Labeled Random Finite Sets-Based Group Target Tracking with Multi-Detection Model

In radar tracking applications for low-altitude group targets (such as bird flocks or drone swarms), the traditional point target measurement model is often unsuitable due to the complex micro-Doppler signatures and multipath effects. Algorithms based on labeled random finite sets (RFS) challenges with multi-detection problems. This paper presents a new method for tracking low-altitude group targets using labeled RFS in radar systems. To mitigate the disturbance from multi-detection, we propose modified assignment cost matrices in the update process, which reduces the number of false associations, and thus achieves quality enhancement of trajectory estimations. The effectiveness of the proposed method is validated through simulation results.