Capacity Evaluation of MIMO-OFDM Systems using Reduced-Complexity ML Detectors in a Spatially Correlated Channel

This paper evaluates reduced-complexity maximum-likelihood (ML) detectors with soft-decision outputs in a multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) system. The reduced-complexity ML detectors, e.g., a sphere decoder (SD) or QR decomposition (QRD)-M algorithm, are very promising candidates as a MIMO detector because they can estimate the ML or quasi-ML symbol with very low computational complexity. These detectors, however, have a difficulty in producing soft-decision outputs because they may lose likelihood required for calculating log-likelihood ratio (LLR), which is related to reliability of detector outputs. We have proposed a simple likelihood estimation method which estimates the lost likelihood by means of a combination of the ML estimation and spatial-filtering method. This paper evaluates the reduced-complexity ML detectors with the proposed method in a spatially correlated MIMO channel to prove its availability in a realistic environment. Computer simulation confirms that the proposed method provides excellent decoding performance and achieves very high system capacity.