Two-sensor tracking of maneuvering transmitters
Hanna Witzgall · 2018
This paper introduces a technique that can be used to track a maneuvering transmitter using the time-difference-of-arrival (TDOA) and frequency-difference-of arrival (FDOA) radio frequency (RF) measurements collected from two receiving platforms equipped with omnidirectional antennas. The challenge this paper addresses stems from the underdetermined nature of the uncooperative tracking problem when using only TDOA and FDOA measurements. Specifically, the number of state parameters that must be estimated (e.g. position and velocity at every measurement cycle) outnumbers the RF observables (e.g. TDOA, FDOA) for the case of a constantly maneuvering (i.e. constantly varying velocity) target. This results in an underdetermined set of equations where there are many “false” state solutions whose measurement models correspond to the observables. These false solutions prevent a conventional least squares solver from finding the correct answer and motivate a new solution. The proposed solution, which is called TWo Sensor Tracking (TWiST), uses a new optimization criterion that enables it to track with fair accuracy a maneuvering target even when the number of observables is less than the number of unknown state variables. The results of this technique will be compared to a conventional least squares solution that assumes a constant velocity transmitter for both the constant velocity and the maneuvering target scenarios. It will be shown that when this constant velocity assumption is violated, the new approach results in substantially better tracking performance.