MIMO-OFDM Based Energy Harvesting Cooperative Communications Using Coalitional Game Algorithm

In this paper, we consider the problem of cooperative communication between relays and base station in an advanced MIMO-OFDM framework, under the assumption that the relays are supplied by electric power drawn from energy harvesting (EH) sources. In particular, we focus on the relay selection, with the goal to guarantee the required performance in terms of capacity. In order to maximize the data throughput under the EH constraint, we model the transmission scheme as a non-transferable coalition formation game, with characteristic function based on an approximated capacity expression. Then, we introduce a powerful mathematical tool inherent to coalitional game theory, namely: the Shapley value (Sv) to provide a reliable solution concept to the game. The selected relays will form a virtual dynamically-configured MIMO network that is able to transmit data to destination using efficient space-time coding techniques. Numerical results, obtained by simulating the EH-powered cooperativeMIMO-OFDMtransmission with Algebraic Space-Time Coding (ASTC), prove that the proposed coalitional game-based relay selection allows to achieve performance very close to that obtained by the same system operated by guaranteed power supply. The proposed methodology is finally compared with some recent related state-of-the-art techniques showing clear advantages in terms of link performance and goodput.