Self-Diagnosis of Radar System State in RSU Applications
Chia-Hsing Yang, Ming‐Chun Lee, Ta-Sung Lee · 2021 IEEE 94th Vehicular Technology Conference (VTC2021-Fall) · 2021
To realize the intelligent transportation, environmental awareness of roadside units (RSUs) is of paramount importance. One of the approaches to enable the environmental awareness of RSUs is to equip RSUs with radar systems. However, as more and more radar systems are installed, manually monitoring whether these radar systems work in their normal states becomes impossible. To resolve this issue, a radar system state self-diagnosis method is proposed in this paper by using the radar sensing information with deep learning techniques. Specifically, by using the proposed feature extraction approach, we first effectively convert the huge amount of radar sensing data into useful features. Then, by using the proposed deep neural network to interpret the extracted features, the radar systems can self-diagnose whether there exist faults on the systems. We verify our proposed method via real-world experiments. Results show that our proposed method can accurately diagnose the radar system and report the faults.