Base line estimation using sparse nonlinear operators

Abstract

Efficient nonlinear operators for base line estimation of high energy physics detector signals are introduced. Due to the high processing rate requirements the operators are simplified using sparse input sequences. The operators are a sparse order statistic filter, a sparse L-filter and a sparse Kronecker L l-filter. Each filter is optimized using a total cost function minimization algorithm based on the conjugate gradient search. The performance of different operators is compared using simulated detector signal sequences.