DBNN, FDNN, Discriminative Learning, and Back-Propagation Neural Networks in DS/CDMA Systems

Conventional single user detector in DS/CDMA (Direct Sequence Code Division Multiple Access) systems involves multiple access interference and near-far effect, which cause the limitation in capacity. The complexity of optimum multiuser detector also grows up exponentially with the number of users. There has been a lot of interest in sub-optimal multiuser detectors with less complexity and reasonable performance. In this paper, we apply decision based neural network (DBNN), fuzzy decision neural network (FDNN), and discriminative learning and back-propagation neural networks employing multilayer perceptron for detection of signals of users in DS/CDMA systems in additive white Gaussian noise (AWGN) channel. We also show that FDNN and discriminative learning neural net have almost the same performance.