Graph Representation Learning of Lightweight IoT Ciphers
2026-08-24 • Machine Learning
Machine LearningCryptography and Security
AI summaryⓘ
The authors studied two lightweight encryption methods called SIMON and SIMECK, which are used in small devices like those in the Internet of Things. They used machine learning techniques to better find and show groups of likely differences that can help break the encryption. By creating special features from partial data and building graphs with different ML methods, they could perfectly spot high-probability differences without mistakes. Their approach also helped visualize these difference clusters clearly, and it worked well across both encryption types, showing it might apply to similar algorithms too.
Lightweight Cryptographic AlgorithmsFeistel CipherDifferential CryptanalysisDifference Distribution TableMachine LearningGraph Representation LearningK-Nearest NeighbourDecision TreesRandom ForestsInternet of Things
Authors
Jonathan Cook, Sabih ur Rehman, M. Arif Khan
Abstract
SIMON and SIMECK belong to a family of Lightweight Cryptographic Algorithms (LCAs) based on the Feistel block cipher, designed for Internet of Things (IoT) devices. As with all Feistel ciphers, they are susceptible to differential cryptanalysis, necessitating rigorous resilience evaluations. While state-of-the-art techniques leverage heuristics and sampling to improve efficiency, little work has applied Machine Learning (ML) guided Graph Representation Learning (GRL) to efficiently identify and visualise high-probability differential clusters. We address this gap by introducing an efficient feature engineering strategy that extracts four differential attributes from a partial Difference Distribution Table (pDDT), revealing structural information concealed in raw differential data. Utilising the enriched features, we construct and compare three ML-guided directed graphs for SIMON$32$ and SIMECK$32$ using K-Nearest Neighbour (KNN), Decision Trees (DT), and Random Forests (RF). To the best of our knowledge, our framework produces the first graph-based visualisation of the differential clustering effect, in which high-probability single-bit differentials form geometrically close clusters in the learned embedding. All three models achieve a precision of $1.0$ in identifying high-probability differentials, confirming zero false positives. KNN achieves the strongest cluster separation, the highest F1 score and the lowest graph construction time of approximately $2.3$ seconds, while DT and RF produce optimal paths with near-perfect regression. The results are consistent across both LCAs, demonstrating the applicability of the framework to other AND-rotation LCA families.