Skip to Main Content
An optimum receiver for multiple-input multiple output orthogonal frequency-division multiplexing (MIMO-OFDM) communication systems accounting for realistic channel estimation is proposed. The receiver is assumed to comply with the emerging IEEE 802.11n standard and its performance is compared against that of a genie receiver (corresponding to ideal channel estimation) and a mismatched receiver (using estimated channel state information in the ideal channel metric). Receiver complexity is addressed in two steps: first, by developing an iterative expression of the receiver metric and second by implementing a spectral approximation, which allows a dramatic reduction of the receiver complexity with unnoticeable degradation. Since the optimum receiver is based on the availability of channel distribution information, it is shown by numerical results that its estimation has a marginal effect on the error performance and does not represent an issue for the receiver implementation.