A performance study of MIMO detectors in the presence of channel estimation errors

In this paper, we investigate the performance of the low-complexity algorithms based on Likelihood Ascent Search (LAS) and Tabu Search (TS) for detection in MIMO systems with channel estimation errors. Here, we briefly review the algorithms for LAS and TS and their variants and subsequently compare them for various MIMO settings. We provide the simulation results for BER performance in 2× 2, 4× 4 and 8× 8 MIMO systems. From the results, we conclude that Tabu Search shows no degradation in BER up to an estimation error variance of 0.05 in lower order systems, while LAS offers the same advantage in higher order systems. However, this trend is also valid for LAS in lower order systems, in the low SNR regime. We corroborate that LAS and TS are worthy candidates for further research.