On Saturday 18 October, we officially presented NetFlowMeterthe new open-source tool for analysing network flows, at the No Hat 2025one of the most relevant Italian conferences in the field of computer security and ethical hacking.
In the field of network intrusion detectionflow analysis by means of Machine Learning techniques is increasingly widespread and to be effective requires a correct reconstruction of network streams from packets, a process entrusted to the so-called flow generator.
During our research, we found important limitations in one of the most widely used tools, CICFlowMeter, such as correctness problems in flow generation and parsing performance.
That is why we released NetFlowMeterby pursuing three key objectives:
- improve the accuracy of reconstructed streams, raising the quality of datasets used in ML models;
- optimise performance to support real-time scenarios and intensive traffic;
- provide an open, transparent and community-validated tool.
At the conference we also presented concrete use cases, showing how NetFlowMeter can make a real contribution to the protection of digital infrastructures.
Our ambition is for this software to support researchers, security teams and the offensive/defensive community in improving the state of the art in traffic monitoring.
We would like to thank the staff of No Hat for the opportunity to share our work with a community so lively and rich in ideas, whose confrontation represents for us a concrete value of growth and continuous improvement.
For the slides and source code of the presentation, you can visit this address:
