Anomaly Data Detection for ADS- B Based on Zero-bias Inception Network

Automatic Dependent Surveillance-Broadcast (ADS-B) technology is a new type of air traffic surveillance technology. With the wide applications of ADS-B technology, its security has attracted great attention. This paper proposes a zero-bias inception model, which uses an improved inception-v3 model combined with a zero-bias layer for feature extraction and anomaly detection. Compared with the traditional anomaly detection algorithms, this method reduces computational complexity, enhances robustness and interoperability, and improves the accuracy of recognition.