ABSTRACT
This chapter examines the problems associated with digital processing and estimation of the time delay between signals received at two spatially separated sensors together with noise. The complete derivation is carried out for discrete signals using the maximum likelihood approach. Then the necessary conditions are obtained for the equivalence to occur between the continuous and discrete theory for a generalized correlator. The chapter analyzes the estimator of time delay for discrete signals. A maximum likelihood (ML) estimator is derived for a purely discrete system with discrete delay parameter. Then the delay is allowed to be a continuous variable with nonperiodic signals and the effect on ML estimations is denoted. The chapter presents and discusses some simulation results in light of the preceding analysis. It tabulates experimental means and variances for various signal-to-noise (S/N) ratio and noise bandwidth conditions. The signal spectrum is considered smooth to avoid ambiguity or bias in time delay estimates by most of the windows.
