Zero-Value Filtering as a Side-Channel Proactive Defense: From Attack to Countermeasure

Tosun et al. (2024) exploit zero-value patterns in incomplete NTT to accelerate attacks, revealing a critical vulnerability. This research reframes ZV-FA as a proactive defense: hardware/software monitors NTT polynomial multiplications for anomalous zero-value patterns (e.g., statistically improbable coefficient distributions) and triggers countermeasures like dynamic masking or algorithm switching. Unlike Praditasari et al.’s (2025) AI-based anomaly detection (which needs benign traces), this method uses algorithmic invariants as signatures of attacks. By integrating with Heinz et al.’s (2023) RNR for side-channel hardening, it could create a multi-layered shield. Impact: Transforms an attack vector into a low-overhead, mathematically grounded defense for NIST-standardized schemes (Kyber/Dilithium).

References:

1. Combined Fault and DPA Protection for Lattice-Based Cryptography. Daniel Heinz, T. Pöppelmann (2023). IEEE transactions on computers.
2. AI-Driven Defense Mechanism for Lattice-Based Post-Quantum Cryptography: Adaptive Mitigation Against Side-Channel Attacks. W. A. Praditasari, Hyungyeop Kim, Hyejin Yoon, Danang Rimbawa, Okyeon Yi (2025). Asia Joint Conference on Information Security.
3. Zero-Value Filtering for Accelerating Non-Profiled Side-Channel Attack on Incomplete NTT-Based Implementations of Lattice-Based Cryptography. Tolun Tosun, E. Savaş (2024). IEEE Transactions on Information Forensics and Security.