustainable operation of satellite computing presents complex challenges. Existing approaches usually prioritize energy considerations while neglecting economic aspects, which restricts satellite networks from achieving their full economic potential. In this paper, we address this gap by investigating task allocation in satellite computing. Our approach encourages satellites to consistently provide resources and optimizes battery usage, enabling the completion of more tasks and ultimately maximizing profit. The task allocation approach involves two key components: task pricing and task scheduling. Firstly, we introduce a unique task pricing algorithm that adheres to economic properties, establishing a direct link between satellite utilization and financial income, ensuring economically viable satellite operations. Moreover, we develop two distinct task scheduling algorithms tailored for offline and online scenarios, exploiting dynamic programming and reinforcement learning respectively. Extensive simulations demonstrate that our proposed algorithms effectively enhance task completion rates and optimize total satellite profit.

No abstract available

Orbital edge computing (OEC) is crucial for supporting space intelligence applications within satellite networks. However, individual satellites face resource constraints, and implementing distributed processing techniques, such as federated learning (FL), across multiple satellites introduces significant scheduling complexity. To address these challenges, we first model the key factors influencing complex satellite networks, including satellite constellations, regional resource demands, inter-satellite communication and routing, energy consumption, and battery aging—a novel aspect invoked by OEC operations. We propose an adaptive aggregation method to fundamentally improve communication efficiency in OEC-based FL. To enhance scheduling performance, we formulate a unified optimization problem that jointly considers data partitioning, resource allocation, and aggregation transmission tasks within a decentralized partially observable Markov decision process (Dec-POMDP) framework. Furthermore, we introduce an episodic-phase-recalling reward shaping (EPRS) method to correlate the influences across these phases. Inspired by multi-task learning, we propose an efficient multi-agent reinforcement learning (MARL) algorithm featuring a multi-head actor-critic (MH-AC) network structure and task-equalized adaptation (TEA) technology, designed to optimize latency, energy consumption, network traffic, and battery aging. Extensive experiments validate the effectiveness of the proposed method, showing a 29.9% reduction in total training time, an 11.5% reduction in network traffic, and superior overall performance compared to rule-based methods.

No abstract available

Efficient offloading and resource management are critical for LEO satellite-assisted multi-access edge computing (MEC) systems. This letter introduces a collaborative model based on satellite groups that enables resource sharing between adjacent forward and backward satellites to address load distribution issues. We propose the satellite cooperation and optimization with differential evolution and Dinkelbach algorithm (SCODA), which minimizes the cost of task devices (TDs) by optimizing offloading decisions, resource allocation, and power control. Our scheme ensures flexible load balancing and enhances resource utilization through a multilayer computing framework. Simulation results show that SCODA outperforms benchmarks, reducing costs and improving efficiency.

No abstract available

The rapid growth in satellite internet services, exemplified by constellations like Starlink, has underscored the imperative for efficient service allocation in satellite networks. This poses significant challenges due to the dynamic and multidimensional nature of user demands, which vary across time and space, alongside the fluctuating availability of satellite resources. Traditional static allocation methods are inadequate for these dynamic conditions, often resulting in inefficient resource utilization and compromised service quality. To address these challenges, this paper presents Demand-Oriented Robust Service Allocation (DORSA) with a multi-phase task matching approach. Our method begins with a static matching phase to establish an initial user demand-service match, followed by a dynamic matching phase that leverages a sparse time-varying bipartite graph model to enhance computational efficiency. The DORSA algorithm incorporates local backup-aware matching to rapidly adapt to demand fluctuations and satellite movements, ensuring continuous and high-quality service provision. Simulation results indicate that our approach reduces the average response latency by 74.5% compared to the baseline approach, and by 54.2% compared to the state-of-the-art task allocation algorithm under dynamic demand conditions within satellite networks.

