Interference Suppression and Energy Efficiency Improvement With Massive MIMO and Relay Selection in Cognitive Two-Way Relay Networks

We analyze a relay assisted wireless communication link between two underlay massive multiple-input multiple-output (MIMO) terminals. Specifically, an amplify and forwarding (AF) two-way relay is optimally selected to maximize the sum rate and to keep the interference on the primary user (PU) below an interference threshold. We first obtain asymptotic signal-to-interference-plus-noise ratio (SINR) values for two scenarios: (1) the relays and the two end nodes use transmit power scaling and (2) only the end nodes use transmit power scaling. For these two cases, we derive optimal power allocations subject to the PU interference constraints. With these optimal power allocations, we analyze how relay selection impacts the outage, the sum rate, and the energy efficiency of the network. For the first scenario, the outage can be reduced to zero with appropriate power allocation and the relay selection can be done offline. For the second scenario, outage will depend on the instantaneous channel state between the relays and the PU. Furthermore, for a realistic power consumption model, we analyze the energy efficiency and show that relay selection will increase it.