A Socio-Technical Approach to Capacity Maximization for Device-to-Device Relay Selection

Device-to-Device (D2D) relaying is considered a promising technology to increase the data rates in next generation networks. We consider the D2D relay selection problem in which cell-edge mobile devices (CMDs), having bad channel conditions to the access point (AP), may forward their data to the AP via relaying mobile devices (RMDs) with better channel conditions. For this purpose, the RMDs sacrifice a fraction of their communication bandwidth and energy to relay the data of the CMDs. A key challenge in D2D relaying is to increase the willingness of RMDs to act as relays to CMDs. To overcome this challenge, considering the technical perspective of bandwidth allocation and transmit power optimization is not enough. In addition, the social perspective is important with the users' different individual motivations to participate, such as strong social relationships between CMDs and RMDs and an altruistic motivation to help CMDs. In this paper, we address the D2D relay selection problem with a socio-technical approach, i.e., we consider the RMDs as individual decision makers whose participation decision is influenced by its preferences regarding technical and social motivations. Furthermore, we formulate a relay selection problem to maximize the expected capacity under the a priori unknown decisions of the RMDs regarding their participation. To solve this problem, we propose a novel decentralized, preference-aware D2D relay selection algorithm, termed DPA-D2D, which is based on game theory. We show that the CMDs' capacity gain is more than 40 % higher compared to state-of-the-art D2D relay selection algorithms.