Abstract

This work presents a new AI-driven Adaptive Lightweight Key Management (ADAL-KM) technique that combines lightweight cryptography operations with an intelligent key-risk assessment approach. An intelligent lightweight adaptive decision-making model is utilized to set proper key lifetimes and session keys distribution within hierarchical device groups. Thus, ADAL-KM saves cryptographic resources while ensuring security. The experimental assessment based on CICIoT2023 IoT traffic shows that ADAL-KM reduces the overheads of key management communication to 57.91 KB at 40 devices, which is 22.6% less than 74.86 KB for the Efficient Key Management System. The computation overhead drops to 121.1 ms from 158.7 ms, and the energy overhead decreases to 12.09 J from 15.43 J, which is a 21.7% decrease. The success rate of establishment of secure sessions by the method proposed is 98.9%, while it is 94.2% for the Efficient Key Management System. The rekeying latency of ADAL-KM becomes 71.0 ms, which is 26.4% lower than 96.5 ms.

Authors

Mohammed Ali Sohail¹, Deepika Lokesh²
Jazan University, Kingdom of Saudi Arabia¹, C. Byre Gowda Institute of Technology, India²

Keywords

Internet of Things, Cryptographic Key Management, Lightweight Cryptography, Artificial Intelligence, Secure Communication