Invertibility of Higher-Order Moment Matrices

Robert D. Nowak and Barry D. Van Veen
Department of Electrical and Computer Engineering
University of Wisconsin-Madison, WI 53706 USA

Abstract --

Signal processing algorithms based on Volterra filter structures often require
that a matrix composed of higher-order moments of the underlying process be 
invertible.  Previously, this problem has been studied for uncorrelated random
variables.  This paper establishes conditions under which a wide class of 
correlated processes have invertible higher-order moment matrices.