Machine-learning based anomaly and intrusion detection for cyber-resilience in microcontrollers

In this presentation, we present an overview of a proposed cybersecurity architecture and software stack, for microcontrollers, incorporating intrusion detection and cyber-resilience to enable critical systems to detect and respond to cyber-attacks. The cyber-resilience security architecture’s is a host-based IDS capable of rapidly detecting a broad range of attacks in real-time with high detection rates and low false positives. This requires new transformative IDS technologies. We present an overview of an IDS approach using IoT-optimized machine learning algorithms to classify execution of software tasks/processes/applications as normal or anomalous in real-time.