Abstract In recent years, the application of deep learning in cryptanalysis has gained significant attention, particularly with the emergence of neural network-based distinguishers. At CRYPTO’19, Gohr demonstrated that neural networks could develop differential distinguishers capable of producing highly competitive attacks against existing methods. Building on this foundation, we propose multiple input polytopic differential neural distinguishers ( PDND s) for the lightweight block ciphers SIMON, SIMECK, and SPECK. Our approach incorporates a novel data generation method that utilizes two polytope differences, resulting in more precise training data and enhanced model accuracy. Through extensive experiments in single-key and related-key scenarios, we evaluate and validate the intrinsic performance of our neural distinguishers. Our results show that PDND s significantly outperform the baseline, polytopic, multiple, and mixture differential neural distinguishers, utilizing a single input difference, in accuracy across various cipher rounds. Notably, our PDND s achieved $$100\%$$ 100 % accuracy for up to 7 rounds of SIMON32 and SIMECK32, and $$99.39\%$$ 99.39 % accuracy for 5 rounds of SPECK32 in the single-key scenario. Additionally, for extended rounds, we achieved accuracy levels of up to 12 rounds for SIMON32, 13 for SIMECK32, and 8 for SPECK32 without requiring staged training. In the related-key scenario, our method further improved performance, introducing 13-round and 15-round RK-PDND s for SIMON32 and SIMECK32, respectively, underscoring the enhanced capabilities of our approach. Furthermore, we demonstrate the effectiveness of our neural distinguishers through a key recovery test, where they successfully distinguish between correct and incorrect keys, confirming the practical applicability of our approach in cryptanalysis.

As Message Queuing Telemetry Transport (MQTT) becomes a widely adopted protocol for IoT communication, the use of traditional Transport Layer Security (TLS) encryption poses challenges due to its high computational demands on resource-constrained devices and one-way authentication. In this paper, we propose a novel lightweight mutual authentication protocol designed for MQTT-based communication in Industrial Internet of Things (IIoT) environments. Our protocol ensures message integrity without the high computational overhead associated with TLS encryption. By leveraging an alternative Authenticated Encryption (AE) method optimized for resourceconstrained devices, we enhance both security and efficiency. To evaluate the performance, we conducted experiments using three Raspberry Pi devices, comparing our protocol to both non-encrypted and TLS-secured MQTT. The results demonstrate a significant reduction in system run-time, from 139 ms to 108 ms, when using the proposed protocol for mutual authentication, compared to default TLS, which only supports oneway authentication. Additionally, simulations conducted in NS3, across various practical scenarios, reveal notable improvements in authentication success rates and reduced delays. Our findings suggest that this protocol offers a promising solution for secure and efficient communication in IIoT systems.

SIMECK is a family of three lightweight block ciphers designed by Yang et al ., following the framework used by Beaulieu et al . from the United States National Security Agency to design SIMON and SPECK. In this study, the authors employ an improved miss‐in‐the‐middle approach to find zero correlation linear distinguishers and impossible differentials on SIMECK48 and SIMECK64. Based on this novel technique, they will be able to present zero‐correlation linear approximations for 15‐round SIMECK48 and 17‐round SIMECK64 and these zero‐correlation linear approximations improve the previous best result by two rounds for SIMECK48 and SIMECK64. Moreover, they attack 27‐round SIMECK48 and 31‐round SIMECK64 based on these zero‐correlation linear distinguishers. In addition, due to the duality of zero‐correlation and impossible differential, they search for the impossible differential characteristics for SIMECK48 and SIMECK64 so that they will be able to present 15‐round SIMECK48 and 17‐round SIMECK64 while the best previously known results were 13‐round impossible differentials for SIMECK48 and 15‐round impossible differentials for SIMECK64. Moreover, they propose impossible differential attacks on 22‐round SIMECK48 and 24‐round SIMECK64 based on these impossible differential characteristics. The results significantly improve the previous zero correlation attack and impossible differential characteristic results for these variants of SIMECK to the best of the authors’ knowledge.