Task scheduling in data relay satellite networks (DRSNs) is subject to dynamic disruptions such as resource failures, sudden surges in task demands, and variations in service duration requirements. These disturbances may degrade the performance of pre-established scheduling plans. To improve adaptability and robustness under such uncertainties, this paper presents a dynamic scheduling model for DRSN that integrates comprehensive task constraints and link connectivity requirements. The model aims to maximize overall task utility while minimizing deviations from the original schedule. To efficiently solve this problem, an ensemble heuristic adaptive contract net protocol (EH-ACNP) is developed, which supports dynamic scheduling strategy adaptation and efficient rescheduling through iterative negotiations. Extensive simulation results show that, in scenarios with sudden task surges, the proposed method achieves a 3.1% improvement in yield compared to the state-of-the-art dynamic scheduling algorithm HMCNP, and it also outperforms HMCNP in scenarios involving resource interruptions. Sensitivity analysis further indicates that the algorithm maintains strong robustness when the disposal rate parameter exceeds 0.2. These results highlight the practical potential of the EH-ACNP for dynamic scheduling in complex and uncertain DRSN environments.

No abstract available

No abstract available

This paper focuses on online management of Distributed Satellite System under dynamic environment. In reality, the inter-satellite communication is limited by practical reasons. The objective is to maximize the total profit of whole system subject to communication time window constraints and observation time window constraints. Firstly, Contract Net Protocol (CNP)-based mechanism is presented to solve this problem. Then, we propose two novel online coordination mechanisms based on Synchronous and Asynchronous communication respectively. The two proposed online coordination mechanisms are compared with CNP-based mechanism. Computational experiments indicate that when the communication resources are scarce, Synchronous Communication-based mechanism is the best choice for the balance between the profit and the communication cost. When the communication resources are relatively sufficient, Asynchronous Communication-based mechanism is preferred for the good performance on profit and percentage of scheduled urgent tasks.

No abstract available

With the rapid development of large‐scale satellite constellations and the intelligent progression of satellites, distributed task planning algorithms have garnered extensive research. However, traditional distributed algorithms face drawbacks such as increased communication burdens and slower response to emergency tasks when dealing with task planning for large satellite clusters. This paper introduces improvements to the traditional contract network protocol to address these shortcomings. Firstly, a method for satellites to determine their own dominant tasks is proposed with the aim of alleviating the superfluous negotiation communication among satellites. Secondly, in the context of batch task upload scenarios, a multitask single‐bid methodology is designed. Furthermore, a provisional task group was constituted to conduct small‐scale intersatellite negotiations for exceptional and urgent tasks for the purpose of minimizing the response time of the system. The simulation results indicate that the algorithm marginally enhances the computing memory of satellites but decreases the intersatellite communication traffic.

Due to the characteristic of fast response, strong pointing ability, and high stability of agile satellites, the space-based situational awareness system composed of agile satellites can enable real-time relay observation of dynamic targets and support early warning of space threats. In order to improve the early warning capability of the constellation for multiple dynamic noncooperative targets and the adaptability in different scheduling scenarios, the cooperative multisatellite scheduling with the pointer network and meta-learning for online optimization is proposed to address the coordinated observation of multiple dynamic targets with online optimization challenges. Specifically, the proposed model with a simplified encoder–decoder structure focuses on enhancing the efficiency of the cooperative scheduling strategy by designing a continuous relay scheduling framework and an improved reinforcement-learning-based policy optimization method with double-critic networks. Moreover, a meta-reinforcement-learning-based increment improving method is designed for continuous adaptability in the fresh scheduling scenario. Ablation and comparative experiments are verified to demonstrate the effectiveness of the proposed cooperative scheduling model.

The significant enhancement of satellite onboard processing capability and drastic proliferation of Artificial Intelligence (AI) applications have fostered decentralized satellite federated learning (DSFL), a transformative paradigm that exchanges and aggregates machine learning (ML) models in satellite clusters for collaborative learning. However, the limited model exchange opportunities caused by intermittent inter-satellite contacts, along with heterogeneous onboard datasets, can lead to ineffective and/or biased model aggregation. To address these issues, it is crucial yet challenging to design an effective DSFL scheduling strategy that determines whether and when to pull models from contact satellites and perform local training for optimizing DSFL performance. In this paper, we propose a contact-based DSFL framework and formulate the DSFL scheduling problem to maximize the accuracy of trained models. As the problem cannot be solved directly, we transform it into a hierarchical Markov game by introducing options for decision agents deployed on individual satellites. Under a learning-to-learn paradigm, we develop a Multi-agent Dueling Double Deep Q Network (MA3DQN)-based intelligent DSFL scheduling strategy. The agents, trained in a distributed and alternating manner, adaptively make scheduling decisions based on instantaneous partial observations of the environment. Simulation results demonstrate the efficiency and adaptability of the MA3DQN-based strategy over three baselines.

