3 ms·
How interesting. Thanks for this info. The SeaGlass method is much more data-heavy, it seems, which is both good and bad. A good baseline is labor-intensive to
by devindotcom 9y ago
How interesting. Thanks for this info. The SeaGlass method is much more data-heavy, it seems, which is both good and bad. A good baseline is labor-intensive to create and requires lots of physical and temporal presence, but once you have it, it's a great way to vet outlier signals. Seems like that's the point you want to apply work like yours.
- Datenstrom 9y agoYes, SeaGlass seems much more advanced of a technique than the commercial solution I discovered which also just consisted of a single board computer and an SDR. Certainly standalone IMSI Catcher detectors with hardcoded metrics would have a hard time keeping up with the rapidly evolving networks and new attacks without more data about what is going on to update them. I was considering the use of ML for more general detection beyond the hardcoded metrics. I think I read that SnoopSnitch was gathering data but only for detection of known attacks not anomalies.