Low earth orbit (LEO) satellites with wide coverage can carry mobile edge computing (MEC) servers with computing power to form the LEO satellite edge computing system, providing computing services for ground users that cannot access the core network. This paper studies the joint optimization problem of communication and computing resource in the LEO satellite edge computing system to minimize the utility function value of the system. Due to the fact that, general optimization tools cannot effectively solve this problem, this paper proposes a deep learning-based bandwidth allocation algorithm. The bandwidth allocation schemes are generated through multiple parallel deep neural networks (DNNs). The utility function values of the system are calculated according to the derived optimal CPU cycle frequency and optimal user transmission power. The bandwidth allocation scheme corresponding to the optimal system utility function value is stored in the memory to further train and improve all DNNs. The simulation results show that the proposed algorithm can achieve good convergence effect and the algorithm proposed in this paper outperforms the other four comparison algorithms with low average time cost.
Huge low earth orbit (LEO) satellite networks can achieve global coverage with low latency. In addition, mobile edge computing (MEC) servers can be mounted on LEO satellites to provide computing offloading services for users in remote areas. A multi-user multi-task system model is modeled and the problem of user’s offloading decisions and bandwidth allocation is formulated as a mixed integer programming problem to minimize the system utility function expressed as the weighted sum of the system energy consumption and delay. However, it cannot be effectively solved by general optimizations. Thus, a deep learning-based offloading algorithm for LEO satellite edge computing networks is proposed to generate offloading decisions through multiple parallel deep neural networks (DNNs) and store the newly generated optimal offloading decisions in memory to improve all DNNs to obtain near-optimal offloading decisions. Moreover, the optimal bandwidth allocation scheme of the system is theoretically derived for the user’s bandwidth allocation problem. The simulation results show that the proposed algorithm can achieve a good convergence effect within a small number of training steps, and obtain the optimal system utility function values compared with the comparative algorithms under different system parameters, and the time cost of the system and DNNs is very satisfactory.