An Experimentally-Based Method For Detection Threshold Determination in HFSWR’s High-Resolution Range-Doppler Map Target Detection

Primary signal processing in High Frequency Surface Wave Radars (HFSWRs) plays an important role because it affects their overall performance. The two most important parameters in high-resolution primary signal processing, which is based on the MUSIC (MUltiple SIgnal Classification) algorithm, are the boundary between the signal subspace and the noise subspace in the eigenvalue decomposition process, and detection threshold used for target detection on range-Doppler (RD) map. With most of the existing radars, detection is generally performed in the RD domain, and then the azimuth is estimated only for the detected targets. Therefore, target detectability is primarily determined by target detectability performance in the RD domain. This paper shows that the determination of the detection threshold based on experimental results is possible by forming Radar Operating Characteristics (ROC) for individual ships, in order to see how the radar actually works in practice, but also to estimate the optimal value of the detection threshold, so that the probability of target detection is satisfactory.