Earth observation resources are becoming increasingly indispensable in disaster relief, damage assessment, and other related domains. Many unpredictable factors, such as changes in observation task requirements, bad weather, and resource malfunctions, may cause the scheduled observation scheme to become infeasible. In these cases, it is crucial to promptly reformulate high-quality observation schemes while exerting minimal negative effects on the previously scheduled tasks. Accordingly, in this study, a bottom-up distributed coordination framework together with an improved contract net is proposed, aiming to facilitate dynamic task replanning for heterogeneous Earth observation resources. This hierarchical framework consists of three levels: 1) neighboring resource coordination; 2) single planning center coordination; and 3) multiple planning center coordination. The observation tasks affected by unpredicted factors are managed along with a bottom-up route from resources to planning centers. This bottom-up distributed coordination framework transfers part of the computing load to various nodes of the observation systems to plan tasks more efficiently and robustly. To support the prompt replanning of multiple tasks to proper Earth observation resources in dynamic environments, we propose a multiround combinatorial allocation (MCA) method. Moreover, a new float interval-based local search algorithm is proposed to quickly obtain a promising replanning scheme. The simulation results demonstrate that the MCA method can achieve a better task completion rate for large-scale tasks with satisfactory time efficiency. In addition, this method can efficiently obtain replanning schemes based on original schemes in dynamic environments.

The provision of real-time information services is one of the crucial functions of satellites. In comparison with the centralized scheduling, the distributed scheduling can provide better robustness and extendibility. However, the existing distributed satellite scheduling algorithms require a large amount of communication between satellites to coordinate tasks, which makes it difficult to support scheduling in real-time. This letter proposes a multiagent deep reinforcement learning (MADRL)-based method to solve the problem of scheduling real-time multisatellite cooperative observation. The method enables satellites to share their decision policy, but it is not necessary to share data on the decisions they make or data on their current internal state. The satellites can use the decision policy to infer the decisions of other satellites to decide whether to accept a task when they receive a new request for observations. In this way, our method can significantly reduce the communication overhead and improve the response time. The pillar of the architecture is a multiagent deep deterministic policy gradient network. Our simulation results show that the proposed method is stable and effective. In comparison with the Contract Net Protocol method, our algorithm can reduce the communication overhead and achieve better use of satellite resources.

Aiming at the problem of large conflict and large number of interactions in the task planning algorithm of large-scale satellite constellation based on contract net protocol, this paper proposes an improved contract net algorithm based on target clustering method. The target clustering method is used to allocate the targets with large possibility of conflict to different satellites. Simulation results show that the new method can effectively reduce the number of communications between agents and reduce the task planning time under the condition of ensuring high task completion rate.

PERTEO is a mission proposal intended to provide real-time EO services to reduce Natural Disaster impact by proposing a heterogeneous small satellite constellation and performing in-orbit processing. For that aim, this mission takes advantage of the enhanced capabilities of AI edge processing by enabling on-demand user services through the "satellite as a service" concept. The design of the constellation includes three types of sensors: SAR, Hyperspectral and Multispectral imager VHR combining their capabilities to achieve almost real-time due to their in-orbit distances. Six spacecrafts are considered in each orbit organized in pairs (180° phased platforms) mounting the same instrument. The proposed solution is responsive achieving almost real-time latencies ( Ex: ≲ 1 min, down to seconds is achieved in the first observation product). Persistence is achieved with a low revisit time (≲ 1 hour any payload; ≲ 3 hour s any specific payload choice).

This article explores the tradespace for a constellation of heterogeneous smart satellites intended to measure soil moisture using a combination of L and P band radars, radiometers, and reflectometers. Orbit inclination, repeat cycle, number of satellites, and number of planes were treated as input variables to create a set of architectures for evaluation. Attempting to optimize multiple output variables (cost, average revisit time, maximum revisit time, and percent coverage) results in a complex tradespace with suitable options at various cost caps. Therefore, several cost ranges are examined to find the best constellation for a given cost cap. It was found that a relatively simple constellation of three satellites in one plane offers acceptable performance at a low cost. This preliminary submission shows results for a homogeneous constellation, while the final paper will include satellites with various instrument configurations.

