Multiuser detection in DS/CDMA systems using neural networks

We analyze and examine the performance of multilayer perceptron neural networks with a backpropagation learning rule as multiuser detectors of DS/CDMA signals in synchronous AWGN and fading channels. The results show a significant reduction in the receiver complexity in comparison with previous studies while achieving the same performance. We compare the neural network performance with the other detectors. We also apply a cross entropy criterion in the neural network and compare its performance with the backpropagation method.