Advancing Intrusion Detection in Fog Computing: Unveiling the Power of Support Vector Machines for Robust Protection of Fog Nodes against XSS and SQL Injection Attacks

Abstract

Fog computing, characterized as a cloud infrastructure in close proximity to end devices, faces substantial security challenges that necessitate robust intrusion detection mechanisms for fog nodes. The resource-constrained nature of fog nodes renders them particularly susceptible to attacks, making the development of efficient intrusion detection systems imperative. In this study, we propose a comprehensive approach to protect fog nodes, taking into account their limited resources. Leveraging the power of Support Vector Machines (SVMs), a widely adopted machine learning technique in IoT security, our method overcomes challenges associated with local optima, overfitting, and high-dimensional data. A thorough literature review underscores the prevalent use of SVMs in IoT security research. Specifically, we focus on addressing two prevalent web attacks: Cross-Site Scripting (XSS) and SQL injection attacks, based on global statistical data. To evaluate our approach, we employ the CSE-CIC-IDS2018 dataset and a pseudo-real dataset. Precision, recall, and accuracy are employed as evaluation metrics, along with the Mean Average Precision (MAP). Our evaluation results demonstrate an exceptional level of accuracy, achieving an impressive 98.28% accuracy in terms of average performance when compared to existing methods. Comparative analysis with state-of-the-art approaches further validates the superior efficacy and efficiency of our proposed method.