Radar OSA Practice Test

What is a ROC curve and how do you interpret Pd vs Pfa?

It shows probability of detection vs false alarm; higher and left is better; area under the curve indicates overall detector performance.

A ROC curve shows how likely you are to detect a target (Pd) as you allow more false alarms (Pfa) by lowering or raising the detector threshold. It captures the trade-off between sensitivity and false alarms.

Interpreting it, you want higher Pd for the same Pfa, which means more reliable detection without paying in extra false alarms. A curve that sits higher and more to the left is better: for a given Pd you achieve a smaller Pfa, and for a given Pfa you achieve a larger Pd. The area under the curve (AUC) summarizes overall performance—the closer the AUC is to 1, the better the detector; an AUC around 0.5 suggests no useful discrimination.

So when comparing detectors, a ROC curve that lies above and to the left of another (or has a larger AUC) indicates superior detection performance for all thresholds.

It shows range vs velocity; steeper curve is better.

It shows noise variance vs SNR; bigger area means worse performance.

It shows processing gain vs PRF; no interpretation.