Moving source localization using doppler frequency shift measurements in sensor network

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] This thesis studies and develops a solution for locating a moving source using Doppler frequency shift measurements in sensor network. The study inquires the problems of inaccurate sensor positions, unknown source frequency, and inaccurate source frequency, on the quality of the location estimate. Iterative solutions that need initial guesses have been considered since the frequency measurements are nonlinearly related to the source trajectories (position and velocity). In this thesis we present three solutions to localize the source, iterative maximum likelihood estimator (IMLE), grid search estimator (GSE), and closed form estimator (CFE). The IMLE achieved the Cramer Rao lower bound (CRLB) performance but it needs initial guess sufficiently close to the actual source trajectories. The GSE searches over only the source position and this makes the processing time much less than normal full dimensional GSE. The proposed CFE does not need initial guess and has very small processing time and it can achieve the CRLB performance under low level of measurements noise.