Design and optimization of a high resolution radar data processing system

Abstract

An implementation of a complete High-Resolution Radar Data Processing System is performed, including Clustering, Gating, measurement-to-track association and
track filtering algorithms. For the last mentioned, various schemes are implemented, such as a classic α-β Filter and a debiasing compensation Kalman Filter. A useful Detection Matrix Simulation tool is developed to test the overall system performance. In addition, the system is tested using real radar data. Subsequently, a method to initialize Kalman Filter in Cartesian Coordinates, using a α-β Filter in Polar Coordinates is proposed, proving to be useful when there are targets in the Radar FOV with initial acceleration or that have a high measurement error due to their distance to the Radar. Finally, based on results obtained in tests, parameter adjustment is perform to optimize the system performance.