Estimation of arbitrary time delays of multichannel synthetic data

A new method for computing arbitrary time delays between multichannel data is introduced based on the idea of using the ordinary cross‐correlation technique. It is shown that, by windowing the data record, events can be isolated and identified between channels. This new technique, called the window‐correlation technique (WCT), is capable of determining arbitrary delays, very accurately, down to a signal‐to‐noise ratio of 0 dB. Finally, results of the WCT are compared to those of the ordinary cross correlation for Gaussian signals corrupted by additive Gaussian white noise.