Blind Separation of Ground Reaction Force Signals

Blind source separation presents an interest for several applications fields. In this study, focus is on the blind separation of biomechanical signals namely ground reaction force signals. We showed recently in [1, 2], that ground reaction force signals are cyclostationary. Thus, the idea of this paper is to make use of the cyclostationary character of ground reaction force signals to separate it and then compare the separation results with those based on stationary statistics. To this end, we introduce frequency domain approaches for either blind source separation or MIMO system identification excited by cyclostationary inputs. These approaches exploit the cyclic spectra matrices of the whitened measurements to identify the mixing system at each frequency bin, up to constant diagonal, frequency dependent permutation and phase ambiguity matrices. Two efficient algorithms to fix the permutation problem and to remove the phase ambiguity based on cyclostationarity are also presented. The approaches exploit the fact that the input signals are cyclostationary with the same cyclic frequency. Simulation examples are presented to illustrate the performances of these approaches. Furthermore, the effectiveness of the proposed methods are tested over real signals and compared with existing methods, as well.