Resource allocation is an important problem influencing the service quality of multi-beam satellite communications.In multi-beam satellite communications, the available frequency bandwidth is limited, users requiremen...Resource allocation is an important problem influencing the service quality of multi-beam satellite communications.In multi-beam satellite communications, the available frequency bandwidth is limited, users requirements vary rapidly, high service quality and joint allocation of multi-dimensional resources such as time and frequency are required. It is a difficult problem needs to be researched urgently for multi-beam satellite communications, how to obtain a higher comprehensive utilization rate of multidimensional resources, maximize the number of users and system throughput, and meet the demand of rapid allocation adapting dynamic changed the number of users under the condition of limited resources, with using an efficient and fast resource allocation algorithm.In order to solve the multi-dimensional resource allocation problem of multi-beam satellite communications, this paper establishes a multi-objective optimization model based on the maximum the number of users and system throughput joint optimization goal, and proposes a multi-objective deep reinforcement learning based time-frequency two-dimensional resource allocation(MODRL-TF) algorithm to adapt dynamic changed the number of users and the timeliness requirements. Simulation results show that the proposed algorithm could provide higher comprehensive utilization rate of multi-dimensional resources,and could achieve multi-objective joint optimization,and could obtain better timeliness than traditional heuristic algorithms, such as genetic algorithm(GA)and ant colony optimization algorithm(ACO).展开更多
Considering the demand of distributed satellite clusters for high-speed information communication in the future,this paper establishes a laser network model based on optical multibeam antenna.At present,there are stil...Considering the demand of distributed satellite clusters for high-speed information communication in the future,this paper establishes a laser network model based on optical multibeam antenna.At present,there are still some networking and reconstruction problems,such as network connectivity,duration,and stability.To address them,the paper develops a multiobjective optimization model for the laser networking of distributed satellite clusters,which aims to maximize network connectivity and network duration and minimize the perturbation of the network connection matrix.The model is constructed under the constraints of multibeam antenna capability,the visibility of satellites in clusters,and network connectivity.From the perspectives of the optimization effect and timeliness of the optimization algorithm,a deep reinforcement learning algorithm is proposed,which is based on a double-layer Markov decision model,to meet the needs of on-orbit intelligent networking and dynamic reconstruction of distributed satellite clusters.Simulation results show that the algorithm features flexible architecture,excellent networking performance,and strong real-time performance.When the optimization results are similar,the proposed algorithm outperforms the nonsorted genetic algorithm II algorithm and the particle swarm optimization algorithm in terms of solution speed.展开更多
基金supported by the National Key Research and Development Program of China under No. 2019YFB1803200。
文摘Resource allocation is an important problem influencing the service quality of multi-beam satellite communications.In multi-beam satellite communications, the available frequency bandwidth is limited, users requirements vary rapidly, high service quality and joint allocation of multi-dimensional resources such as time and frequency are required. It is a difficult problem needs to be researched urgently for multi-beam satellite communications, how to obtain a higher comprehensive utilization rate of multidimensional resources, maximize the number of users and system throughput, and meet the demand of rapid allocation adapting dynamic changed the number of users under the condition of limited resources, with using an efficient and fast resource allocation algorithm.In order to solve the multi-dimensional resource allocation problem of multi-beam satellite communications, this paper establishes a multi-objective optimization model based on the maximum the number of users and system throughput joint optimization goal, and proposes a multi-objective deep reinforcement learning based time-frequency two-dimensional resource allocation(MODRL-TF) algorithm to adapt dynamic changed the number of users and the timeliness requirements. Simulation results show that the proposed algorithm could provide higher comprehensive utilization rate of multi-dimensional resources,and could achieve multi-objective joint optimization,and could obtain better timeliness than traditional heuristic algorithms, such as genetic algorithm(GA)and ant colony optimization algorithm(ACO).
文摘Considering the demand of distributed satellite clusters for high-speed information communication in the future,this paper establishes a laser network model based on optical multibeam antenna.At present,there are still some networking and reconstruction problems,such as network connectivity,duration,and stability.To address them,the paper develops a multiobjective optimization model for the laser networking of distributed satellite clusters,which aims to maximize network connectivity and network duration and minimize the perturbation of the network connection matrix.The model is constructed under the constraints of multibeam antenna capability,the visibility of satellites in clusters,and network connectivity.From the perspectives of the optimization effect and timeliness of the optimization algorithm,a deep reinforcement learning algorithm is proposed,which is based on a double-layer Markov decision model,to meet the needs of on-orbit intelligent networking and dynamic reconstruction of distributed satellite clusters.Simulation results show that the algorithm features flexible architecture,excellent networking performance,and strong real-time performance.When the optimization results are similar,the proposed algorithm outperforms the nonsorted genetic algorithm II algorithm and the particle swarm optimization algorithm in terms of solution speed.
