Efficient MIMO Detection Based on Eigenspace Search with Complexity Analysis

A low-complexity, effective detection method for multiple-input multiple-output (MIMO) systems based on eigenspace search and linear equalization-based detection schemes is proposed in this paper. Based on the observation that solutions yielded by linear detectors are corrupted by color noise, the proposed method introduces a new constellation search procedure to augment linear detectors. Specifically, calibrated search is conducted around the initial solution yielded by linear detectors, in the directions guided by the eigenvectors corresponding to the dominant eigenvalues of the covariance matrix of the color noise to identify improved solutions. Complexity analysis is performed to understand the cost of this search procedure. Simulation results demonstrate that the proposed scheme yields an approximate 5 dB gain over linear equalization-based detectors in terms of symbol error rate (SER), at moderate additional computational cost.