With the progress of technology and the growth of social demand, the application range of civil and commercial earth observation satellites is becoming more and more extensive, the services provided are more diversified and refined, the types of payload are complex and diversified, the satellite resolution is greatly improved, and the observation mission is more mobile and flexible. This makes the data transmission requirements for Earth observation missions more and more diverse, and puts forward higher requirements for network transmission capabilities such as fast response, low delay, high reliability, and high density connection. Based on the development trend of earth observation satellites in recent years, this paper analyzes the data transmission requirements of different types of Earth observation satellites. Driven by demand, a heterogeneous star cluster network architecture featuring hierarchical decoupling, on-demand chain construction and parallel support for dynamic observation tasks is constructed through multi-path comparative analysis. On this basis, This paper proposes an integrated management architecture of observation tasks and network resources. Finally, the key problems and technical challenges in the construction of heterogeneous cluster networks are summarize.

The efficient management of mega-constellation satellite resources and the rapid planning of observation missions are critical driving force for the advancement of space technology. To address the dimensionality explosion problem in the solution space for regional observation mission planning of mega-constellations and to satisfy timely demands, a task planning method based on an A2C (Advantage Actor-Critic) neural network with dynamic temporal relation Mask (TRMA2C) is proposed. Firstly, a discrete state space related to the quality of observation windows is designed, and a hybrid optimization objective function that integrates task completion rate, time window quality, and the timeliness of observation activities is constructed. Secondly, the TRM is designed for application in the process of policy gradient updates and value function estimation. The effectiveness and efficiency of the TRM-A2C method are validated through testing and comparative experimental simulations. This approach thereby provides theoretical and technical support for the operation and management of Chinese mega-constellations.

This paper explores constellation-level autonomous mission planning techniques for distributed network, aiming to address the complexity of mission planning brought about by the increase in the number of remote sensing constellation satellites in the rapidly developing commercial space sector. The article firstly outlines the challenges in constellation mission planning, including the increasing computation requirements, dynamic constellation capacity, the poor adaptability of node destruction, and the difficulty of establishing target selection rules to satisfy the rapid changes of the mission. In this paper, we propose an autonomous mission planning algorithm based on a dual-core processor which includes a mission receiving mechanism, a satellite on-board conflict adjudication based on a point system, a mission management process, a mission conflict processing, and a mission execution and feedback mechanism. The resource consumption and performance indexes of the on-orbit mission planning algorithm are analyzed and the constellations collaborative mission planning algorithm in- orbit validated by using two remote sensing satellites, and the validation results show that that the autonomous mission planning by fast orbital extrapolation is able to execute the imaging tasks precisely.

The rapid expansion of satellite constellations and the growing density of orbital debris have intensified the need for advanced computational tools capable of optimizing orbital architectures, collision avoidance, and long-term space sustainability. Traditional optimization methods are increasingly constrained by the combinatorial complexity of constellation design variables—including orbital planes, phasing, revisit requirements, and communication coverage—and the nonlinear dynamics of debris propagation. This study introduces a hybrid quantum–classical optimization framework that integrates quantum approximate optimization algorithms (QAOA), quantum annealing, and classical multi-physics orbital simulation to improve decision-making in constellation deployment and debris mitigation. Quantum subroutines accelerate the search for globally efficient orbital configurations, while classical solvers handle high-fidelity astrodynamics, propagation models, and mission constraints. Preliminary simulation results indicate that the hybrid framework yields more efficient constellation geometries, reduces collision risk, and enhances debris avoidance planning compared to classical baselines. These findings highlight the potential of quantum-assisted optimization to support safer, more resilient, and sustainably managed space systems.

