Task Migration Optimization Algorithm in Mobile Edge Computing.

With the rise of the Internet of Things(IoT) and big data, more and more mobile devices and sensors are generating massive amounts of data that need to be processed and analyzed. Mobile edge computing improves the quality of service by setting servers at the edge of the network to provide services to end devices, thereby enhancing the quality of user experience. This paper presents a migration optimization scheme for multi-mobile users in a multi-edge server scenario, modeling the multi-user task migration process as a Markov decision process and considering the user mobility factor, with joint user states as states, joint migration decisions as actions, and global rewards as rewards to fully consider the energy consumption constraints in the migration process. In this paper, we propose a K-PPO algorithm based on deep reinforcement learning, which introduces a centralized critical critic network and counterfactual benchmark based on the PPO algorithm to improve the efficiency of policy learning. The goal of the algorithm is to minimize the average completion time of all tasks. Through comparative experiments and a series of evaluations with existing baseline methods, this paper demonstrates the effectiveness of the proposed task migration algorithm.