This paper presents a new multi-target data association method for automoti
ve radar which we call the order statistics joint probabilistic data associ
ation (OSJPDA). The method is formulated using the association probabilitie
s of the joint probabilistic data association (JPDA) filter and an optimal
target-to-measurement data association is accomplished using the decision l
ogic algorithm. Simulation results for heavily cluttered conditions show th
at the tracking performance of the OSJPDA filter is better than that of the
JPDA filter in terms of tracking accuracy by about 18%.