A Maximum-Likelihood Decoder With a New Reduction Strategy for MIMO Channel Systems

An efficient maximum-likelihood decoder with a new reduction strategy is proposed for linear MIMO channel systems. Unlike the current reduction strategies which only reorder the columns of the channel matrix, the new reduction algorithm employs the so called integer Gauss transformations to reduce the off-diagonal entries of the upper triangular factor of the QR decomposition of the channel matrix. Simulation results show that this new decoding algorithm can be much more efficient than existing algorithms.