Model-Based Intrusion Detection System for CAN Protocol Anomaly Detection

This project aims to create a model-based intrusion detection system (IDS) for Controler Area Networks (CAN)

Getting Started

Installation

1. Clone the repository
2. Install the required Python packages

pip install -r requirements.txt

Functionality

Checkout code/presentation.ipynb to get an idea of the functionalities of the code base.

Dataset Download

1. Create a data directory in the project root.
2. Create a .env file in the project root and add the full path of your data directory. (Sometimes jupyter notebooks has issue with "../")

DATA_PATH=DATA-DIR-PATH-GOES-HERE

3. Use CANDataset to download and process the dataset. The data will be processed and saved to parquet files and will be pulled from these subsequently. Example use:

from data_prep.CANDataset import CANDataset
from dotenv import load_dotenv

load_dotenv()
data_path = os.getenv('DATA_PATH')
dataset = CANDataset(data_path, log_verbosity=1)