Blind Minimum Mean Squared Error (MMSE) Multiuser Detection with Reliable Subspace Estimation (RSE)

In this paper, we propose a new blind minimum mean squared error (MMSE) receiver for the downlink CDMA system. The improved receiver is based on the principle of splitting the signal space into three subspaces associated with the desired signal, the interfering signals and the white noise. The interference is suppressed by projecting the received signal onto the subspace orthogonal to the interference subspace. The receiver estimates the interference subspace from selected sample by subtracting the reliable estimate of desired signal. Reliability is determined by the log likelihood ratio (LLR) of the estimate given by the blind MMSE receiver. To estimate the interference subspace more reliably, we only use the received samples for which the desired signal can be estimated with low probability of error. We construct a blind-MMSE and matched filter (MF) receiver based on this concept and both of them are compared against conventional blind multiuser detectors and are shown to perform better. The new subspace splitting receivers for efficient small block size for used the transmitted data, making them suitable for applications which require low time delays.