Title: An architectural framework for interoperable security anomaly detection in edge computing and SDN using gradient-boosted trees

Authors: R. Gowrishankar; V.J. Arulkarthick; A.P. Janani; K. Sundaresan; U. Barakkath Nisha; R. Yasir Abdullah; Thrivikram Bathini

Addresses: Kalaignarkarunanidhi Institute of Technology (KIT), Coimbatore, Tamil Nadu, India ' JCT College of Technology, Coimbatore, Tamil Nadu, India ' Dr. Mahalingam College of Engineering and Technology, Pollachi, Coimbatore, Tamil Nadu, India ' Karpagam Institute of Technology, Coimbatore, Tamil Nadu, India ' Sri Krishna College of Engineering and Technology, Coimbatore, Tamil Nadu, India ' Sri Krishna College of Engineering and Technology, Coimbatore, Tamil Nadu, India ' Sree Rama Engineering College, Tirupati, Andhra Pradesh, India

Abstract: In the ever-changing environment of edge computing and Software-Defined Networking (SDN), it is crucial to have strong security measures in place to protect against new threats and ensure compatibility between different systems. Conventional anomaly detection algorithms frequently have difficulties in dealing with the intricacy and dynamic characteristics of such situations. This paper presents a specialised architectural framework designed for edge computing and SDN, utilising the sophisticated capabilities of Gradient-Boosted Trees (GBT) to achieve efficient anomaly detection. Our framework aims to improve security in edge computing and SDN infrastructures by utilising the predictive capabilities and adaptability of GBT. GBT offers enhanced precision in identifying security anomalies when compared to conventional approaches by examining extensive and diverse data streams prevalent in such contexts. The effectiveness of the suggested architecture in simulated edge computing and SDN systems is proven through empirical evaluations. The results demonstrate substantial enhancements in the detection of anomalies and a decrease in false positive occurrences, underscoring the capability of GBT in strengthening security within intricate and ever-changing ecosystems. This research enhances security solutions specifically designed for edge computing and SDN environments. The suggested architecture establishes the foundation for future research and development endeavours focused on tackling the changing security challenges in edge computing and SDN domains.

Keywords: edge computing; SDN; software-defined networking; security anomaly detection; GBT; gradient-boosted trees; interoperability; framework architecture.

DOI: 10.1504/IJGUC.2025.148540

International Journal of Grid and Utility Computing, 2025 Vol.16 No.5/6, pp.481 - 492

Received: 28 Jun 2024
Accepted: 06 Aug 2024
Published online: 11 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article