Gotta Detect 'Em All: Fake Base Station and Multi-Step Attack Detection...
summary
FBSDetector identifies fake base stations and 21 multi-step attacks from layer‑3 traces using stateful LSTM with attention, packet‑level then trace‑level classification, and graph neural networks on packet‑flow graphs. It trains on POWDER‑collected FBSAD/MSAD datasets, runs as an Android app, and is validated with USRP B210s and srsRAN. Reported results: 96% accuracy, 2.96% FPR for FBS; 86% accuracy, ~3.3% FPR for MSAs, 835 KB memory and <1 ms inference.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:17Fake base station challenges
- 2:21Building FBS dataset
- 3:24FBS detection framework
- 5:27Multi-step attack detection
- 6:29Deployment results
- 8:15Hardware validation
- 9:18Robustness against attacks
- 10:21Summary
related talks
24mEnhancing Physical Perimeter Defense Using SDR
22mHacking LTE Public Warning Systems
44mNetsec is Dead(?): Modern Network Fingerprinting for...
16mA System to Detect Forged-Origin BGP Hijacks
01h 02mHacking the Wireless World with Software Defined Radio - 2.0
53mHacking the Wireless World with Software Defined Radio - 2.0
30mSoftware defined radio based Global Navigation Satellite System real time spoofing....
29mYou Can’t Hide: How a Device Can Give Away a Fraudster’s Secrets.