Implications of threshold effects for coherent and incoherent processing techniques of time delay estimation

In recent work, the Ziv-Zakai Lower Bound (ZZLB) has been shown to yield a tighter bound, and hence, a more accurate estimate of performance than the Cramer-Rao Lower Bound (CRLB), for time delay estimators. Above a threshold signal-to-noise ratio (SNR), the ZZLB converges to the CRLB; below the threshold, the ZZLB predicts a significantly larger variance for the time delay estimate than the CRLB. The impact of this threshold effect on signal processing techniques for time delay estimation is considered in this work. It is shown that, at low SNR, coherent processing yields significant performance gains over incoherent processing. Simulation results are presented to corroborate these performance predictions. When relative motion between source and sensors is present, tne received signals must be pre-processed to compensate for the relative time compression due to motion, unless short coherent processing times are used. One such compensation technique is described, and preliminary simulation results are presented to demonstrate the effectiveness of this technique.