Přístupnostní navigace
E-application
Search Search Close
Bachelor's Thesis
Author of thesis: Bc. Jaroslav Cempírek
Acad. year: 2025/2026
Supervisor: Ing. Petr Ilgner, Ph.D.
Reviewer: Ing. Minh Tran
This bachelor's thesis addresses anomaly detection in network traffic of Internet of Things (IoT) devices in a home environment, with emphasis on user privacy protection. The theoretical part describes the architecture of IoT networks, typical communication protocols, anomaly detection methods ranging from statistical approaches to machine learning algorithms, and legal aspects of network data processing under the GDPR regulation. The practical part implements a test environment with three IoT devices (IP camera, smart bulb, temperature sensor) whose traffic is mirrored from a MikroTik router to a Raspberry Pi using the TZSP protocol. Statistical features were extracted from more than 21 000 captured network flows using CICFlowMeter, and a separate Isolation Forest model was trained for each device. The detection capability was validated both on simulated anomalies and during a three-month deployment in a real home network, during which the system captured phenomena such as concept drift following device firmware updates. The system is complemented by a web dashboard, automatic identification of new devices by MAC address, and a model retraining mechanism. The thesis concludes with a discussion of the limitations of the approach and recommendations for enhancing user privacy.
Internet of Things, IoT, anomaly detection, machine learning, Isolation Forest, network traffic, CICFlowMeter, Raspberry Pi, GDPR, privacy protection, concept drift
Date of defence
16.06.2026
Result of the defence
Defended (thesis was successfully defended)
Grading
A
Process of defence
Student prezentoval výsledky své práce a komise byla seznámena s posudky. Student obhájil bakalářskou práci a odpověděl na otázky členů komise a oponenta. Otázky: 1) Přetrénování modelů se dnes spouští ručně. Podle jakých signálů by systém mohl sám rozpoznat, že u některého zařízení nastal concept drift a je třeba model přetrénovat? 2) Výběr 27 příznaků z 82 jste odvodil z popisné úvahy nad komunikačními vzory. Jak byste kvantitativně doložil, že právě tyto příznaky nejlépe oddělují normální provoz od anomálního? 3) Jaká je tedy úspěšnost?
Language of thesis
Czech
Faculty
Fakulta elektrotechniky a komunikačních technologií
Department
Department of Telecommunications
Study programme
Information Security (BPC-IBE)
Composition of Committee
doc. Ing. Jan Jeřábek, Ph.D. (předseda) doc. Ing. Ivo Lattenberg, Ph.D. (místopředseda) Mgr. Martin Erlebach (člen) Ing. Eva Holasová, Ph.D. (člen) Ing. Petr Ilgner, Ph.D. (člen) Ing. Kryštof Zeman, Ph.D. (člen) Ing. Martin Rusz, Ph.D. (člen)
Supervisor’s reportIng. Petr Ilgner, Ph.D.
Grade proposed by supervisor: A
Reviewer’s reportIng. Minh Tran
Grade proposed by reviewer: B
Responsibility: Mgr. et Mgr. Hana Odstrčilová