Frequency domain blind MIMO system identification based on second- and higher order statistics

We present a novel frequency-domain framework for the identification of a m ultiple-input multiple-output (MIMO) system driven by white, mutually indep endent, unobservable inputs. The system frequency response is obtained base d on singular value decomposition (SVD) of a matrix constructed based on th e power-spectrum and slices of polyspectra of the system output. By appropr iately selecting the polyspectra slices, we can create a set of such matric es, each of which could independently yield the solution, or they could all be combined in a joint diagonalization scheme to yield a solution with imp roved statistical performance. The freedom to select the polyspectra slices allows us to bypass the frequency-dependent permutation ambiguity that is usually associated with frequency domain SVD, while at the same time allows us compute and cancel the phase ambiguity, An asymptotic consistency analy sis of the system magnitude response estimate is performed.