A Low-Complexity Upgrade of the Linear Detector for MIMO Channels via Partial Decision Feedback

The BLAST-ordered decision-feedback (BODF) detector is a nonlinear detection strategy for multiple-input multiple-output (MIMO) channels that can significantly outperform a linear detector. However, for some applications even the BODF detector is too complex. We propose the partial decision-feedback (PDF) detector, a stripped-down version of the BODF detector that only feeds back one decision. The PDF detector performs close to the BODF detector, with complexity comparable to the linear detector. For example, over a 3-input 3-output Rayleigh-fading channel with 64-QAM inputs, the PDF detector is 21% less complex than the BODF detector, yet requires only 0.3 dB more average signal energy to reach a bit-error rate of 10^-3