Low Complexity Signal Detection in MIMO Systems

The Maximum Likelihood Detector (MLD) offers an optimal bit-error-ratio for an un-coded multiple-input-multiple-output (MIMO) communication system. The computational complexity, however, of MLD increases significantly for detection of higher order modulated signal streams with multiple transmit antennas. To reduce this complexity while simultaneously also achieving near optimal performance, we propose a novel technique that modifies the sub-optimal Zero Forcing Detector (ZFD) by using the minimum eigen vector of the matrix $\mathbf{HH}^{H}$ , where H is the channel matrix and $\mathbf{H}^{H}$ is its Hermitian. The performance in terms of BER and computational complexity of the proposed technique is compared to that of the MLD. Simuations show that the BER performance of the proposed technique is near optimal whereas the computational complexity is signiflcantly reduced.