Public defence in Networking Technology, M.Sc. Mehrnoosh Monshizadeh

M.Sc. Mehrnoosh Monshizadeh will defend the thesis "Machine Learning Techniques to Detect Known and Novel Cyber-attacks" on 22 May 2023 at 12 (EET) in Aalto University School of Electrical Engineering, Department of Information and Communications Engineering, in lecture hall TU2, Maarintie 8, Espoo.

Opponent: Prof. Jerry Chun-Wei Lin, Western Norway University of Applied Sciences, Norway
Custos: Prof. (emeritus) Raimo Kantola, Aalto University School of Electrical Engineering, Department of Information and Communications Engineering

Intrusion Detection Systems (IDS) are considered well-known tools for monitoring and detecting malicious traffic in communication networks. However, traditional IDSs rely on known signature and lack the ability to detect novel attacks. This problem has motivated researchers to incorporate machine learning (ML) algorithms in the IDS architecture. However, in order to train algorithms in ML-based IDSs, obtaining reliable datasets with appropriate characteristics is a major challenge. Due to the lack of labelled datasets, ML-based IDs suffer from overfitting problem which make them inefficient for real time intrusion detection. In ML models, overfitting occurs when a model that is trained with a dataset may not have the same outcome with another dataset. Furthermore, in real-life scenarios, considerable amount of incoming data does not belong to any known category; and for unknown traffic, dividing data into the classes without having information on the nature of the traffic is challenging. In addition, annotating large datasets is very costly and hence we can label only few examples manually. On the other hand, the 5G+ and 6G networks are expected to deliver massive connectivity to numerous IoT/IoE devices, where huge amount of data needs to be analyzed by AI enabled mechanisms. Consequently, a mature scaling architecture must be considered as a mandatory feature in ML-based IDSs. And lastly, AI becomes crucial part of 5G+ architecture and technologies, not only as enabler, but also as defender and offender. However, AI can also be employed to launch intelligent attacks where, Adversarial attacks on AI systems are becoming more and more of security concern in these networks. Consequently, mitigation for adversarial attacks requires more intelligent defence systems empowered by AI methods that with minimum human interaction learn the attacker algorithms behaviour and weaken its process. This study proposes an intelligent, modular, robust and scalable security solution to dynamically detect known and unknown cyber-attacks targeting mobile networks.

Thesis available for public display 10 days prior to the defence at: https://aaltodoc.aalto.fi/doc_public/eonly/riiputus/
Doctoral theses in the School of Electrical Engineering: https://aaltodoc.aalto.fi/handle/123456789/53

Contact information of doctoral candidate: