Citation

Umar, Muhammad and Rehman, Habiba and Ahmed Sami, Adnan and Siraj, Mohammad and Channa, Muhammad Ibrahim and Kamal, Shahid (2026) Advancing Cybersecurity in IoT Networks: A Q-Learning Approach to Mitigate Routing Attacks. In: 2026 5th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET), 22-23 May 2026, Sukkur, Pakistan.

Text 11.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

The Internet of Things (IoT) is a system of interconnected devices that exchange information through the Internet. Such networks are important as they can be automated, allow real-time monitoring, and make decisions based on data, hence opening up new opportunities in smart homes, healthcare, agriculture, transportation, and smart cities. Despite the significant advantages, IoT networks are still susceptible to security threats, specifically routing attacks, thereby disrupting communication, manipulating the routing paths and causing packet loss, service degradation, and compromised data integrity. Conventional signature-based or rule-based Intrusion Detection Systems (IDS) lacks the ability to adapt to new attack patterns. Therefore, this study proposes an adaptive and intelligent Q-Learning-based IDS (Q-IDS) scheme that dynamically classifies the IoT nodes as Malicious node and benign, based on the behaviors. Every node has a record of the trust value (Q -value) of its neighbouring nodes. The reward of +n is given to cooperative behaviour and the penalty of −m to malicious behaviour. Whenever a Q -value is below the predetermined threshold, the node is considered as a malicious and isolated from the network; otherwise, the node is treated as trustworthy. The simulated results in a dynamic IoT network shows that Q-IDS scheme can accurately identify malicious nodes and improve network performance in terms of throughput, end-to-end delay, and routing load. This research contributes to the field of cybersecurity, as it provides an effective and secure solution to mitigate routing attacks and enhancing the security posture of modern networks.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	routing attacks, Cyber Security, machine learning, temporal difference learning, accuracy
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	31 Jul 2026 06:24
Last Modified:	31 Jul 2026 06:24
URII:	http://shdl.mmu.edu.my/id/eprint/16412

Downloads

Edit (login required)