Computation Offloading and Resource Allocation in UAV-Assisted Satellite Network Systems

In order to support the Internet of Things (IoT) systems, it is necessary to allow IoT user equipment (UE) to offload its delay-sensitive computational tasks to a server both UAV-assisted networks and satellite networks have been considered for computation offloading but mostly studied separately. In this paper, we investigate computation offloading and resource allocation in a UAV-assisted satellite network, where an IoT UE can offload its task to a UAV-carried edge server or a satellite edge server. To minimise the maximum total service delay of all IoT UEs, which includes the transmission delay and processing delay, we propose a Q-learning-based task offloading decision optimisation algorithm in conjunction with the computation resource allocation strategies for the UAV-assisted satellite network. Simulation results show that our proposed optimisation algorithm achieves a much lower total service delay of all IoT UEs than the benchmarks.