IJARSAI INTERNATIONAL JOURNAL OF ADVANCED RESEARCH SCIENTIFIC ANALYSIS & INFERENCES
UPID No: 1125P1
Exploring the Role of Explainable AI in Attack Modeling for IoT Networks
Article Title: Exploring the Role of Explainable AI in Attack Modeling for IoT Networks
Author(s): Blessy Thomas, Connected Systems and Intelligence Lab, School of Computer Science and Engineering, Digital University Kerala, Technopark Phase IV, Thiruvananthapuram, Kerala, India, Email: blessy.res21@duk.ac.in
Discipline: Computer Science
Paper Information: Received: 02 Aug 2024| Revised Submission: 08 Sep 2024 | Accepted: 03 Dec 2024| Available Online: 15 Mar 2025
DOI:
Abstract: In an era marked by the pervasive presence of Internet of Things (IoT) networks, it is critical to ensure the strong security of these autonomous systems. The integrity, confidentiality, and availability of IoT networks are threatened by new, complex security concerns that have arisen as a result of the rapid expansion of these networks. Vulnerabilities in the IoT are often exploited by the attacker to gain unauthorized access endangering privacy and security of the network. Identifying potential attack paths in the connected ecosystem is essential for implementing strong security measures. With attack modeling, security experts can evaluate the potential attack paths in the network. We propose a new framework for attack modeling in IoT networks by incorporating Explainable AI(XAI). The XAI algorithm LIME ( Local Interpretable Model Agnostic Explanations) has been included in the proposed framework to enhance the explanation and comprehension of the critical attack paths predicted. A use case is discussed and various graphs with explanations are provided to evaluate the performance of the attack model against the evolving dynamics of the IoT network. XAI provides a crucial layer of defense for protecting IoT networks from cyber hazards by offering interpretable insights into strong security decisions.
Index Terms - Machine Learning, Deep Learning, Explainable AI, IoT Network Security, Vulnerability Analysis, Attack Modelling.
Download Full Article
This website uses cookies.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.
Papers are welcome for January-June Issue
You can Submit Online. Click on the Submit Paper Button