Abstract

This thesis introduces an analytical method for large-system design and analysis of multistage linear multiuser (MLMU) receivers for direct-sequence (DS-) code division multiple access (CDMA) systems. The figure of merit to evaluate performance is the signal to interference plus noise ratio (SINR) that is calculated for a case where the number of active users and the processing gain tend towards infinity, while their ratio remains finite. The large-system performance will be a function of the number of the interference cancellation stages, the moments of the eigenvalues of the covariance matrix, the system load and the SNR. Compared to other recently-proposed methods, our proposed method offers a more efficient approach to finding the large-system performance by introducing a new expression for the covariance matrix of the interference term. A simplified multistage scheme to estimate the multipath channel gains is also introduced. It will be shown that by applying this new method for data detection and channel estimation, the performance of the MLMU receiver will converge to that of the minimum mean-squared error (MMSE) receiver.