Profiling IoT botnet activity

Almazarqi, Hatem (2024) Profiling IoT botnet activity. PhD thesis, University of Glasgow.

Abstract

Undoubtedly, Internet of Things (IoT) devices have evolved into a necessity within our modern lifestyles. Nonetheless, IoT devices have proved to pose significant security risks due to their vulnerabilities and susceptibility to malware. Evidently, vulnerable IoT devices are enlisted by attackers to participate into Internet-wide botnets in order to instrument large-scale cyber-attacks and disrupt critical Internet services. Tracking these botnets is challenging due to their varying structural characteristics, and also due to the fact that malicious actors continuously adopt new evasion and propagation strategies. This thesis develops BotPro framework, a novel data-driven approach for profiling IoT botnet behaviour. BotPro provides a comprehensive approach for capturing and highlighting the behavioural properties of IoT botnets with respect to their structural and propagation properties across the global Internet. We implement the proposed framework using real-world data obtained from the measurement infrastructure that was designed in this thesis. Our measurement infrastructure gathers data from various sources, including globally distributed honeypots, regional Internet registries, global IP blacklists and routing topology. This diverse dataset forms a strong foundation for profiling IoT botnet activity, ensuring that our analysis accurately reflects behavioural patterns of botnets in real-world scenarios. BotPto encompasses diverse methods to profile IoT botnets, including information theory, statistical analysis, natural language processing, machine learning and graph theory.

The framework’s results provide insights related to the structural properties as well as the evolving scanning and propagation strategies of IoT botnets. It also provides evidence on concentrated botnet activities and determines the effectiveness of widely used IP blacklists on capturing their evolving behaviour. In addition, the insights reveal the strategy adopted by IoT botnets in expanding their network and increasing their level of resilience. The results provide a compilation of the most important autonomous system(AS) attributes that frequently embrace IoT botnet activity as well as provide a novel macroscopic view on the influence of AS-level relationships with respect to IoT botnet propagation. Furthermore, It provides insights into the structural properties of botnet loaders with respect to the distribution of malware binaries of various strains. The insights generated by BotPro are essential to equip next generation automated cyber threat intelligence, intrusion detection systems and anomaly detection mechanisms with enriched information regarding evolving scanning, establishment and propagation strategies of new botnet variants. Industry will be equipped with even more improved ways to defend against emerging threats in the domains of cyber warfare, cyber tourism and cybercrime. The BotPro framework provides a comprehensive platform for stakeholders, including cybersecurity researchers, security analysts and network administrators to gain deep and meaningful insights into the sophisticated activities and behaviour exhibited by IoT botnets.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
• Colleges/Schools: College of Science and Engineering > School of Computing Science
• Supervisor's Name: Marnerides, Dr. Angelos
• Date of Award: 2024
• Depositing User: Theses Team
• Unique ID: glathesis:2024-84102
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 22 Feb 2024 14:22
• Last Modified: 22 Feb 2024 14:24
• Thesis DOI: 10.5525/gla.thesis.84102
• URI: https://theses.gla.ac.uk/id/eprint/84102

Related URLs

• Conference proceeding
• Conference proceeding