MSE-Based Transceiver Design for Decentralized Transmit Power Minimization of Multi-Cell MIMO Cognitive Radio Systems

Abstract

This paper considers multiple-input multiple-output (MIMO) cognitive radio (CR) networks in which multiple primary and secondary systems coexist. We propose a transceiver design algorithm to minimize the total transmit power of secondary systems subject to the signal to interference plus noise ratio (SINR) constraint for each secondary user and the interference constraint for each primary user. Due to the complexity of backhaul among secondary systems, perfect knowledge about the channel state information (CSI) may not be available in practice. To overcome this problem, we decompose the transmit power minimization problem into cascade problems of decentralized beamformer design with the interference constraint and power allocation with the SINR constraint. For beamformer design, we formulate the total mean squared error of data symbol minimization problem, to enhance the feasible rate and only use the local CSI via dual decomposition. The power allocation problem is established as linear programming. Simulation results show that, compared to the conventional centralized algorithm, the proposed decentralized algorithm achieves much higher feasible rate with slightly increased transmit power.