Robust interference channel transmission using sparsity enhanced mismatch model

Herein a maximin-SNR robust precoder design is proposed for MIMO cognitive radio transmission when inaccurate channel estimate is present. The proposed scheme exploits the structural property of the statistical transmit covariance matrix to increase the degree of freedom in the mismatch model related to the channel matrix. Simulation results show that the proposed scheme outperforms previously proposed methods in terms of SER under certain conditions. The performance gain is obtained at the cost of increased interference level toward the primary user, but without violating the average interference power constraint. Analytical and simulation results are given to support the efficacy of the proposed scheme. Furthermore, analysis is carried out to differentiate the performance given by the lower bound method and S-Procedure method.