Guided Local Search in High Performance Detectors for MIMO Systems

Abstract—This work analyzes efficient non-spreading (a)synchronous MIMO detection topologies under realistic channels which results in high throughput and good perfor-mance complexity trade-off. In this sense, we look for near-optimum efficient MIMO detections suitable for (un)coding schemes. Main system and channel parameters are analyzed, such as increasing number for transmitter and receiver an-tennas, number of iterations for convergence under AWGN and Rayleigh fading channels. Two heuristic local search MIMO detectors are compared with other near-optimum de-tectors, specifically SDR (semidefinite relaxation), expectation-maximization (EM), and linear multiuser detectors. Besides, the MIMO detectors performances under large MIMO systems (high number of transmitter and/or receiver antennas) are analyzed. The performance complexity tradeoff results have indicated promising features for the guided local search (GLS) procedures in high capacity MIMO detectors. Keywords-MIMO system, heuristic detectors, semidefinite relaxation. I