Modern commercial off-the-shelf (COTS) electronics face increasing reliability challenges, particularly in space applications where harsh radiation environments and the lack of serviceability accelerate degradation. This paper presents a hierarchical health management approach for heterogeneous systems, enabling self-health-awareness and graceful degradation through cross-layer monitoring and control. The proposed architecture distributes fault detection, analysis, and decision-making across multiple levels, from execution units to system-level controllers, integrating hardware- and software-based health monitoring modules. As a case study, we detail the implementation of this approach on Zynq UltraScale+ MPSoC devices, leveraging both the processing system and programmable logic for distributed fault management without compromising mission performance. The design supports predictive fault mitigation, health-aware task scheduling, and adaptation to partial hardware failures, thereby enhancing reliability and extending operational lifetime. Potential applications span aerospace, automotive, telecommunications, and other safety-critical domains, with the proposed framework offering a modular and reusable health management ecosystem adaptable to diverse system architectures.

With the development of natural language and neural network technology, intelligent satellite constellation management methods will have great research value. Traditional management methods rely heavily on the commanders' proficiency in the command system, and there are problems such as comprehension bias due to insufficient level of understanding, long time of task issuance, which bring many difficulties to the subsequent determination of requirements and collaborative planning, and have limitations such as high computation volume and slow response speed when dealing with large-scale constellations. Using natural language processing technology, a natural language processing based architecture is proposed for large-scale remote-sensing constellation management. On this basis, a neural network-based assisted decision-making algorithm is proposed to quickly provide the commander with selectable satellites, optimize the decisionmaking process and shorten the decision-making time, and assist the commander in monitoring the entire mission process. Simulation experiments show that the algorithm has high computational efficiency while the probability that the recommended satellite can actually perform the task reaches 98.52%, and the architecture proposed in this paper has good application prospects in the management of large-scale constellations.

Earth Observation (EO)satellite constellation design deserves further investigation for optimizing configurations that enhance space mission performances. In recent years, there has been considerable interest in reducing the System Response Time (SRT)of EO satellites - the interval between request submission and availability of the image product - in order to achieve rapid response in case of natural or man-made disasters or matters involving defense and natural security. This key performance indicates to the user when, after the request submission, the image produced will be available to him. The best way to improve this performance metric is using heterogeneous constellations, where two different functional constellations are cross-linked; one is mainly for imaging and the other is a communication constellation that is dedicated to relaying commands delivery from Earth station to imaging satellites and data collection back to Earth. This scheme has been proposed before in the previous work to explore its potential enhancement of system performance, or to evaluate the network performances by comparing candidate relay constellations for servicing remote sensing satellites. However, methods for satellite constellation design of this scheme have not been introduced. Since the best heterogeneous configuration may require studying several constellation combinations, this paper presents a framework capable of generating thousands of heterogeneous constellation configurations based on predefined Design Variable (DV)ranges and sizing those configurations in terms of the predefined Measure of Performances (MOPs). Using Systems Tool Kit (STK)and its various add-on modules, we introduce multiple solutions to configure both the imaging and relay constellations of the heterogeneous constellation systems that can achieve their objectives and improve the overall system performance by reducing the SRT. One of these solutions is an imaging constellation of 8 satellites equally distributed in 2 different planes, a Sun-Synchronous Orbit (SSO)and a Mid-Inclination Orbit (MIO). We select this constellation based on the global daily coverage percentage and the satellite optical sensor parameters. In order to reduce the maximum SRT, we select a relay constellation in a Medium Earth Orbit (MEO)on an Equatorial plane and a location of a Ground Station (GS)as a receiving and transmitting Earth site.

: A multi-cluster collaborative cognition-based information system architecture was proposed from four aspects ： overall architecture ， hardware architecture ， software architecture and protocol architecture ， in response to Queqiao constellation system ’ s characteristics ， such as super-large space-time scale ， extreme spatial environment ， and serious limitation of onboard re‐ sources ， as well as the challenges of heterogeneous function fusion and resource collaborative scheduling ， intelligent autono‐ mous operation under intermittent link connection ， and interconnection and elastic networking between satellites. With the perception-decision-execution loop at its center ， it will effectively enhance cislunar space information interaction and intelligent collaboration capabilities ， realize adaptive topology management ， multi-source data fusion ， and autonomous constellation sys‐ tem operation ， and provide reference for the design of future constellation information systems.

