Research on DRL-based Computation Offloading Algorithm in Integrated Terrestrial-Satellite Networks
With the rapid development of the Low Earth Orbit(LEO)satellite network and Mobile Edge Computing(MEC)tech-nology,by deploying MEC servers in LEO satellites,computation offloading services can be provided for remote areas where there is a lack of terrestrial MEC servers.However,as the number of ground users increases,the complexity of integrated terrestrial-satellite net-work computation offloading scenarios has grown significantly.Existing studies have difficulties dealing with scenarios characterized by high arrival rates and complex tasks.To solve this problem,a Deep Reinforcement Learning(DRL)-based Parallel Computation Offload-ing(DPCO)algorithm is proposed.This algorithm models the computation offloading problem with the optimization objective of minimi-zing the total offloading delay.It also considers the impact of Amdahl's law on computational performance and allocates computing re-sources of satellite MEC servers to enable multi-task parallel processing.Additionally,the DPCO algorithm transforms the model into a Markov Decision Process(MDP)and solves it using the Advantage Actor-Critic(A2C)algorithm.Finally,the performance of the DP-CO algorithm is verified by simulation.Simulation results show that the algorithm effectively addresses existing methods'deficiencies and provides valuable references for designing computation offloading algorithms in integrated terrestrial-satellite networks.
国家科技期刊平台
NSTL
万方数据
