Identification of single events in the HPGe detector: Comparison of various methods based on the analysis of simulated pulse shapes

TL;DR

This paper compares three methods—library matching, single-parameter, and neural networks—for identifying single events in HPGe detectors using simulated pulse shapes, aiming to improve detection in neutrino double beta decay experiments.

Contribution

It evaluates and compares the effectiveness of three different pulse shape analysis methods for single event identification in HPGe detectors.

Findings

Neural network method shows promising separation capabilities.

Library matching provides high accuracy with known pulse shapes.

Single-parameter method is simpler but less effective than others.

Abstract

Various methods of identification of single events in the HPGe detector are compared on the basis of a program especially designed to simulate pulse shape in a semi-conductor germanium detector. Capabilities of three following methods are shown: (1) an application of the library of single pulse shapes, (2) the single-parameter method, and (3) a separation on the basis of artificial neural networks. The analysis was done in the context of an application of all above-mentioned techniques in the energy interval around the expected neutrino less double beta decay of Ge-76 in the Heidelberg-Moscow experiment.