Abiodun Olaluwe Dissertation Proposal Defense, Monday, December 15, 2025 @ 10:00 am Central Time
December 15 @ 10:00 am - 11:00 am
COMMITTEE CHAIR: Dr. Annamalai Annamalai
COMMITTEE CO-CHAIR: Dr. Akshay Kulkarni
TITLE: A MACHINE LEARNING-ASSISTED FRAMEWORK FOR SIDE-CHANNEL ANALYSIS ON POST QUANTUM CRYPTOGRAPHY
ABSTRACT: The advancement of quantum computing has rendered classical cryptographic systems such as RSA and elliptic curve cryptography increasingly vulnerable, necessitating the development of post-quantum cryptographic (PQC) algorithms capable of withstanding both classical and quantum adversaries. While PQC algorithms such as CRYSTALS-Kyber, CRYSTALS-Dilithium, and Hamming-based Quasi-Cyclic (HQC) offer strong mathematical security, their hardware implementations remain susceptible to side-channel attacks (SCAs) that exploit physical leakages in power, timing, or electromagnetic emanations. This dissertation proposes a Machine-Learning-Assisted Side-Channel Analysis (ML-SCA) Framework to systematically evaluate and mitigate leakage in PQC hardware. The study follows a three-phase methodology: (i) Leakage Characterization using Test Vector Leakage Assessment (TVLA) and RMS envelope analysis to identify points of interest (POIs); (ii) Machine-Learning Profiling employing convolutional and recurrent neural networks to model complex leakage patterns and quantify resistance using success rate (SR) and guessing entropy (GE) metrics; and (iii) Countermeasure Evaluation by integrating adaptive masking and fault-injection testing within FPGA-based PQC implementations. Using the ChipWhisperer CW305 platform, the research develops reproducible, vendor-neutral evaluation procedures that bridge cryptographic theory with physical resilience. The framework provides a quantitative basis for certifying quantum-resistant hardware and contributes novel insights into machine-learning-based defense mechanisms. Ultimately, this work strengthens the scientific and practical foundation for ensuring the implementation security of PQC in the era of quantum computing.
Keywords: Post-quantum cryptography, side-channel analysis, machine learning, hardware security
Room Location: Electrical Engineering Building Room 315D
