DistillGuard: Comparative Study on IoT Intrusion Detection

This repository contains the implementation and results of the DistillGuard framework (Knowledge Distillation for IoT Security) applied to three major datasets:

1. ToN-IoT (Heterogeneous)
2. RT-IoT (Real-Time)
3. N-BaIoT (Botnet Traffic)

Project Structure

• DistillGuard-ToN-IoT/: Implementation for ToN-IoT.
• DistillGuard-RT-IoT/: Implementation for RT-IoT.
• DistillGuard-N-BaIoT/: Implementation for N-BaIoT.

Key Results

Dataset	Teacher Acc	Student Acc	Compression	Robustness
ToN-IoT	96%	95.9%	4x	Low
RT-IoT	99%	99.5%	4x	Low
N-BaIoT	17% (Fail)	82.6% (Success)	4x	Low

See DistillGuard-N-BaIoT/comparison_study.pdf for the full report.

Usage

Each directory contains its own README.md and scripts:

1. data_processor.py
2. train_distillguard.py
3. evaluate.py

Requirements

• PyTorch
• Pandas
• Scikit-Learn
• Matplotlib