Complexity reduction of RBF multiuser detector for DS-CDMA using a genetic algorithm.

The optimal receiver for detecting direct sequence code division multiple access (DS-CDMA) signals suffers from computational complexity that increases exponentially with the number of users. Several suboptimal multiuser detectors (MUDs) have been proposed to overcome this problem. Due to the nonlinear nature of the decision boundary of the optimal receiver, it is known that nonlinear receivers outperform linear receivers. Radial basis function (RBF) MUD is a nonlinear suboptimal receiver that can perfectly approximate this decision boundary and it needs no training since it is fully determined when the spreading codes of all users and the channel impulse response (CIR) are known. However, the RBF MUD suffers from structural complexity since the number of hidden nodes (center functions) in its structure increases exponentially with the number of users. In this study, we propose a new method to minimize the number of center functions of the RBF MUD using a genetic algorithm (GA) and the least mean squares (LMS) algorithm. With simulations performed in AWGN and multipath channels it is shown that the proposed method immensely reduces the complexity of the RBF MUD with a negligible performance degradation. [ABSTRACT FROM AUTHOR]