Optimization Strategy for Task Offloading and Service Caching in Mobile Edge Computing Based on Deep Reinforcement Learning

This paper proposes a joint optimization solution for task offloading and service caching in a multi-edge server environment to address the issue of limited computational and storage resources in edge computing. A resource-constrained optimization model is formulated and transformed into a mixed-integer nonlinear programming problem. To tackle this problem, an improved Deep Deterministic Policy Gradient (DDPG) algorithm is introduced. This algorithm utilizes deep neural networks to approximate the policy and value functions and incorporates experience replay and target networks to enhance training stability. The agent interacts with the environment, observes the current state, selects an action accordingly, and receives a reward. The policy network is updated based on the observed reward using experience replay. Experimental results demonstrate that the proposed DDPG-based joint optimization method achieves excellent performance in terms of reducing task execution latency and alleviating network load in the edge computing environment. In the context of edge computing, the limited computational and storage resources of edge servers can lead to increased task processing latency when too many tasks are offloaded to them.