Low-Complexity Near-Optimal Signal Detection in Underdetermined Large-MIMO Systems
The authors presented a tabu search based detection algorithm which achieved improved performance in underdetermined MIMO systems by exploiting multiple random restarts and a threshold based stopping criterion. Unlike the generalized sphere decoder, the proposed algorithm scaled well for large number of antennas. In addition, it exhibits near-optimal performance with large number of antennas; e.g., near-ML BER performance was shown for 16?12, 16?8, 32?24, 32?16 UDMIMO with 4-QAM/16-QAM. A proposed low-complexity ML lower bound aided the assessment of the nearness of the proposed algorithm performance to ML performance.