Identification of Time Varying Delays in Stochastic System via Combined Non-Linear Estimation and Viterbi Algorithm

This paper introduces a class of estimators for the identification of time varying delays in linear discrete time stochastic systems. The proposed new technique which is suboptimum in maximum likelihood is realized by parallel nonlinear filtering coupled with a Viterbi Algorithm (VA). This class of finite memory nonlinear state estimators (NLSE) provide the components needed by the VA for the delay sequence identification which approaches maximum likelihood with increase in the memory length.