An Adaptive Receiver for Stbc in Frequency Selective Channel with Improved Robustness and Pilot Requirements

Abstract: The semi-blind recursive least squares (RLS) based adaptive receiver has been designed to perform joint interference suppression and equalization for space time block codes (STBC) in a frequency and time selective channel. In order to decrease pilot requirements in the training block, this paper introduces a linear predictor (LP) algorithm to do a forward prediction of the channel coefficients based on a smaller pilot block. We then introduce a QR-decomposition (QRD) based algorithm to improve the performance of the receiver at higher Doppler frequencies. The simulation results show that the addition of the LP does not affect the frame error rate (FER) of the overall system. Linear prediction requires a smaller number of pilot symbols in order to provide the channel estimates for a given burst. Hence the LP increases the overall throughput of the system by decreasing the pilot symbols required. The simulation results show that this is a more effective method for improving the system performance as the overall FER of the combined QRD-LP receiver is decreased significantly. Finally, by comparing the QRD based receiver and the RLS based receiver at higher Doppler frequencies, we verify that the QRD receiver has superior FER performance under these conditions.