Base Line Normalization of High Energy Physics Detector Signals Using Sparse Median Operations

Abstract

Efficient median-based operations for normalization of base line fluctuations in high energy physics detector signals are introduced. High processing rates are required in real-time implementations of digital signal processing systems for future high energy physics detectors. By shortening the median operator using sparse input and sparse recursive input the number of comparisons needed to obtain the operator output is greatly reduced. The sparse recursive median operator is comparable in base line normalization performance to the standard median (SM) operator, which is shown using threshold decomposition and stack filter representation. The different algorithms are applied to simulated detector signal sequences and the performances are compared using a mean squared error (MSE) based criterion.