Few tools exist for designing constellations of heterogeneous satellites. A new modular tool for total mission design of heterogeneous constellations, including spacecraft design, orbit selection, and launch manifestation, is proposed. The component modules and algorithms are discussed, including a novel crossover method for genetic algorithms and a novel constraint formulation for launch manifestation of maneuverable vehicles. Finally, the expandability of the tool to multiple domains and various applications is highlighted. The main benefit of the described tool is that it allows analysts to quickly and easily design complex system architectures for space systems with a variety of objectives, providing cost savings and enabling timely responses to changing mission needs.

In this paper, we argue that deploying many mission-specific satellite mega-constellations incurs significant monetary and environmental costs. Instead, we propose launching a single or a small number of mega-constellations equipped with heterogeneous computing, communication, storage, and sensing capabilities, allowing them to offer a broad range of services to customers who no longer need to launch their own satellites. We argue that the hardware technology for building such platforms is already widely accessible. Thus, we highlight the algorithmic and systems challenges that the community needs to address to enable cost-efficient and secure constellation-as-a-service platforms. We also develop a simulator that allows for experimenting with different scheduling algorithms for constellation-as-a-service platforms.

As a significant part of spacecraft, earth observation satellites play an important role in space information applications. Recently, user demands and scale of satellite constellations have considerably increased. Multi-satellite earth observation mission scheduling has become a practical problem that needs to be solved urgently in the current satellite application field. Aiming at the problem of heterogeneous satellite earth observation mission scheduling, this paper analyzes the earth observation satellite scheduling resource constraints, establishes an optimized objective function, and proposes a multi-satellite mission scheduling model. On this basis, an adaptive genetic and tabu hybrid search algorithm is proposed, the hybrid algorithm introduces an adaptive trimming strategy in solution space that can effectively overcome the shortcomings of genetic algorithm's weak local search ability and strong initial solution dependence of tabu search, and flexibly respond to different scales of scenarios. The simulation results show that the hybrid algorithm can obtain a global optimal solution in relatively high speed with average task completion rate of 94.6% above, and the scalability, adaptability, efficiency and effectiveness of the model is verified.

No abstract available

We propose a distributed model predictive control (MPC) framework for coordinating heterogeneous, nonlinear multi-agent systems under individual and coupling constraints. The cooperative task is encoded as a shared objective function minimized collectively by the agents. Each agent optimizes an artificial reference as an intermediate step towards the cooperative objective, along with a control input to track it. We establish recursive feasibility, asymptotic stability, and transient performance bounds under suitable assumptions. The solution to the cooperative task is not predetermined but emerges from the optimized interactions of the agents. We demonstrate the framework on numerical examples inspired by satellite constellation control, collision-free narrow-passage traversal, and coordinated quadrotor flight.

The rapid advancements in computing and communication capabilities of Low Earth Orbit (LEO) satellites have made it feasible to execute complex and collaborative inorbit computation missions. Transformer-based large AI models (LAMs), known for their exceptional performance in in-context learning (ICL) and prompt-based reasoning, have attracted significant attention, providing powerful intelligence across sectors such as industry and aerospace. However, the significant parameter volume of LAMs poses a substantial challenge for direct deployment on satellites with constrained computing power and energy provision. To address this, the looped Transformer model reduces parameter requirements through layerwise parameter sharing, achieving performance comparable to vanilla Transformer-based LAMs in ICL tasks. Despite this efficiency, the limited and heterogeneous space-borne computing and storage capabilities complicate the orchestration for balanced workload allocation during multi-satellite cooperation. In this paper, we propose SAI, a collaborative multi-satellite space AI system that exploits the memory efficiency of the looped Transformer and the inherent parallelism in batch data processing. SAI enables accelerated on-satellite inference by integrating heterogeneous onboard resources and introducing a novel hybrid approach combining data and pipeline parallelism. This approach supports cross-satellite cooperation with parallelism planning and asynchronous inter-batch overlapping, significantly reducing inference latency and enhancing resource efficiency. Furthermore, SAI optimizes inference latency by formulating it as a shortest-path problem, effectively solved via Dijkstras algorithm. Extensive evaluations demonstrate SAIs superior performance in reducing inference latency and runtime memory usage compared to existing baselines.

