Magnitude Squared Coherence Based SCA
Magnitude Squared Coherence is a signal processing tool that indicates how well two time domain signals match one with the other by tracking linear dependencies in their spectral decomposition. This paper introduces different ways of using the Magnitude Squared Coherence for Side Channel Analysis. This distinguisher has several advantages over well-known distinguishers. Following, many Side-Channel Attacks have been proposed in the literature. Surprisingly, most of them directly work with time domain traces. Easy connection with engineering intuition could explain this trend despite some inherent drawbacks of such an approach. Among them, the impossibility of capturing the complete leakage scattered over many time samples of raw traces appears as a limitation.