QoS-Aware Service Migration in Edge Cloud Networks

Abstract

Recently, the edge cloud computing has gained a wide attention, because its proximity to users can provide low latency services and reduce network traffics volumes. However, the performance of edge cloud is highly affected by user's mobility. Therefore, to prevent the performance degradation, we propose QoS aware service migration scheme in edge-central cloud network, where the optimal target server is selected and service migration is conducted with the consideration of service migration and network cost, QoS and resource utilization in server. To select the optimal target server, we formulate a Markov decision process (MDP). Evaluation results demonstrate that the proposed scheme can reduce migration cost and network traffic while required QoS in terms of latency is guaranteed.