This work investigates resource optimization in heterogeneous satellite clusters performing autonomous Earth Observation (EO) missions using Reinforcement Learning (RL). In the proposed setting, two optical satellites and one Synthetic Aperture Radar (SAR) satellite operate cooperatively in low Earth orbit to capture ground targets and manage their limited onboard resources efficiently. Traditional optimization methods struggle to handle the real-time, uncertain, and decentralized nature of EO operations, motivating the use of RL and Multi-Agent Reinforcement Learning (MARL) for adaptive decision-making. This study systematically formulates the optimization problem from single-satellite to multi-satellite scenarios, addressing key challenges including energy and memory constraints, partial observability, and agent heterogeneity arising from diverse payload capabilities. Using a near-realistic simulation environment built on the Basilisk and BSK-RL frameworks, we evaluate the performance and stability of state-of-the-art MARL algorithms such as MAPPO, HAPPO, and HATRPO. Results show that MARL enables effective coordination across heterogeneous satellites, balancing imaging performance and resource utilization while mitigating non-stationarity and inter-agent reward coupling. The findings provide practical insights into scalable, autonomous satellite operations and contribute a foundation for future research on intelligent EO mission planning under heterogeneous and dynamic conditions.

The increasing demand for comprehensive, accurate, and rapid satellite remote sensing data has necessitated the development of heterogeneous satellite constellations. These constellations, equipped with diverse payloads, are designed for cooperative ground observation. Unlike traditional homogeneous constellations, heterogeneous constellations often incur higher risks due to varying altitudes and inclinations among their sub-constellations. This diversity increases the likelihood of potential disruptions and necessitates significant long-term maintenance costs. This paper introduces a novel approach to constellation design that concurrently optimizes observation capabilities and maintenance costs. The design challenge is reformulated as a multi-objective optimization (MO) problem, tackled using a combination of System Tool Kit (STK) and MATLAB integration. The approach utilizes the Non-dominated Sorting Genetic Algorithm II (NSGA-II) alongside a fuzzy set algorithm to achieve a balance across multiple criteria. The optimization model considers four main objectives: minimizing the average maximum revisit time, maximizing the total duration of inter-satellite link connectivity, and reducing the maintenance costs associated with constellation configuration, which includes minimizing the total drift in right ascension of the ascending node and satellite altitude adjustments. Two sets of simulations are conducted to evaluate the method: one using the multi-objective optimization framework (MO) and the other based on a single-objective optimization (SO). The results reveal that while the MO-based designs result in a longer average maximum revisit time compared to SO, they excel in terms of inter-satellite connectivity. Additionally, these designs demonstrate a reduction in the overall maintenance costs. This showcases the effectiveness of the proposed method in optimizing trade-offs between different operational objectives, thus enhancing the sustainability and efficiency of satellite constellations.

Beam hopping and power allocation are key issues in Multi-Beam Satellite (MBS) systems. However, existing research could not effectively simulate the dynamics of satellite systems and reduce the problem scale. Therefore, this paper proposes a beam hopping and power allocation method based on heterogeneous graphs, aiming to optimize the utilization of resources and traffic throughput in MBS with minimal consumption. First, we construct a heterogeneous graph model to represent the satellite system, where nodes represent illumination cells and satellite beams, and edges represent communication links between beams and illumination cells. Then, the beam hopping and power allocation problem is modeled as a Markov Decision Process (MDP) and extended to a heterogeneous graph-based dynamic beam pattern and power allocation model (HGBP). Finally, the results of simulation experiments prove that compared to the benchmark algorithm, the proposed HGBP can effectively improve system throughput and energy efficiency.

No abstract available

The cutting-edge applications of cyber-physical power systems (CPPS) must transmit large volumes of data packets collected by massive remote terminal units (RTUs) to the control center. To develop high-reliability and self-sustainable communication networks for the RTUs deployed in hard-to-reach areas, we propose an RTU satellite-terrestrial multi-hop network with energy cooperation for remote CPPS. Specifically, data packets generated by RTUs are either transmitted to faraway base station (BS) in a multi-hop manner or uploaded to satellite network, and each RTU harvests ambient renewable power with the capacity to transfer harvested energy to the relay RTU. We then develop a multi-agent learning-based packet routing and energy cooperation approach (MAQMIX-PREC) to maximize the network throughput by jointly optimizing relay selection, sub-slot partition, and channel allocation. This approach effectively decouples the decision-making and coordinates the training among RTUs in the RTU multi-hop network. Experimental evaluations illustrate that the proposed approach achieves congestion-awareness and energy cooperation, and outperforms benchmark methods in terms of training convergence, network throughput, and traffic intensity.

