Publication Date

8-2025

Date of Final Oral Examination (Defense)

5-8-2025

Type of Culminating Activity

Dissertation

Degree Title

Doctor of Philosophy in Computing

Department

Computer Science

Supervisory Committee Chair

Nasir Eisty, Ph.D.

Supervisory Committee Member

Amit Jain, Ph.D.

Supervisory Committee Member

Edoardo Serra, Ph.D.

Abstract

Data breaches remain a critical threat in high-security sectors such as finance, healthcare, and government, where protecting sensitive data is paramount. This research investigates and addresses key mechanisms of data leakage, focusing on covert channels and physical attacks that exploit system vulnerabilities. First, I demonstrated a novel data exfiltration technique using clock modulation in x86 CPUs, showing that dynamic manipulation of CPU frequencies can enable high-speed covert communication. Second, I developed a hardware-level security framework to protect sensitive assets—such as private keys—from cold boot and similar physical attacks. Experimental results confirmed the framework’s ability to maintain data confidentiality even under direct hardware access. Third, I implemented an AI-based classification model that maps software vulnerabilities from the CVE (Common Vulnerabilities and Exposures) database to their root causes in the CWE (Common Weakness Enumeration), effectively identifying recurring coding flaws that lead to security breaches. Collectively, this work advances the field of cybersecurity by providing validated solutions to mitigate data leakage at both hardware and software levels.

Comments

Shariful Alam, ORCID: 0009-0004-6206-2634

DOI

https://doi.org/10.18122/td.2389.boisestate

Available for download on Sunday, August 01, 2027

Share

COinS