Abstract

The work begins with a study of several most frequently referred to sub-optimal multiuser receivers. The four structures of interest are the decorrelator, the minimum mean square error receiver, the multistage parallel interference cancellation receiver, and the successive interference cancellation receiver. The performance of the receiver structures is compared through simulation studies on the basis of bit error rate in additive white Gaussian noise, and near-far channels. In the second part of this research, array processing techniques are investigated as a means for improving the performance. We evaluate the system performance of some of the existing adaptive blind array schemes for CDMA receivers, and a new adaptive blind algorithm is derived. The knowledge of the desired user's signature sequence is exploited to design the blind receiver, where no training sequence needs to be transmitted. The blind algorithm is formulated using a constrained energy minimization criterion, and adaptation is carried out by using the LMS algorithm combined with the gradient-projection algorithm to incorporate the quadratic constraint. The multiple access interference coming from the directions other than that of the desired user can be significantly suppressed by using spatial filtering property of antenna arrays. The interference rejection capability of the proposed receiver is examined in situations where a single interferer dominates the received power. It is shown that this combination can significantly improve the system performance and outperform all existing multiuser detectors as well as smart antenna systems. Finally, we present a new structure combining two-dimensional RAKE receiver and the multiuser detection for further enhancement of the system performance in multipath fading channels.