An Adaptive Algorithm for Blind MMSE Equalization of SIMO Channels

In this work, an adaptive method for estimating blind MMSE equalizer with arbitrary delay of SIMO channel is proposed. By using an RLS-similar algorithm to recursively update correlation matrix and its pseudo inverse, the proposed algorithm ensures convergence and is not sensitive to initialization. Compared with other batch-type approaches, the new adaptive method doesn´t require directly computing pseudo inverse or singular value decomposition (SVD), consequently doesn´t need rank estimation nor channel order estimation. Furthermore, the property of fast convergence enables our new method to be used in real time communication applications. Simulation examples are provided to demonstrate the performance of the new method.