Date of Award

Summer 6-5-2026

Document Type

Thesis (Undergraduate)

Department

Computer Science

First Advisor

Sergey Bratus

Second Advisor

Rahul Sarpeshkar

Abstract

Glitching, or Fault Injection, is an effective hardware hacking technique used in industry but largely unstudied in academia. Industry-led study of glitching is result-driven and overwhelmingly proprietary, meaning that a scientific approach to glitching is seen infrequently and published even less.

This thesis aims to be part of the effort to bring glitching to the attention of the academic side of the cyber security community. Specifically, this work aims to characterize and explore the probabilistic nature of fault injection that has been previously overlooked. Although from a purely result-driven point of view probabilistic nature of a phenomenon suggests unreliability and is viewed as a drawback, this research aims to examine and better formalize the unpredictability present in glitching setups.

Firstly, a more formalized approach to glitching, as well as an overview of existing work, is provided. Secondly, the potential of glitching research is demonstrated through data collection and analysis of multiple experiments.

Download

Share

COinS