Statistical Signal Processing and Detector Optimization in Project 8

Despite the unambiguous discovery of non-zero neutrino masses from flavor oscillation experiments, a direct measurement of the absolute mass scale of the neutrino remains elusive to experimentalists. Project 8 is a tritium endpoint experiment utilizing Cyclotron Radiation Emission Spectroscopy (CRES), a novel, high-precision spectroscopic technique, in order to establish the absolute neutrino mass scale. In this document, I investigate the statistically motivated limits to CRES signal detection and parameter estimation, as well as the resultant consequences on optimal detector configuration. I implement and test an application of the Viterbi algorithm for CRES signal reconstruction, yielding the first derived limits on the minimal detection criteria. I then present an original derivation of the Cramér-Rao Lower Bound of the start frequency resolution for realistic CRES signals, along with estimators yielding near-optimal performance. Finally, these improved detection and reconstruction algorithms lead into a discussion of optimal detector design.Ph.D