Multiuser Detection of Synchronous MC-CDMA in Multipath Fading Channels Using Hopfield Neural Networks

摘要

Sub-optimum multiuser reception using Hopfield Neural Network for synchronous Multicarrier Code-Division Multiple Access signals in a multipath fading channel is studied with respect to near-far ratio. We have shown that by the appropriate choice of Hopfield Neural Network parameters from the channel state information, the Hopfield network can collectively resolve the multipath fading effects and the multiple-access interference in the system. Moreover, the Hopfield Neural Network demonstrates multiple-access interference resilient performance regardless of the number of paths resolved at the receiver. We have also investigated the bit-error rate performance of the system with respect to channel estimation errors. Results show that performance of the proposed detection scheme is influenced by the correctness of the estimated channel state information.