DL for IoT security incident detection

Traditional security solutions (such as encryption, authentication, access control, and network security) are ineffective for IoT devices. In recent years, DL/ML-based solutions have become very popular alternatives to traditional solutions. DL/ML-based solutions can monitor IoT devices and their networks intelligently and detect various new or zero-day attacks. Importantly, DL/ML can detect and/or predict various devices and network level security incidents through anomaly detection. By gathering, processing, and analyzing data about various normal and abnormal activities of devices/things and their networks, these DL/ML methods can identify various security incidents, including IoT device and network level intrusions. In the following sections, we briefly present a few DL models that are useful in IoT device and network level IDS.

...