Partial maximum likelihood receiver with instantaneous SNR-based subspace search for multistream MIMO

In this paper we propose a novel PML (partial maximum likelihood) receiver with instantaneous SNR-based subspace search for multistream MIMO (multi-input multi-output) systems. The soft-decision outputs of the conventional ZF receiver are used as the initial estimates for PML. In contrast to the OSIC (ordered successive interference cancellation) making only one hard decision based on soft-decision estimates, the proposed PML receiver set up a subspace for ML search with multiple candidate constellation points. The number of candidates in the subspace search is based on the instantaneous SNR at each data stream. Finally, the hard-decision estimates can be obtained by minimizing the ML function over the subspace search. Numerical results proved that when the number of transmit antennas is less than 4, the proposed PML considerably outperforms the conventional OSIC with approximately the same or even less complexity.