Biblio

This paper focuses on the design and development of attack models on the sensory channels and an Intrusion Detection system (IDS) to protect the system from these types of attacks. The encoding/decoding formulas are defined to inject a bit of data into the sensory channel. In addition, a signal sampling technique is utilized for feature extraction. Further, an IDS framework is proposed to reside on the devices that are connected to the sensory channels to actively monitor the signals for anomaly detection. The results obtained based on our experiments have shown that the one-class SVM paired with Fourier transformation was able to detect new or Zero-day attacks.