Representation and analysis of complex chirp signals

Abstract

In this paper, we present a difference equation model with time-varying coefficients for complex-valued multicomponent chirp signals. This model may be applied to real-valued signals as a special case. We show the relation between chirp parameters, i.e., instantaneous frequency (IF) and amplitude (IA), and difference equation coefficients. We present a recursive method for the estimation of the signal poles from which IF and IA are obtained. We show that the estimates may be improved by proper choice of the initial values. We present examples to illustrate the performance of the proposed algorithms.