KEYWORD |
Big data analytics and AI for network traffic monitoring and anomaly detection
keywords BIG DATA, DATA ANALYTICS, INTERNET, MACHINE LEARNING
Reference persons DANILO GIORDANO, MARCO MELLIA
External reference persons Martino Trevisan
Research Groups DAUIN - GR-04 - DATABASE AND DATA MINING GROUP - DBDM
Thesis type DATA ANALYSIS, MACHINE LEARNING, RESEARCH THESIS WITH A COMPANY
Description The cooperation with the company TIM can be seen here. For Internet Service Providers (ISPs), monitoring network traffic is an important tool for understanding network performance, identifying customer concerns, and troubleshooting problems. In this direction, one of the goals is to monitor network performance to understand issues related to objective metrics measurable with Quality of Service (QoS) metrics and the customer's subjective Quality of Experience (QoE).
This work will examine network monitoring data an ISP collects during major live-streaming sporting events on a network with thousands of customers.
The thesis will focus on characterizing the data to understand network patterns and performance and finally design unsupervised machine learning pipelines to group customers or identify customers with unusual behaviour
Aim of the work
- characterize the data to understand network patterns and performance
- Machine learning is used for clustering and anomaly detection to identify groups of users or users with unusual behaviour.
Required skills Python, Machine Learning, Spark
Deadline 26/12/2025
PROPONI LA TUA CANDIDATURA