GNSS SHIELD v3.0 — Spoofing & Jamming Detection

Full-stack deep learning dashboard with real trained models, 5 satellite canvas animations, and secure login authentication.

Project Structure

gnss-shield-v3/
├── backend/
│   ├── app.py               ← Flask API + custom Keras layers + auth
│   ├── requirements.txt
│   ├── models/              ← Your 5 trained .keras models (included)
│   │   ├── cnn_baseline.keras
│   │   ├── cnn_lstm_hybrid.keras
│   │   ├── autoencoder_model.keras
│   │   ├── Attention_CNN_CBAM.keras
│   │   └── senet_cnn.keras
│   └── uploads/
└── frontend/
    ├── src/
    │   ├── App.js
    │   ├── components/
    │   │   ├── LoginPage.js          ← Animated orbital login
    │   │   └── GNSSSpoofingDetector.js ← Full dashboard
    └── package.json

Quick Start

Backend

cd backend
pip install -r requirements.txt
python app.py
# Runs at http://localhost:5000
# Loads all 5 models automatically

Frontend

cd frontend
npm install
npm start
# Runs at http://localhost:3000

Login Credentials

Role	Username	Password
Admin	admin	gnss2024
Analyst	analyst	analyst123

To change: edit USERS dict in backend/app.py, update hash with:

import hashlib
print(hashlib.sha256("new_password".encode()).hexdigest())

Models (all .keras files included)

Model	File	Accuracy
CNN Baseline	cnn_baseline.keras	98.73%
CNN-LSTM Hybrid	cnn_lstm_hybrid.keras	97.24%
Autoencoder	autoencoder_model.keras	85.50%
CBAM (Best)	Attention_CNN_CBAM.keras	99.42%
SE-Net CNN	senet_cnn.keras	99.15%

The CBAM and SE-Net models use custom Keras layers (ChannelAttention, SpatialAttention, SEBlock) defined directly in app.py — no additional files needed.

API (all protected with Bearer token)

POST /api/login         — Get auth token
POST /api/logout        — Invalidate token
GET  /api/health        — Backend status + loaded models
GET  /api/models        — List all models
POST /api/predict       — Analyze .json or .csv file
GET  /api/predict/demo  — Demo prediction

Input Format

Upload .json GNSS observation files with keys: cn0_G1, doMes_G1, prMes_G1, etc. Or upload .csv with tabular features.

---