In radar systems, target detection is required for situational awareness, and constant false alarm rate (CFAR) detectors are mainly used to detect targets. In recent radar systems, multiple target situations such as swarm drones, vehicle radars, ground target detection/tracking radars, sea target detection/tracking radars, and aerial target detection/tracking radars occur frequently, and accordingly, the importance of target detection performance in multiple target situations is increasing. Since the detection performance degrades when using conventional CFAR detectors in multiple target situations, many studies are being conducted to improve the detection performance in multiple target situations. In addition, it is necessary to reduce the computational complexity to implement CFAR detectors with improved detection performance in multiple target situations in real-time radar systems.
In this dissertation, CFAR detectors that reduce the computational complexity and improve the detection performance in multiple target situations were studied, and in particular, CFAR detectors with a new structure were proposed using the average value of the selected samples and the ordered statistics of sub-reference cells to improve the detection performance in multiple target situations. The first method, which uses the average value of selected samples, requires no prior information, has low computational complexity, and shows high detection performance in multiple target situations. This dissertation presents the concept and the structure of the method, compares detection performance in multiple target situations using computer simulations, and presents detection performance comparison results using flight test data. The second method is a method using the ordered statistics of sub-reference cells, which does not require prior information, has low computational complexity, and shows high detection performance in multiple target situations. This dissertation presents the concept and the structure of the method, compares detection performance in multiple target situations using computer simulations, and presents detection performance comparison results using flight test data. Comparing and verifying the proposed CFAR detectors and the conventional CFAR detectors using the computer simulations and the flight test data, the proposed CFAR detectors have low computational complexity, do not require prior information, and have improved detection performance in multiple target situations. Based on the results presented in this dissertation, the proposed CFAR detectors are expected to apply to real-time processing radar systems and exhibit high detection performance in multiple target situations.