Coherent noise source identification in multi channel analysis

TL;DR

This paper presents a new covariance matrix-based method for identifying coherent noise sources in multi-channel detector data, offering an alternative to PCA, with validation through simulations for CALICE calorimeters and broader applicability.

Contribution

A novel covariance matrix analysis method for coherent noise source identification, specifically developed for CALICE calorimeters, and validated via simulations for general multi-channel applications.

Findings

Effective in identifying noise sources in CALICE calorimeters

Validated through simulations demonstrating reliability

Applicable to any multi-channel analysis

Abstract

The evaluation of coherent noise can provide useful information in the study of detectors. The identification of coherent noise sources is also relevant for uncertainty calculations in analyse where several channels are combined. The study of the covariance matrix give information about coherent noises. Since covariance matrix of high dimension data could be difficult to analyse, the development of analysis tools is needed. Principal Component Analysis (PCA) is a powerful tool for such analysis. It has been shown that we can use PCA to find coherent noises in ATLAS calorimeter or the CALICE Si-W electromagnetic calorimeter physics prototype. However, if several coherent noise sources are combined, the interpretation of the PCA may become complicated. In this paper, we present another method based on the study of the covariance matrix to identify noise sources. This method has been developed for the study of front end ASICs dedicated to CALICE calorimeters. These calorimeters are designed and studied for experiments at the ILC. We also study the reliability of the method with simulations. Although this method has been developped for a specific application, it can be used for any multi channel analysis.