Socially optimal distributed user association for multi-hop machine-to-machine communications

In order to accommodate a massive number of machine-to-machine (M2M) communication devices (MTCDs) in cellular networks, we propose an efficient association and multi-hop relaying scheme, with rate control, between users and access points. Overall transmission rate is maximized and power consumption is minimized, while ensuring load balancing and proportional fairness among users. In this context a distributed algorithm is proposed that solves this joint association and relaying problem, only requiring limited control information exchange between nearby users. With formulation as a game, it is shown that the proposed algorithm converges to a social optimum, with a subgame perfect equilibrium. Numerical simulation results show significant improvements in the network data rate, by a factor of at least 6, and fairness, up to a factor of 14, among the MTCDs by the proposed association method when compared to other potential schemes.