Optimal and sub-optimal iterative cross-layer energy efficient schemes for CR MIMO systems with antenna selection

In this paper, energy efficient transmit antenna selection in underlay cognitive radio (CR) multiple-input multiple-output (MIMO) systems from a cross layer perspective, is investigated. Conventional energy efficient antenna selection optimizes an energy efficiency (EE) metric defined in terms of the system capacity and energy consumption, making it a physical layer criterion. In this work, we propose an EE metric that is defined in terms of the system throughput and energy consumption. Throughput considers characteristics from the data link and physical layers, making the proposed EE metric a cross layer quantity. The subset of transmit antennas that maximizes the proposed EE criterion, subject to interference constraints at the primary user receiver (PU RX), is selected. Additionally, the proposed algorithm optimizes the EE by adapting the number of active transmit antennas and transmit power, under a spectral efficiency constraint. A sub-optimal reduced complexity iterative algorithm is also developed. Simulation results show that the cross layer approach results in overall improved system performance, measured in terms of throughput, packet error rate and transmit power, as compared to conventional energy efficient antenna selection.