A Machine Learning-Based Framework for Assessing Cryptographic Indistinguishability of Lightweight Block Ciphers

TL;DR

Indistinguishability is a foundational cryptographic security property ensuring that ciphertexts from two messages encrypted under the same key are indistinguishable with probability not significantly better than $0.5$. The paper introduces MIND-Crypt, a machine learning–based framework that employs CNNs, LSTMs, BiLSTMs, and ResNets to test whether ciphertexts from SPECK32/64 and SIMON32/64 in CBC mode under Known Plaintext Attacks reveal distinguishable patterns. Across extensive experiments, all DL models perform at or near the random baseline ($ ext{Accuracy} \,\approx\,$ $0.5$), with memorization effects observed when IV entropy is reduced, thereby reinforcing the cryptographic security of these lightweight ciphers under ML-based indistinguishability assessments. These findings provide practical reassurance for IoT deployments relying on such ciphers and are complemented by plans to release code and data for reproducibility; the study also outlines directions for future exploration of more advanced or quantum-inspired ML techniques.

Abstract

Indistinguishability is a fundamental principle of cryptographic security, crucial for securing data transmitted between Internet of Things (IoT) devices. This principle ensures that an attacker cannot distinguish between the encrypted data, also known as ciphertext, and random data or the ciphertexts of the two messages encrypted with the same key. This research investigates the ability of machine learning (ML) in assessing indistinguishability property in encryption systems, with a focus on lightweight ciphers. As our first case study, we consider the SPECK32/64 and SIMON32/64 lightweight block ciphers, designed for IoT devices operating under significant energy constraints. In this research, we introduce MIND-Crypt, a novel ML-based framework designed to assess the cryptographic indistinguishability of lightweight block ciphers, specifically the SPECK32/64 and SIMON32/64 encryption algorithm in CBC mode (Cipher Block Chaining), under Known Plaintext Attacks (KPA). Our approach involves training ML models using ciphertexts from two plaintext messages encrypted with same key to determine whether ML algorithms can identify meaningful cryptographic patterns or leakage. Our experiments show that modern ML techniques consistently achieve accuracy equivalent to random guessing, indicating that no statistically exploitable patterns exists in the ciphertexts generated by considered lightweight block ciphers. Furthermore, we demonstrate that in ML algorithms with all the possible combinations of the ciphertexts for given plaintext messages reflects memorization rather than generalization to unseen ciphertexts. Collectively, these findings suggest that existing block ciphers have secure cryptographic designs against ML-based indistinguishability assessments, reinforcing their security even under round-reduced conditions.

Figures (1)

• Figure 1: The MIND-Crypt assessment framework - Investigating the indistinguishability of SPECK32/64 and SIMON32/64 lightweight block ciphers. Two plaintext messages encrypted with the same key, represented in binary format, form the basis for training a DL model.