AI ‐Enhanced Lightweight Key Agreement and Secure Access for Power Wireless Sensor Networks
Abstract
ABSTRACT The research focuses on the secure communication of power wireless sensor networks in smart grid environment. Aiming at the problem of the traditional key negotiation mechanism with high resource overhead, complex protocol, and difficult to adapt to the pain point of resource‐constrained devices, a Lightweight Key Agreement and Secure Access Protocol Model for Power‐aware Wireless Sensor Networks (LKAPM) is designed and optimized. The model combines wireless physical layer feature information to generate a shared key, and adopts a challenge‐answer mechanism based on timestamps and authenticated random numbers. It can ensure secure access with low computation and communication overhead without relying on traditional heavy encryption algorithms. This study constructs a network experimental platform that simulates the power terminal environment and systematically compares LKAPM with existing protocols. The experimental results indicated that LKAPM was 27.36 ms in average handshake delay and 608.24 bytes in communication overhead. The node computation time was 4.91 ms, and the average memory occupation was controlled at 6.73 KB. Meanwhile, the percentage of successfully establishing secure connections in the 50‐group attack environment test reached 99.16%, which was better than most of the existing light‐weight security protocols. In supporting wireless physical layer security, LKAPM could effectively resist replay attacks and man‐in‐the‐middle attacks, and had strong privacy protection capability. The research results have practical significance for improving the data communication security of power terminals in smart grids. In particular, it has good application value in key research directions such as mobile network security, smart grid and power/energy system security, cyber‐physical security, Internet of Things and machine‐to‐machine technology.
Fields of Study & Subjects
Security in Wireless Sensor Networks
Applied Advanced Technologies
Advanced Technologies and Applied Computing
Publication details
- Journal / source
- Security and Privacy
- Volume
- 9
- Issue
- 1
- Publication date
- 1 January 2026
- Language
- EN