An adaptive WPE multiuser detector

Weighted polynomial expansion (WPE) detectors can be used to approximate linear multiuser detectors to avoid the need to compute matrix inversion. For finite stage WPE detectors, weight vector optimization is an important process since it determines how effective the linear multiuser detectors can be approximated. In this letter, the optimal weight vector for the finite stage WPE detector is derived based on the mean-squared error (MSE) criteria. In addition, a low-complexity adaptive WPE detector based on the least-squares criteria is proposed to simplify the computation of the optimum weight vector. Simulation results show that the proposed adaptive WPE detector outperforms the existing WPE method.