Investigation of Sensor Bias and Signal Quality on Target Tracking with Multiple Radars

The tracking of airborne targets from low-cost Internet of Things (IoT) sensors, such as radars, is a problem of increasing interest due to the proliferation of drones. Such low-cost IoT sensors have problems with sensor bias and poor signal quality which may impact detection and tracking performance. In this paper, we investigate the impact of sensor imperfections on tracking performance. In particular, we consider the scenario of two ground-based sensors measuring the elevation/bearing/range of three airborne targets in clutter. The measurement uncertainty of the second sensor was altered between test cases to emulate sensor bias, then the results of the multi-target tracking algorithm were compared using the Generalized Optimal Sub-Pattern Assignment, Single Integrated Air Picture, and uncertainty metrics. Results indicate that bias in the range measurement tends to decrease the track’s robustness against clutter. A sensor with equally poor performance in the elevation/bearing/range measurements scored the lowest in all investigated metrics. The clutter density parameter of the investigated multi-target tracking problem, the Joint Probabilistic Data Association filter, was altered and found to have nearly negligible effect on the track quality.