Local tracking in the ATLAS muon spectrometer

The LHC, the largest hadron collider accelerator ever built, presents new challenges for scientists and engineers. With the anticipated luminosity of the LHC, it is expected to have as many as one billion total collisions per second, of which at most 10 to 100 per second might be of potential scientific interest. One of the two major, general-purpose experiments at LHC is called ATLAS. Since muons are one of the important signs of new physics, the need of their detection has lead to the construction of a stand- alone Muon Spectrometer. This system is located in a high radiation background environment (mostly neutrons and photons) which makes the muon tracking a very challenging task. The Muon Spectrometer consists of two types of precision chambers, the Monitor Drift Tube (MDT) chambers, and the Cathode Strip Chambers (CSC). In order to detect the muon and estimate its track parameters, it is very important to detect and precisely estimate its local tracks within the CSC and MDT chambers. Using advanced signal processing tools, this dissertation is focused on two main objectives: a. To design and implement better algorithms for muon detection and track parameter estimation within the CSC and MDT chambers. b. To evaluate the performance of muon detection and parameter estimation algorithms. In developing novel local muon track finding algorithms, we exploit the inherent high efficiency of both the MDT and the CSC detectors, and use the detect-before-estimate approach, which first detects muon tracks and then estimates the track parameters of the found tracks. The detection algorithms are based on novel modifications of the Hough transform, which for the MDT is shown to be an approximation of the Generalized Likelihood Ratio Test (GLRT). In order to precisely estimate the muon momentum, the track finding is followed by a track fitting operation. While the use of the traditional least squared (LS) is sufficient for accurate track fitting in the MDT, a more sophisticated algorithm is required for good performance in noisy environment of the CSC. We present novel generic algorithms for iii muon hit position estimation and track fitting in Multi Wired Proportional Chambers (MWPC). The proposed approach is based on prior knowledge of the hit-cluster charge distribution of a muon. First, the muon hit position is estimated using the LS or the Expectation-Maximization (EM) techniques. Then, each cluster is identified as a "clean" or a "dirty" one, and used in a novel iterative weighted least squares algorithm, which is shown to have significantly better performance than the current used algorithm in the present of high radiation background. The performances of the suggested algorithms are theoretically and experimentally evaluated using simulations and real data. They are shown to be superior to any existing algorithms in a high radiation background.