The ambiguity problem arising in multisensor data association

In many applications, (for example: air traffic control; robotics; ocean surveillance and medical diagnostics) one needs to process and analyse large volumes of data. The multisensor data association problem is concerned with determining which measurements from one or more sensors actually relate to the same object. This problem can be formulated as a multidimensional assignment problem. In the application of bearingonly data association, the solution of this NP-hard problem gives a list of targets and their position. Unfortunately, this solution though mathematically correct may not be physically valid as false targets known as ghosts may occur. This situation occurs whenever the number of targets is greater than or equal to the number of sensors. When ghosts are present, the ambiguity problem can arise. The ambiguity problem is the focus of this paper and we establish bounds on various sensor-target configurations. Australasian Journal of Combinatorics 27(2003), pp.107–127