This project demonstrates and compares three Constant False Alarm Rate (CFAR) radar detection algorithms using synthetic 1D range data:
- LO-CFAR (Locally Optimal Rank Detector)
- Standard CFAR (mean-based)
- OS-CFAR (median-based)
- Simulates radar signal embedded in Gaussian noise
- Sliding window detection across 128-sample signal
- Adjustable noise standard deviation (σ)
- Displays and compares:
- Raw signal and injected targets
- Detection statistics for each method
- Thresholds for detection
Below is an example output showing the detection curves for noise standard deviation σ = 1.0:
- Python 3.x
- numpy
- matplotlib
- scipy
Simply run the script to generate and display detection results for 3 different noise levels (low, medium, high). Each detection method is shown with appropriate thresholds and visual comparison.