Human-to-Device Authentication Framework
A Biometric Locking Paradigm for IoT and the Connected Person
Paper and Patent Under Review.
Overview: As the internet-of-things (IoT) era begins, there is a significant need for low-cost access control schemes that allow humans (i.e., connected people) to activate and maintain IoT services and systems. Traditional biometric access control systems remain, however, vulnerable to physical attacks resulting in template theft/privacy, illegal access, etc. This work aims at protecting a system (i.e., device) by locking its functionality with a key produced by combining human and hardware signature. The system can be unlocked again to its original functionality only if the correct key is provided.
Fig. Impact of proposed technique on a RGB-GRAY image conversion system (a) Input to the system locked by proposed locking mechanism, (b) output from the locked system when correct unlocking key is provided, and (c, d, e) output from the system when wrong keys are applied.
The proposed system can be very useful in securing the functionality of a system. In the figure above, a grayscale conversion system is considered as a system to be protected. The proposed locking technique has been applied on the system to lock its functionality to protect it. For an RGB image (Fig. a) as the input to the locked system, when the correct key is provided, the system can successfully convert it to the grayscale version (Fig. b). However, when multiple wrong keys are provided, the conversion gets distorted as shown in (Fig. c, d, e).