Feature-Based Angle-Only Measurement to Radar Track Association

Abstract

Angle-only tracking of airborne targets is challenging due to the ambiguity caused by the absence of a measure for range. If a colocated radar and angle-only sensor work collaboratively, it is possible to associate measurements and tracks using not only kinematics, but also radio frequency (RF) derived features. Combining kinematic distances and RF features in a probabilistic framework enables the stitching together of disparate measures in a coherent manner. In this article, we quantitatively formulate a probabilistic RF feature-based kinematic association technique which improves performance over kinematic association alone. We demonstrate a significant increase in accuracy using this RF feature-based kinematic association technique in challenging radar tracking scenarios with closely spaced targets.