A new suboptimal multiuser detector based on orthogonal polynomial approximation

We propose a new suboptimal detector that uses orthogonal polynomial approximation to determine the stage weighting factors at each interference cancellation (IC) stage. The performance of the new detector using the proposed scheme is compared with that of other sub-optimal detectors in the near-far and additive white Gaussian noise (AWGN) environments. Simulation results show that the average convergence rate of the proposed method for obtaining the optimum solution through estimation of the minimum (and maximum) eigenvalue of the code correlation matrix is faster than for other sub-optimal detectors based on IC. In addition, the method described here is also able to obtain better BER performance, with the same number of stages, than conventional parallel interference cancellation in the above environments.