通感算融合网络中差异化任务卸载与资源分配联合研究OA

To address the challenges of heterogeneous user demands and low resource allocation efficiency in integrated sensing,communication,and computing(ISCC)networks,this paper proposes a mobile edge computing(MEC)-assisted ISCC framework that incorporates service caching,differentiated task offloading,and resource allocation.Under constraints of data rate and sensing accuracy,the objective is to minimize the overall task processing cost by jointly optimizing service caching,task offloading,and resource allocation.The optimization problem is reformulated as a partially observable Markov decision process,and a multi-agent deep deterministic policy gradient(MADDPG)algorithm is employed to adaptively meet diverse user requirements.The algorithm autonomously determines optimal strategies for service caching,task offloading,and resource allocation.Compared with existing task offloading methods,the proposed MADDPG-based approach reduces the average task processing cost by approximately 13.25%and improves the local cache hit rate by about 18.33%,demonstrating its effectiveness in enhancing system performance.