hardware chip
hardware chip

PPREDATOR

Duration
December 2023 until December 2029
Project type
Research

Prediction of security vulnerabilities in hardware designs is of critical importance for manufacturers, since once fabricated, a hardware implementation is impossible to patch. Furthermore, vulnerabilities in hardware implementations (including side-channel leaks and fault injection vulnerabilities) can be used to effectively bypass security mechanisms and put chips or systems at major risk. 

Challenges in hardware security

Ensuring the security of modern hardware designs is challenging due to their complexity, aggressive time-to-market demands, and the variety of attacks introduced against hardware designs. Predicting and eliminating side-channel leaks requires a dedicated team with a broad range of expertise, such as signal processing, statistics, and cryptography, as well as keeping up with the development of a very dynamic research area - adding to the costs of secure hardware development. Existing commercial electronic design automation tools for hardware design can optimise implementations with respect to power consumption, silicon area and operation speed, but have no built-in support for physical security. 

Project aim

The aim of this project is to provide the semiconductor industry and the security evaluation industry with a set of metrics and a framework to aid analysis and mitigation of hardware security vulnerabilities.

Results

This project has two tracks: predicting vulnerabilities against side-channel analysis and fault injection. 

Related to the former we introduced ISALeak, a target-agnostic framework for identifying and diagnosing leakage in fully masked implementations. Our next step is to address the leakage simulation problem for hardware implementations by using an OpenTitan board, which is the industry's first open-source project for a silicon Root of Trust. 

With respect to mitigations to fault injection we are considering AI-assisted Fault Detection Mechanisms as a collaboration with Ikerlan, focusing on glitching FI attacks. Voltage fault injection attacks are a potent threat to embedded devices as they exploit brief, hard-to-detect power fluctuations causing errors or bypassing security mechanisms. We plan to leverage AI-based detectors to identify glitches by observing the clock signals that are manipulated by voltage glitching mechanisms.

Funding

NWO TTW

Partners

Contact information

More information or questions? Please get in touch with Lejla Batina