RCS-Aided Tracking: Does it Always Improve Data Association?

Single-frame measurement-to-track association algorithms that make unique decisions often have difficulty with closely-spaced (but resolved) targets. With little upon which to base decisions, it is not surprising that mistakes are made. One approach proposed in the literature for mitigating this problem is to augment the kinematic measurement-to-track likelihoods with amplitude information. However, this can actually be detrimental to the measurement-to-track association process for radar tracking. This paper analytically computes the probability of error for RCS-aided measurement-to-track association for Rayleigh targets and shows numerical results for fixed-amplitude and Rician targets. Since Rayleigh targets are typical in many radar applications, the last section addresses the number of pulses from a Rayleigh target that must be integrated to obtain reliable measurement-to-track data association with RCS.