A TIME-DELAY ESTIMATOR BASED ON THE SIGNAL INTEGRAL - THEORETICAL PERFORMANCE AND TESTING ON ECG SIGNALS

We present a theoretical and experimental performance study of a metho d for time delay estimation (TDE), based on the signal integral (TDE-S I). The TDE-SI method considers the delay between two transient signal s as the difference between the center of mass of these signals. The m ethod has three special cases: In the first, time is the mass coordina te and the signal s(t) is the mass distribution (estimate ($) over cap D-s,); in the second case, the squared signal s(2)(t) is the mass dis tribution (estimate ($) over cap D-s2); and the last is a variant of t he second. The bias and the standard deviation (a) of the estimate hav e been evaluated when the signal is contaminated by Gaussian white noi se. ($) over cap D-s is not biased but the a of the TDE is higher than that obtained when working with ($) over cap D-s2. Moreover, the ($) over cap D-s2 estimate is biased. The special case of a bias-corrected estimate (($) over cap D-s2') is presented; this ($) over cap D-s2', yields a sigma of its TDE lower than the estimate ($) over cap D-s. He nce, ($) over cap D-s2' is the most suitable of the three TDE-SI optio ns for TDE. Theoretical estimations are validated by simulation result s with artificially generated signals and by real so-called QRS comple x waves (ventricular activity) from an electrocardiographic (ECG) sign al.