How osama336/network-anomaly-detection-system works
NADS is a Streamlit network-monitoring application. Administrators select an interface and packet count, start live capture, and view packet classifications and visual summaries. The documented backend uses TShark and a Keras model, with Pandas preprocessing and Plotly visualizations; the README also describes anomaly alerts and report generation. The available source excerpt shows capture, queueing, model loading, and dashboard classification, but omits parts of the implementation, so some documented capability links remain unverified.
Main entry points: NADS.py Streamlit controls
Components 12
Monitoring interface
- Streamlit controls
NADS.py
Traffic capture
- Capture runner
NADS.py - Live packet capture
NADS.py - Packet queue
NADS.py
Anomaly detection
- Feature preprocessing
NADS.py - Keras model
- Packet classification
NADS.py
Analysis and insights
- Packet results
NADS.py - Anomaly alerts
NADS.py - Report generation
NADS.py - Traffic visualizations
NADS.py
Other components
- Administrator
Connections 12
- Administrator → Streamlit controls: uses controls
README.md - Streamlit controls → Capture runner: starts capture
NADS.py - Capture runner → Live packet capture: runs capture
NADS.py - Live packet capture → Packet queue: enqueues packets
NADS.py - Streamlit controls → Packet queue: drains packets
NADS.py - Streamlit controls → Packet classification: classifies packets
NADS.py - Packet classification → Feature preprocessing: preprocesses features
NADS.py - Packet classification → Keras model: uses model
NADS.py - Streamlit controls → Packet results: displays packets
NADS.py - Streamlit controls → Anomaly alerts: shows alerts
README.md - Streamlit controls → Report generation: generates reports
README.md - Streamlit controls → Traffic visualizations: renders charts
NADS.py