How cyberjsec47/signal-sentinel works
Project type and purpose
• This is an embedded machine-learning detection system (“Signal Sentinel”) that passively detects RF-jamming attacks.
• It consists of two main environments:
a. A development/training environment (desktop/server) where you collect data, extract features, and train a classification model.
b. A production/embedded environment (Raspberry Pi + RTL-SDR) that runs the trained model in real time, classifies signals, and raises alerts or triggers mitigation integration.
File structure
Top-level files:
• Calculations.py, FeatureReduction.py, Functions.py – modules for signal processing and feature preparation.
• ModelTraining.py – generates/trains the ML model using preprocessed data.
• Main.py – orchestrates end-to-end pipeline in a general (possibly desktop) environment.
• requirements.txt – Python dependencies for development.
RaspberryPi/ directory:
• RaspberryInterface.py – interfaces with GPIO or network for alerts/mitigation.
• Functions.py – likely a subset of processing routines tailored to the Pi.
• requirements.txt – runtime dependencies for the embedded device.
README for additional insights
• Purpose: protect IoT or vehicles from RF jamming by detecting anomalous signal patterns.
• Architecture: passive detection only; integrates with other systems to handle mitigation.
• Hardware: Raspberry Pi running Python + RTL-SDR dongle.
• Software stack: Python + ML libraries (e.g., scikit-learn or TensorFlow), SDR libraries (e.g., pyrtlsdr), GPIO/network modules.