Performance Analysis Framework of ML MIMO Receiver over Correlated Rayleigh Fading Channel

In this paper, a novel performance analysis frame-work of Maximum Likelihood (ML) receiver for Vertical Bell Labs Layered Space-Time (V-BLAST) architecture over correlated Rayleigh fading channel is presented. Based on the statistical property of squared Euclidean distance, the exact expressions for Pairwise Error Probability (PEP) over transmit, receive, and joint correlated fading channel are derived. The approximate expression of the PEPs for high signal-to-noise ration (SNR) region is used to quantitatively analyze the loss of the diversity order and the penalty of SNR. Three interesting results of this analysis are: 1) the loss caused by transmit and receive correlation is independent. 2) the transmit correlation has no impact on the diversity order, and the loss of SNR caused by transmit correlation is determined by the transmit correlation matrix and modulation scheme. 3) the loss of diversity order due to receive correlation is equals to the difference of the numbers of receiver antenna and the rank of receive correlation matrix, under the assumption that the receive correlation matrix has full rank, the penalty of SNR is jointly determined by the determinant of the receive correlation matrix and the number of receive antenna. Simulation results are given to corroborate the theoretical analysis.