Secure edge-based smart grid communication using lightweight authentication modeling with autoencoders and real-world data
Abstract The growing use of edge computing in intelligent grids presents security issues as a result of the limited processing capabilities of edge devices and the intensifying complexities of cyberattacks. The conventional authentication schemes based on cryptographic protocols have excessive compu...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-06-01
|
| Series: | Discover Computing |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s10791-025-09643-w |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract The growing use of edge computing in intelligent grids presents security issues as a result of the limited processing capabilities of edge devices and the intensifying complexities of cyberattacks. The conventional authentication schemes based on cryptographic protocols have excessive computational overhead, rendering them inappropriate for real-time grid communication. This paper presents a light-weight authentication scheme based on a five-layer deep autoencoder for anomaly-based authentication, facilitating secure and efficient communication in resource-limited edge-based smart grids. Temporal sequencing and feature normalization are utilized to optimize the model to enhance detection accuracy while minimizing computational complexity. Moreover, an adaptive thresholding scheme is incorporated to increase the robustness of the model against changing cyber-attacks, maintaining high reliability in dynamic smart grid scenarios. Experimental verification on the ICS dataset shows that the proposed framework attains 99.37% detection accuracy, actual rejection rate of 99.4%, and false rejection rate as low as 0.50%, outperforming the state-of-the-art solutions. Moreover, the framework facilitates real-time authentication with latency of ≤ 2 ms and ultra-low energy expenditure (0.015 W per authentication decision), being 30% faster and 50% more power-efficient compared to state-of-the-art solutions. The model performs at a high level even under heavy communication loads and demonstrated its scalability to 50,000 + edge devices. These findings set a new standard in smart grid security, delivering a low-latency, scalable, and energy-efficient authentication solution for deployment in real-world scenarios. The system is tailored to meet Middle East and Turkey regional cybersecurity challenges, tackling infrastructure asymmetry and environmental limitations in smart grid systems. |
|---|---|
| ISSN: | 2948-2992 |