A reduced-complexity lattice-aided decision-feedback detector

Two powerful techniques for improving the performance of a detector in a multiple-input multiple-output communications channel are decision feedback and lattice reduction. We propose a new detector, the DOLLAR detector, that combines decision feedback and lattice reduction in a novel way. The DOLLAR detector is based on a simple lattice reduction algorithm that is designed to exploit the power of decision feedback. The combination of the new lattice reduction algorithm and decision feedback leads to near-optimal performance with complexity comparable to the BLAST-ordered decision-feedback (BODF) detector. For example, over a 4-input 4-output Rayleigh-fading channel with 16-QAM inputs, the DOLLAR detector outperforms the BODF detector by 6 dB while increasing complexity by 16%.