Target Track Estimation using Measurements Acquired with a Small Number of Sensors
J. Arnold, Yaakov Bar‐Shalom, Richard Estrada, Ronald A. Mucci · 1983
An analysis procedure based on the information contained in sets of measurements has been developed and tested for evaluation of target localization with a pair of sensors. Initially, this tool was used to determine if the information contained in a set of measurements is sufficient to identify the localization parameters such as position, course, and speed. However, the analysis also provides important performance measures such as algorithm convergence indicators and position and velocity parameter uncertainty ellipses. Examination of performance for various test cases provided important technical insight relating localization accuracy to target trajectory relative to sensor positions, measurement type, and measurement quality. The results demonstrated [6] that reliable localization based on time difference of arrival (TDOA) measurements occurs primarily for targets passing between the sensors. Also, the TDOA measurements were observed to produce an elongated position uncertainty ellipse whereby an accurate estimate of bearing and an inaccurate estimate of range result. On the other hand, reliable localizations were observed for frequency measurements obtained in the vicinity of closest point of approach (CPA). The corresponding uncertainty ellipses are not elongated although the uncertainty in bearing generally appears to be greater than the range uncertainty for frequency measurement sets. The simultaneous use of both types of measurements, i..e., frequency and TDOA, was observed to produce an uncertainty ellipse contained by the overlapping regions of the individual measurement ellipses, which demonstrates the improvement in performance afforded by complementary measurement sets.