低轨卫星物联网中基于PPO-CRA的任务卸载与资源分配算法OA

To address the limited computational capacity of lightweight ground terminals and constrained satellite communication in satellite Internet of Things(SIoT)scenarios,this paper proposes a three-tier SIoT network architecture with an edge control gateway(ECG)as relay,and develops a task offloading and resource allocation algorithm combining proximal policy optimization with convex resource allocation(PPO-CRA).The optimization objective of minimizing system cost is decomposed into two subproblems:task offloading,modeled as a Markov Decision Process and solved via PPO to dynamically generate offloading strategies;and resource allocation,addressed through Lagrange multiplier methods for convex optimization.Simulation results demonstrate that PPO-CRA achieves strong performance in system cost reduction with robust adaptability and stability across varying environmental conditions.