A study of the ambiguity problem in footstep bearing estimation using tri-axial geophone

This paper investigates the ambiguities that exist in the bearing estimate of human footstep signals acquired using a single tri-axial geophone. Estimated angles for a series of footsteps from three different walking patterns are evaluated. It is observed that there are two kinds of ambiguities, viz., energy and wave polarization. A systematic study of these ambiguities are presented using covariance matrix analysis (CMA) as the base algorithm. Existing techniques to resolve these ambiguities are analysed with their limitations.