Multi-user detection for DS-CDMA communications
IEEE Communications Magazine
Fourth-generation mobile initiatives and technologies [Guest Editorial]
IEEE Communications Magazine
Incremental backpropagation learning networks
IEEE Transactions on Neural Networks
| Hi-index | 0.00 |
Multiuser detection paradigm gives efficient techniques for separation of users in CDMA system specially, when the system has a high MAI due to the non-orthogonality of the used codes. Many techniques are presented for multiuser detection but most of them are static, i.e. assuming time-invariant channel which is not the practical case. Some of the presented techniques are dynamic but their transfer function is linear which doesn't give the optimum or near optimum BER as required. This paper presents a dynamic multiuser detection technique based upon backpropagation neural networks using sigmoid transfer function. Simulation is performed to compare the proposed technique with other neural networks techniques and other static techniques. The simulation results show that the proposal causes BER enhancement over the considered techniques.