Time Delay Estimation

Publisher Summary The estimation of time delay or time difference has become an important problem in digital signal processing, for example, an ideal active radar or sonar that employs a single omnidirectional sensor to transmit and receive signals can measure the time difference between the time a signal was transmitted and the time a backscattered signal was received to estimate the range of a radar or sonar target. On the other hand, an ideal passive sonar or radar generally employs an array of omnidirectional sensors for the sole purpose of receiving acoustic or electromagnetic radiation from distant targets. When the radiation is received at the passive array, time delay estimation methods are used to estimate the travel time of an acoustic or electromagnetic wavefront between the sensors. Wavefront travel time gives the range and direction of a radiating target. In either case, the estimation of these time delays is often corrupted by ambient and receiver-generated noise and multipath and finite-length observation intervals. This chapter provides an overview of the time delay estimation problem. It discusses the active and passive radar and sonar problems, beamforming, and active and passive localization. It also discusses the theory and algorithms involved in estimating the intersensor time delays. In real-world applications, there are no stationary point target models and simple backscattering. Besides, there are multipath, finite-length observations, and nonstationary, non-Gaussian noise. The chapter highlights the basic principles and some current real-world problems, namely, the fundamental limitations of the passive theory when narrowband energy is present and the problem of tracking a time-variant time delay.