S2R: Service Trading Based Response Time Optimization in Mobile Edge Computing

摘要

New caching forms coupled with Mobile Edge Computing (MEC) supports new computing architecture at the physical proximity of the end-users. One of the similar architectures is mobile cloudlet, where a smartphone offers computations to another smartphone. However, the mobile edge computing architecture still has lots of issues related to response time optimization and resource provisioning. This paper focuses on a novel auction strategy to optimize resource utilization and as a result, proposed scheme minimizes the response time. The auction scheme is developed using multiple-leader and multiplefollower Stackelberg game model, which optimizes the utility of both users and mobile edges using non-cooperative approach for users and co-operative approach for mobile edges. Finally, the proposed scheme is simulated in NS3 environment and compared with existing schemes to validate the performance of proposed scheme. We observe that with the proposed scheme, there is a steep increase in overall utility in terms of response time and resource utilization.