Selective Detection for MIMO-OFDM Transmission

We present a novel technique for the detection of spatially multiplexed transmitted symbols in uncoded multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system. The method, termed selective detection, observes errors caused by a simple zero- forcing (ZF) on different OFDM sub-carriers, and selectively performs list sphere detection (LSD) on sub-carriers with larger errors. Compared with the full-complexity LSD, the proposed algorithms could significantly reduce the receiver complexity at a small signal-to-noise ratio (SNR) loss. For example, simulation results for a MIMO-OFDM system using four transmitters and four receivers with 16 quadrature amplitude modulation (16-QAM) show that 93% of SNR gain that can be achieved by the use of LSD over ZF can be realised if only 33% of sub-carriers are intelligently selected for LSD.