Social-Aware D2D Relay Networks for Stability Enhancement: An Optimal Stopping Approach
He Zhang, Zehua Wang, Qinghe Du · IEEE Transactions on Vehicular Technology · 2018
Device-to-device (D2D) relay network is regarded as a promising approach to extend the coverage area of cellular networks or improve the service quality for cell-edge users. However, the mobility of relay devices will negatively affect the stability of D2D communications. Aiming to improve the stability in mobile D2D communications and hence deliver more data traffic in D2D relay networks, in this paper, we exploit users' social behaviors to dynamically detect better relay devices and substitute them for the existing ones. As the mobility of relays pose a critical challenge in achieving the tradeoff between the relay probing time and the optimality of relay selection, we propose the optimal stopping approach for mobile devices conducting the relay re-selection. Specifically, we embed social tie and centrality relationships between a potential relay and the end user to predict the date delivery performance of the potential relay in the D2D relay network. Our optimal stopping approach can be applied to different network topologies for mobile users with different social behaviors. The simulation results show that our proposed scheme can achieve the optimal tradeoff between the time used in relay probing and the optimality of the selected relay. Results also show that both short-term and long-term stability can be obtained and the amount of relayed data traffic can be maximized.