Ensuring security is a challenging task in the physical world. Mathematically robust cryptographic algorithms are just the first step towards a secure system. In practice, cryptographic implementations may leak the secret key through hardware defaults such as variation in power consumption, electromagnetic radiation, or timing, or faults in computation. These are collectively known as Side-Channels. Attacks based on side-channels are especially crucial for automotive and cyber-physical systems (CPS) with several potential attack surfaces. Attacks on the cryptographic components, which are essentially the root of trust, may leave the entire software stack vulnerable. The goal of this project is to investigate, and provably mitigate side-channels for implementations of cryptographic algorithms.
Another security issue regarding the hardware layer of CPS and automotive is the threat from malicious and stealthy hardware modifications commonly known as Hardware Trojan Horses (HTHs). Potential sources of HTHs are third-party or open-source hardware intellectual property cores, and untrusted fabrication facilities. The exploits of HTHs can be diverse -- from denial-of-service to the leakage of the secret. On the other hand, detection of HTHs can be notoriously difficult, mainly due to its small size and inherently stealthy nature, which bypasses the standard hardware testing methodologies. In this project, we aim to develop and utilize test methodologies for HTHs.
(a)
(b)
(c)
(d)
(e)
Instruments: (a) Laser-based fault injection; (b) Electromagnetic and power side-channel; (c) ESP Anweshak tool for side-channel and fault attacks; (d) FIBS-SEM station; (e) Temperature Chamber.