Kalman Filter Based Tracker Study For Muon-Electron Conversion Experiment

Abstract

The search for muon to electron conversion with a sensitivity of the order 10^-17 requires a several order of magnitude increase in muon intensity and a high resolution, sigma = 0.1 MeV/c of the electron's momentum. We present results of a pattern recognition and track momentum reconstruction algorithm that relies on a Kalman filter approach. Background from captured protons, neutrons, photons and muon decay in orbit were generated by GEANT. The effective average straw tube background rate was 800 kHz. The pattern recognition proceeds in two stages. In the first, simple considerations using only straw tube center coordinates, without drift time information, were applied to reduce the background to a manageable level. Then the drift time information is incorporated and a Deterministic Annealing Filter applied to reach the final level of background suppression and to provide a starting point for the track momentum reconstruction using the Kalman filter. This procedure reduces the simulated background by a factor 800 with small, (2.7%), losses in real tracker hits. The momentum resolution of the tracker is sigma = 0.12 MeV/c and the acceptance for muon conversion events with momentum above 103.6 MeV/c is 22%. These numbers do not differ significantly from the values obtained without background. The expected number of events from muon decay in orbit (main background) in which the decay electron has momentum greater than 100 MeV/c is 0.3, compared to 6.5 muon-electron conversion events above the same threshold for R_mu_e =10^-16.