Radar detection in interfering target environment : algorithm design and real-time implementation

Abstract

The main contribution of this thesis involves two parts. The first one concerns the proposition of two novel algorithms for radar target detection referred to as, respectively, FAOSOSD, Forward Automatic Order Selection Ordered Statistics Detector, based on the minimization of the information theoretic criteria, and ALC-CFAR, Adaptive Linear Combined CFAR, based on an adaptive linear combination of the CA-CFAR and the OS-CFAR detectors thresholds. These algorithms present the ability to sense automatically the environment changes, especially in presence of interfering targets, by adapting their thresholds to ensure better performance compared to classical detectors. The second part deals with the proposition of new and efficient architectures suitable for the implementation of the proposed detectors on the TMS320C6711 DSP processor. The real time processing constraints and the radar signal digitalization effect on the detection quality have been evaluated and discussed. The obtained results show that the developed detectors are suitable for practical applications in radar detection.