CROSS: A Crowdsourcing based Sub-Servers Selection Framework in D2D Enhanced MEC Architecture

In a Device-to-Device (D2D) mobile edge computing system, the mobile devices can share computation and communication resources with each other. A mobile device can offload its computation task to a nearby mobile device. In order to release the computation workload of mobile edge servers and enhance the mobile edge service coverage quality, a set of mobile devices can be selected as sub-servers and provide mobile edge service to nearby devices. Based on the crowdsourcing technique, we have proposed a framework, CROSS, to select sub-servers in a D2D enhanced mobile edge computing system. Two major problems in CROSS framework, the sub-server selection problem and the payoff allocation problem, have been formulated. The first problem is proved to be NP-Hard and solved by an approximation algorithm. The second problem is solved by an auction mechanism. The performance of the CROSS framework is evaluated by experiments. The experimental results show that the CROSS framework is efficient and effective.