Large-Scale Multi-Cluster MIMO Approach for Cognitive Radio Sensor Networks

Abstract

This paper proposes a large-scale cooperative multiple-input multiple-output (CMIMO) beamforming scheme for uplink (UL) access in broadband cognitive radio wireless sensor networks (CR-WSNs) sharing the same spectrum with a primary network and employing orthogonal frequency-division multiplexing. The CR-WSN is divided into clusters each consisting of cooperative nodes that form a virtual antenna array. Using particle swarm optimization (PSO), each cluster seeks the optimal transmit weight vectors that maximize the UL channel capacity of each cluster, while controlling the interference levels to the primary network. Under the assumption of very large number of sensor nodes at each cluster, semi-analytic expressions for the symbol error rate and the ergodic channel capacity of the CMIMO-based CR-WSN are derived and validated with Monte-Carlo simulation. The PSO-based capacity-aware (PSO-CA) scheme is compared with the one based on the traditional gradient search scheme (GS-CA) and the results show that PSO-CA requires considerably less computational complexity while achieving essentially the same level of performance as the GS-CA.