Low-Complexity Turbo Block Decision Feedback Equalization for MIMO Systems

This paper proposes a low-complexity turbo block decision feedback equalizer (BDFE) with reliability-based successive soft interference cancellation (SSIC) for multiple-input multiple-output (MIMO) systems. Turbo equalization in MIMO systems needs to combat both intersymbol interference (ISI) due to frequency-selective fading and multiplexing interference (MI) among data streams transmitted by different antennas. The proposed MIMO BDFE performs SSIC on both ISI in the time-domain and MI in the space-domain. For an SSIC receiver, the order in which the symbols are detected and cancelled is critical to the system performance, yet it is often overlooked in the literature. We propose a new reliability-based detection ordering scheme, where symbols with higher a priori reliability are detected before those with lower a priori reliability. The reliability information can be obtained from the a priori probability at the input of the soft-decision equalizer. The reliability-based ordering scheme is enabled by the turbo equalization structure, and it yields an extra equalization gain that is unavailable in a conventional equalization system. Furthermore, the adoption of non-linear BDFE in MIMO systems enables a low-complexity sequence-based log-likelihood ratio (LLR) evaluation, which is superior to the symbol-based LLR evaluation method employed by most other low-complexity turbo equalizers.