Approximate Bayesian tracking of two targets that maneuver in and out formation flight

The paper evaluates four recently developed advanced target tracking algorithms on their performance in maintaining tracks of two targets that maneuver in and out formation flight, whereas the sensor and measurement extraction chain produces false and possibly unresolved or missing measurements. The baseline algorithm uses a joint IMM for the joint maneuver modes of the two targets, in combination with an enhanced version of JPDA that takes the coupling between target state estimates into account. The other three algorithms have been developed as further enhancements over this baseline algorithm. The first incorporates a track coalescence avoidance approach, the second takes unresolved measurements into account, and the third incorporates both enhancements. The effectiveness of the four advanced filters is evaluated through Monte Carlo simulations