Detection of cyberattacks affecting IoT devices

State: Open

The proposed thesis focuses on designing, implementing, and validating a framework based on Machine and Deep learning and in charge of detecting cyberattacks affecting devices with limited resources (such as Raspberry Pis). With that goal in mind, we will start studying the state-of-the-art regarding intelligent solutions able to detect anomalies produced by malware affecting device behaviours. After that, we will evaluate relevant dimensions and characteristics able to create behavioural profiles. A particular malware will then be used to infect our device, and appropriate machine and deep learning techniques will be integrated into the proposed framework. Finally, the performance of the proposed solution will be evaluated.


[1] A Survey on Device Behavior Fingerprinting: Data Sources, Techniques, Application Scenarios, and Datasets —> https://arxiv.org/abs/2008.03343


[2] On the Feasibility of Online Malware Detection with Performance Counters—>  https://dl.acm.org/doi/pdf/10.1145/2508148.2485970 

20% literature review, 30% Design, 20% Implementation, 15% Evaluation , 15% Documentation
Basic knowledge in Linux, Machine and Deep Learning

Supervisors: Dr Alberto Huertas

back to the main page