Service Migration Strategy Based on Multi-Attribute MDP in Mobile Edge Computing

In order to solve the problem of service interruption caused by user movement and limited range edge nodes, a migration algorithm based on multi-attribute Markov decision process was proposed for mobile computing. By performing migration, distance between is always kept small range. addition, in prevent quality from being affected frequent users, return function model defined comprehensively considering quality, resource demand service, cost, income users each node, premise taking into account cost conditions, which did not only make up deficiency scheme solely distance. The number migrations also reduced, single target server no longer used. candidate set constructed user’s motion trajectory, Q-Learning used problem. Simulation results show that can reduce ensure balance migrations.