MMSE-based iterative processing with imperfect channel and parity check in MIMO systems

It is known that the acquisition of the complete channel state information at receivers is difficult in multiple-input multiple-output (MIMO) systems. Channel estimation error is unavoidable in practical applications. Under imperfect channel conditions, the channel estimate is directly applied to the equalisation process in the conventional minimum mean-square error (MMSE)-based turbo equalisation scheme. A few studies treat the channel estimation error as an independent component from the channel estimate and slightly enhanced performance is achieved. Unlike the existing work, the authors derive the MMSE-based iterative processing conditioned on channel estimate. Moreover, they note that in low-density parity check coded systems, the parity-check procedure is also involved. The pass in parity check indicates that the message bitstream is successfully recovered. This information can be utilised to reduce the overall computational complexity by degrading the MIMO size since the unknown parameters are reduced. By extending the analysis in a small-scale MIMO system to a large-scale one, they propose to utilise the normalised transmission power in the development. Numerical results show the proposed schemes outperform the existing schemes in terms of system bit error rate and computational complexity performance.