Many multi-robot applications require allocating a team of heterogeneous agents (robots) with different abilities to cooperatively complete a given set of spatially distributed tasks as quickly as possible. We focus on tasks that can only be initiated when all required agents are present otherwise arrived agents would be waiting idly. Agents need to not only execute a sequence of tasks by dynamically forming and disbanding teams to satisfy/match diverse ability requirements of each task but also account for the schedules of other agents to minimize unnecessary idle time. Conventional methods, such as mix-integer programming generally require centralized scheduling and a long optimization time, which limits their potential for real-world applications. In this work, we propose a reinforcement learning framework to train a decentralized policy applicable to heterogeneous agents. To address the challenge of complex cooperation learning, we further introduce a constrained flashforward mechanism to guide/constrain the agents' exploration and help them make better predictions. Through an attention mechanism that reasons about both short-term cooperation and long-term scheduling dependency, agents learn to reactively choose their next tasks (and subsequent coalitions) to avoid wasting abilities and to shorten the overall task completion time (makespan). We compare our method with State-of-the-Art heuristic and mixed-integer programming methods, demonstrating its generalization ability and showing it closely matches or outperforms these baselines while remaining at least two orders of magnitude faster.

Coalition structure is an effective cooperation architecture for task implementation in the field of multi-robot systems. Nevertheless, the presence of uncertainty inherently complicates the decision-making process for robots and may potentially result in suboptimal coordination. To tackle this challenge, this paper considers an uncertain multi-robot coalition formation scenario in which task information (the types of all tasks) is incompletely known to robots. Given the local beliefs of robots, the problem of multi-robot coalition formation under uncertainty is formulated as a coalition formation game. In this game, each robot is a rational and self-interested player and tends to join a coalition according to its preference. A polynomial-time coalition formation algorithm is proposed to identify the social agreement, i.e., Nash stable partition, among the robots. The convergence of the proposed algorithm is strictly guaranteed as long as the communication topology of the considered system is strongly connected. The coalition formation game is then extended to a dynamic game, and we propose a belief updating algorithm that enables robots to update their beliefs as the game is played repeatedly. Simulation results demonstrate the effectiveness of our proposed algorithms, and the robots will eventually learn the true type of each task.Note to Practitioners—The work reported in this article will be beneficial for deploying multi-robot systems to support cooperative surveillance applications. In these scenarios, robots can form stable coalitions to perform surveillance tasks, even when the task information is incompletely known due to sensor noise or limited sensor range. The problem of multi-robot coalition formation is known to be NP-hard, which becomes even more challenging when uncertainty is taken into account. This paper proposes game-based algorithms to solve the problem of multi-robot coalition formation under uncertainty. Some practical schemes are introduced as benchmarks to further illustrate the performance improvement brought by our proposed algorithms. The proposed algorithms are further evaluated through a real-world experiment, demonstrating their effectiveness for practical application.

No abstract available

Multi-robot systems have demonstrated significant potential in accomplishing complex tasks, such as cooperative pursuit, search-and-rescue operations. The emergence of heterogeneous robots with diverse capabilities and characteristics shows superior adaptability compared with homogeneous teams. However, in practical applications, global information is typically inaccessible, and composite teams must contend with partial observability and coordination difficulties. To address the issue in heterogeneous multi-robot systems, we propose a novel Intention-Guided reinforcement learning approach with Dirichlet Energy constraint (IGDE). Specifically, an intention-guided module is designed to derive long-horizon strategies based solely on local observations, enabling foresighted decision-making. In addition, a Dirichlet energy constraint is incorporated into the communication process to enhance the diversity of environmental cognition among different classes of robots. Heterogeneous robots perform class-aware actions driven by distinct cognitive representations, thereby enhancing cooperative efficiency. Notably, our approach alleviates the need of prior knowledge and heterogeneity modeling. Extensive comparative experiments and ablation studies verify the effectiveness of the proposed framework. Additionally, real-world deployment is conducted to demonstrate the practicality.