Interference cancellation using multiple antenna techniques for cognitive radio network

Abstract

The explosive growth in wireless communication services and fixed spectrum allocation policies by government agencies have led to spectrum scarcity. Besides, recent research has indicated that the actual occupancy of most licensed frequency bands is quite low leading to underutilization and waste of valuable frequency resources. Cognitive radio has emerged as a promising solution which enables the unlicensed (secondary) user to establish a communication link in licensed band under the condition that there is no or minimal interference to the licensed user. This dissertation addresses the challenge of overcoming interference effect due to coexistence of primary network and secondary network by developing several novel strategies, ranging from adaptive antenna array until the precoding technique of singular value decomposition with iterative algorithm. Two networks that communicate with their counterpart by sharing their spectrum are considered, known as spectrum sharing. In underlay spectrum sharing, the inconsistency of channel condition and simultaneous transmission by primary and secondary users occupying the same frequency band resulted interference effect still remains an issue. Joint interference cancellation and equalization for Space-Time Block Coded Orthogonal Frequency Division Multiplexing (STBC-OFDM) transmission is presented. To mitigate the interference from primary user, pre- and post-FFT adaptive antenna array is employed at secondary user receiver. Then, the optimum antenna weights that place nulls at the primary transmitters are determined. The optimum weight determination for beamformers is based on the minimum mean square error (MMSE) criterion. Simulation results demonstrate that post-FFT is optimum in terms of maximizing signal-to-noise-and-interference power ratio (SINR) and provides a lower bit error rate (BER) under existence of several primary users. Two strategies of jointly-optimized uncoordinated beamforming algorithms for cognitive radio networks are addressed in this dissertation. The optimum weights are designed to maximize the achievable rate and achievable sum rate for both primary and secondary links under the condition that the cross interference at both receivers are totally nullified. Simulation results have shown that employing gradient algorithm at the cognitive link is optimum in terms of maximizing achievable rate and provides a higher sum rate performance as compared to the discrete search. An improved strategy for jointly-optimized uncoordinated beamforming over cognitive radio network is also proposed. The proposed design is based on singular value decomposition and iterative water-filling at the primary link while employing discrete search and gradient algorithm at the secondary link. The interference cancellation is performed at the secondary user and no coordination between the primary and secondary users is required as uncoordinated beamforming is being employed in the network. The new optimum weights are designed to maximize the achievable rates for both primary and secondary links and their achievable sum rate performance of the network under the condition that the cross interference at both receivers are totally nullified. Simulation results show that the proposed design achieves more than two times increases in performance of achievable rates of both primary and secondary links and total sum rate of the network as compared to the previous reported work.