A Novel Data Association Algorithm For Ghost Elimination In Passive Radar Systems

In this paper, a novel data association algorithm is developed for detecting and localizing multiple targets. The fusion of the measurements involving angle-of-arrival (AoA) and time-of-arrival (ToA) generated by the passive sensors is accomplished effectively. The ghost problem faced during this fusion is solved by clustering these measurements and assigning scores to each of them. Score assignment is performed using AoA values and hyperbola intersections generated by ToA values. In addition, entropy is used for eliminating ghost clusters more efficiently. Then, clusters which have the highest scores are used to estimate target positions by applying maximum likelihood estimation. This algorithm is tested with different number of targets and different noise levels.