Current cybersecurity solutions have been inadequate for emerging threats, as recent developments have seen hardware exploits introduced in an untrusted supply chain. Detecting hardware tampering can be an extremely challenging task for traditional cyber security solutions. Power Fingerprinting (PFP) is a novel approach for integrity assessment of critical embedded systems which can detect malicious intrusions at all levels of the execution stack, including hardware, firmware, and software. PFP leverages analog side channel signals, e.g. electromagnetic emissions/power consumption, and machine learning to establish baselines of normal patterns and detect malicious. The core PFP technology has been demonstrated on different platforms, including network infrastructure, control systems, and flight computers.In this project we proposed to demonstrate the technical and market feasibility of using PFP technology to characterize normal digital system operation and identify of anomalous behavior that might indicate hardware or software compromise. Current PFP solutions include analytics tools to establish baselines from normal operation and detect anomalies, as well as monitoring appliances to collect the side-channel signals. These solutions have been commercialized and deployed in different scenarios and pilots.Power Fingerprinting,intrusion detection,side channel analysis,machine learning,supply chain
