|
|
--- |
|
|
license: apache-2.0 |
|
|
datasets: |
|
|
- SilverDragon9/UNSW_TON-IoT_Train_Test_IoT_Datasets |
|
|
- SilverDragon9/UNSW_TON-IoT_Train_Test_OS_Datasets |
|
|
language: |
|
|
- en |
|
|
- el |
|
|
metrics: |
|
|
- accuracy |
|
|
- f1 |
|
|
- precision |
|
|
tags: |
|
|
- IOT |
|
|
- CyberSecurity |
|
|
- Intrusion |
|
|
- Detection |
|
|
- IDS |
|
|
library_name: sklearn |
|
|
--- |
|
|
|
|
|
--- |
|
|
## 🌐 Overview |
|
|
|
|
|
**Sniffer.AI** is an AI-powered Intrusion Detection System (IDS) for **IoT networks**, designed to detect and classify suspicious behavior across smart devices in real-time. /n |
|
|
Built on ensemble machine learning models trained on the **UNSW TON_IoT dataset**, it classifies activity into `Normal` or one of **7 attack types**. |
|
|
|
|
|
> 📡 Target Devices: Fridge, GPS Tracker, Garage Door, Thermostat, Weather Station |
|
|
> 📁 Output can be saved for **offline analysis and archiving** |
|
|
|
|
|
|
|
|
## 📦 Key Features |
|
|
|
|
|
| Feature | Description | |
|
|
|----------------------------------|-------------| |
|
|
| 🧠 Ensemble Models | RF, XGBoost, AdaBoost, Bagging, Decision Trees | |
|
|
| 🧪 Predicts Threat Category | Normal vs 7 Attack Types | |
|
|
| 🕒 Timestamps Every Detection | Provides real-time date & time in output | |
|
|
| 💾 Downloadable Results | Output can be saved as `.csv` or `.json` | |
|
|
| 🌐 Edge Ready | Lightweight enough for IoT Gateway deployment | |
|
|
| 📚 Dataset Used | [UNSW TON_IoT](https://research.unsw.edu.au/projects/toniot-datasets) | |
|
|
|
|
|
## 🔍 Attack Categories |
|
|
|
|
|
- text |
|
|
- Normal |
|
|
- Backdoor |
|
|
- DDoS |
|
|
- Injection |
|
|
- Password Attack |
|
|
- Ransomware |
|
|
- Scanning |
|
|
- XSS |
|
|
|
|
|
## 📊 Sample Output Format |
|
|
|
|
|
| Date | Time | Prediction | |
|
|
|--------------------|----------------|----------------| |
|
|
| 2025-04-11 | 14:35:22 | Scanning | |
|
|
|
|
|
--- |
