Blind adaptive multiuser detection using a recurrent neural network

Multiuser detection has gained much attention in recent years for its potential to improve greatly the capacities of CDMA communication systems. A recurrent neural network is presented for solving the nonlinear optimization problem involved in multiuser detection in CDMA. Compared with other neural networks, the presented neural network can converge globally to the exact optimal solution of the nonlinear optimization problem with nonlinear constraints and has relatively low structural complexity. Computer simulation results are presented to show the optimization capability. The performance in CDMA communication systems is also studied by means of simulation.