Optimum Filter Synthesis with DPLMS Method for Energy Reconstruction

TL;DR

This paper introduces two extensions to the DPLMS method for synthesizing optimum filters, enhancing energy reconstruction accuracy in single and multi-channel systems through faster computation and better noise correlation handling.

Contribution

The paper presents novel extensions to the DPLMS method that improve synthesis speed, filter quality, and multi-channel parameter estimation capabilities.

Findings

DPLMS filters outperform traditional filters in energy estimation.

Multi-channel DPLMS filters leverage inter-channel noise correlations.

Extensions significantly improve synthesis efficiency and estimation accuracy.

Abstract

Optimum filters are granted increasing recognition as valuable tools for parametric estimation in many scientific and technical fields. The DPLMS method, introduced some twenty years ago, is effective among the synthesis algorithms since it derives the optimum filters directly from the experimental signal and noise waveforms. Two new extensions of the DPLMS method are here presented. The first one speeds up the synthesis phase and improves the energy estimation by synthesizing optimum filters with automatically designed flat-top length. The second one improves the quality of parameter estimation in multi-channel systems by taking advantage of the inter-channel noise correlation properties. The theoretical and functional aspects behind the DPLMS method for optimum filter synthesis are first recalled and illustrated in more detail. The two new DPLMS extensions are subsequently introduced from the theoretical viewpoint and more thoroughly considered from the applicative perspective. The DPLMS optimum filters have been applied first to simulated signals with various amounts and characteristics of superimposed noise and then to the experimental waveforms acquired from a solid-state Ge detector. The results obtained are considered from both the absolute viewpoint and in comparison with those of more traditional, suboptimal filters. The results demonstrate the effectiveness of the two new DPLMS extensions. For single-channel energy estimations, the optimum filters provide comparatively better results than the other filters. The DPLMS multi-channel optimum filters further enhance the quality of the estimations, compared to single-channel optimum filters, with non-negligible inter-channel noise correlation. The effectiveness and robustness of the DPLMS method in synthesizing high-quality filters for energy estimation will be tested soon within leading-edge multi-channel physics experiments.