Adaptive blind deconvolution of MIMO channels using higher-order statistics

This paper is concerned with the problem of adaptive deconvolution and estimation of the matrix impulse response function of a multiple-input multiple-output system given only the measurements of the vector output of the system. The system is assumed to be driven by a spatially and temporally i.i.d. non-Gaussian vector sequence (which is not observed). An iterative, inverse filter criteria based approach was previously developed using the third-order and/or fourth-order normalised cumulants of the inverse filtered data at zero-lag. The approach was input-iterative, i.e. the inputs were extracted and removed one-by-one. The matrix impulse response was then obtained by cross-correlating the extracted inputs with the observed outputs. In this paper an adaptive implementation of the above approach is developed using a stochastic gradient approach. Simulation examples are presented to illustrate the proposed approach.