Task Offloading and Resource Allocation Strategies Based on Proximal Policy Optimization

In order to meet the service demands of new network applications and scenarios, computational offloading techniques in MEC have become a popular research issue. Aiming at solving the problems of delay deterioration and server resource scarcity caused by multi-users offloading tasks on edge servers at the same time, this paper firstly constructs a multi-terminal and single-server offloading model for tasks based on the queuing theory.Considering the user's delay experience under the delay-sensitive condition, introduces the weighted sum of the waiting probability and the task completion time as the optimization objective. On this basis, a proximal policy optimization-based computational offloading and resource allocation strategy is proposed. Finally, the simulation findings indicate that the algorithm proposed in this paper exhibits superior performance in optimizing the delay experience of user when compared to alternative algorithms.