Speaker
Description
We developed a software trigger utilizing the Pearson linear correlation coefficient or so-called $r$ which quantifies the resemblance between a slice of continuous data stream and the signal template. The $r$ trigger method was found to outperform conventional approaches using the pulse amplitude in terms of lowering the threshold in a time-varying noise environment. Combining with a simple optimal filter and applying to the AMoRE-I continuous data, we achieved an energy threshold at lower than 5 keV with a tolerable false trigger rate. Although it has been developed for a cryogenic detector experiment, the method can be widely utilized for the signal recognition provided that sufficient computing power is available to handle the data sampling rate. The detailed method of the $r$ trigger, its performance, and considerations for the practical applications are presented.