Optimal Filtering in the ATLAS Hadronic Tile Calorimeter

In this paper we present an Optimal Filtering algorithm to reconstruct the energy, time and pedestal of a photomultiplier signal from its digital samples. The optimal filter method is applied in multiply-sampled signals and improves the energy reconstruction by minimizing the impact on resolution of both electronic noise, intrinsic to the calorimeter, and the pile-up which depends on the luminosity. The OF implementation and performance have been studied for the hadronic Tile Calorimeter (TileCal) of the ATLAS detector. The algorithm is tested and compared with other reconstruction algorithms from two different types of data: calibration runs coming from the TileCal Charge Injection System and physics events - pions, electrons and muons - acquired during the testbeam program of the TileCal detector. The results are promising specially in the regions where the electronics noise contributes significantly to the resolution.