TY - JOUR TI - Antennas re-clustering and target handoff for multiple radars system DO - https://doi.org/doi:10.7282/T3KH0QSJ PY - 2017 AB - Widely distributed multiple radar systems have been shown to offer enhanced localization performance. With smaller radar footprint, the ability to employ larger number of transmit and receive antennas opens new opportunities. In previous research, a subset selection scheme has been proposed for antenna clustering that minimizes the number of transmit and receive antennas required to achieve a preset accuracy performance. The study indicated that some transmit and receive antenna pairs contribute more than others to the localization performance. This thesis concentrates on handoff techniques that enable the transition of target tracking from one antenna cluster to another. As a target moves in an area covered by a grid of multiple radars, its relative position with respect to an existing tracking antenna cluster (or antenna subset) is changing, affecting the accuracy capabilities of the existing antenna cluster. Thus, at some point, there is a need to update the antenna cluster, keeping a useful antenna subset while replacing other antennas with ones that will keep localization accuracy within a given range. Re-clustering methods are proposed to address target handoff within antennas belonging to a larger grid. Low complexity re-clustering algorithms are proposed for handoff purposes which enable a constrained replacement of antennas. These fast approximation algorithms are based on the optimization of the Cramer Rao bound (CRB) and constrained by the number of antennas that may be replaced at any given time. It is shown that this method performs close to optimal and can be implemented in a decentralized fashion. KW - Electrical and Computer Engineering KW - Radar--Antennas KW - Antenna arrays LA - eng ER -