shuji-oh / similarity_can_ids

"Normal and Malicious Sliding Windows Similarity Analysis Method for Fast and Accurate IDS against DoS Attacks on In-Vehicle Networks" IEEE Access 2020

similarity_CAN_IDS: Normal and Malicious Sliding Windows Similarity Analysis Method for Fast and Accurate IDS against DoS Attacks on In-Vehicle Networks

Overview

The similarity_CAN_IDS is a state-of-the-art DoS attacks detection method on CAN bus.

Shuji Ohira, Araya Kibrom Desta, Ismail Arai, Hiroyuki Inoue, Kazutoshi Fujikawa, "Normal and Malicious Sliding Windows Similarity Analysis Method for Fast and Accurate IDS against DoS Attacks on In-Vehicle Networks.", IEEE Access, Vol.8, pp.42422-42435, 2020.

@article{ohira2020similarity_based_IDS,
  title={Normal and Malicious Sliding Windows Similarity Analysis Method for Fast and Accurate IDS against DoS Attacks on In-Vehicle Networks},
  author={Ohira, Shuji and Araya, Kibrom Desta and Arai, Ismail and Inoue, Hiroyuki and Fujikawa, Kazutoshi},
  journal={IEEE Access},
  volume={8},
  pages={42422--42435},
  year={2020},
  publisher={IEEE}
}

Description

We demonstrated that an entropy_CAN_IDS (conventional method) cannot detect an entropy-manipulated attack in which an adversary adjusts the entropy of a DoS attack to a normal value. Thus, we proposed the similarity_CAN_IDS that is a state-of-the-art detection method on the CAN bus. The proposed method use not entropy but similarity to detect intrusion detection.

Directory Structure

similarity_CAN_IDS
┣━ off-line_learning_phase
┃ ┣━ README.md
┃ ┣━ output_similarity.py
┃ ┣━ eval_similarity_CAN_IDS.py
┃ ┗━ output_params.py
┣━ on-line_detection_phase
┃ ┣━ similarity_CAN_IDS.c
┃ ┣━ Makefile
┃ ┣━ lib.c
┃ ┣━ lib.h
┃ ┗━ terminal.h
┣━ CIDs.txt
┣━ optimazed_params.txt
┗━ README.md

Requirement

python3, gcc

Usage

$ git clone https://github.com/shuji-oh/similarity_CAN_IDS  
$ cd similarity_CAN_IDS/off-line_learning_phase  
$ python3 output_params.py ../test_data/test_data.log  
$ cd ../on-line_detection_phase/  
$ make  
$ ./similarity_CAN_IDS can0  

Contribution

Author

shuji-oh

similarity_can_ids's Issues

Improve indent of python scripts in off-line learning phase

Extending for 29-bit ID CAN