Blind identification of MIMO channels based on correlative coding: performance analysis

We derive the asymptotic performance analysis of a recently introduced analytical method for blind identification of multiple-input multiple-output (MIMO) convolutive systems excited by inputs with distinct second-order spectra; the spectral diversity in the second-order moments is induced by correlative filters located at the sources. We establish the asymptotic normality of the MIMO channel estimator and obtain a closed-form expression for the mean square error (MSE) of the channel estimates. The theoretical study permits us to investigate the impact of certain key system parameters on the resulting MSE and to obtain guidelines for the optimum design of the correlative filters. Computer simulations are carried out to validate the theoretical expressions.