Matching-Theory-Based Multi-User Cooperative Computing Framework

In this paper, we propose a matching theory based multi-user cooperative framework to minimize the overall energy consumption of all the user equipments (UEs), where the UEs can be classified into the following roles: resource demander (RD), resource provider (RP), and standalone UE (SU). We first determine the role of the UE by leveraging the roommate matching method. Then, by adopting college admission based algorithm, we divide the UEs into multiple cooperation groups, each consisting of one RP and multiple RDs. Next, we propose the rotation swap operation to further improve the performance without deteriorating the system stability. Finally, we present an effective task offloading algorithm to minimize the energy consumption of all the cooperation groups. The simulation results verify the effectiveness of the proposed scheme.