Phase I project will be the investigation and specification of the necessary algorithms and platform for detection and mitigation of spoofing attacksand compromised sensing in sensor networks. The solution addresses the three main areas of sensing security: (a) detection and mitigation with corresponding customization; (b) support for collaborative, distributed detection across Ndes; and (c) hardware platform with standard interfacing options and utilization as a standalone remote system or as an add-on at-the-edge system. This platform is envisioned to contain several machine learning computational blocks focused on detection and mitigation of attacks that attempt to block sensor readings or inject incorrect data into the data stream. The success of Phase I will be based on the availability of an architectural design, as well as, a proof-of-concept demonstration, including validating anti-spoofing ML solution for several sensing types including GPS.
