Fuzzing+Hardware Performance Counters-Based Detection of Algorithm Subversion Attacks on Post-Quantum Signature Schemes
NIST is standardizing Post Quantum Cryptography (PQC) algorithms that are resilient to the computational capability of quantum computers. Past works show malicious subversion with cryptographic software (algorithm subversion attacks) that weaken the implementations. We show that PQC digital signature codes can be subverted in line with previously reported flawed implementations that generate verifiable, but less-secure signatures, demonstrating the risk of such attacks. Since, all processors have built-in Hardware Performance Counters (HPCs), there exists a body of work proposing a low-cost Machine Learning (ML)-based integrity checking of software using HPC fingerprints. However, such HPC-based approaches may not detect subversion of PQC codes. A miniscule percentage of qualitative inputs when applied to the PQC codes improve this accuracy to 98%. We propose grey-box fuzzing as a pre-processing step to obtain inputs to aid the HPC-based method.