Horizontal SCA Attacks against kP Algorithm Using K-Means and PCA
This addresses security vulnerabilities in cryptographic systems for practitioners, though it is incremental as it applies known ML techniques to side-channel analysis.
The paper tackles the problem of side-channel attacks on ECDSA implementations by developing two machine learning-based methods (K-means and PCA) to reveal cryptographic keys, achieving 100% accuracy on unprotected implementations and up to 98.3% accuracy on those with countermeasures.
Side Channel Analysis attacks take advantage of the information leaked from the implementations of cryptographic algorithms. In this paper we describe two key revealing methods which are based on machine learning algorithms: K-means and PCA. We performed the attacks against ECDSA implementations without any prior knowledge about the key and achieved 100% accuracy for an implementation without any countermeasures against horizontal attacks and 88.7% accuracy for an implementation with bus address sequencing. In the scenario where the kP operation inputs are controlled by the attacker (as during signature verification), we achieved 98.3% accuracy for the implementation with countermeasures.