Performance improvement of QO-STBC over time-selective channel for wireless network

Abstract

Signal detection for vehicle-to-vehicle and wireless network application has motivated new design for conditions in which channel is time-selective fading. Since the physical layer (PHY) forms the foundation of the communication protocol stack, the performance of this layer affects all high layers of the system stack. Quasi-orthogonal space-time block code (QO-STBC) can provide full rate transmission and partial diversity under low decoding complexity. Previous papers on QO-STBC assume that the channels are slow fading or remain static over the length of the codeword. However, time-selective channels do exist, and in this case, the decoder proposed in Jafarkhani (2001) cannot be used to achieve a proper error performance. In order to mitigate the severe performance degradation, the zero forcing (ZF) and minimum mean square error (MMSE) decoders are employed in this paper. We also propose a zero forcing interference cancellation decision-feedback equalizer (ZF-IC-DFE) and a minimum mean-square error interference cancellation decision-feedback equalizer (MMSE-IC-DFE) via Cholesky factorization of the channel Gram matrix after performing interference cancellation. By feeding back past decisions on previously detected symbols, DFE schemes can achieve an additional performance gain compared to their corresponding linear decoders.