Maximizing energy efficiency in analog network coding based two-way relay-assisted system

In this paper, we investigate the energy efficiency maximization problem of analog network coding (ANC) based two-way relay (TWR) assisted cellular network under quality of service (QoS) constraints through transmit power optimization. Circuit power cannot be ignored since it influences energy efficiency greatly. However, the difficulty is that maximizing energy efficiency is shown to be non-convex in TWR-assisted network, and cannot be equivalent to the convex minimizing transmit power because of the effect of circuit power. A two-stage maximum energy efficiency (MEE) method is proposed to tackle this problem effectively. Closed form expressions of both the optimal transmit power and energy efficiency are derived. We prove the optimality of MEE method through both analytical and simulation results, and show that it can increase energy efficiency significantly. Furthermore, it is theoretically proved that the optimal relay selection (RS) criterion is the Max-Harmonic-Mean (MHM) criterion in multiple relay scenarios.