A Distributed Denial-of-Service (DDoS) attack is a malicious attempt to disrupt the normal traffic of a targeted server, service or network by overwhelming the target or its surrounding infrastructure with a flood of Internet traffic.
There are different ways to prevent these attacks, but none of them can be cent-percent accurate, our motive is to design a firewall for avoiding DDoS attacks. For that, accurate detection of the attack is essential.
In this proposed system, we are using multiple machine learning-based algorithms to detect DDoS attack.
Clone the project
git clone https://github.com/MidanAhmed/ddos-detection.gitOpen the project directory in Jupyter Notebook, the following files must be present:
main.ipynbgenerateCSV.ipynbgui.ipynbdataset.csv
Run the generateCSV.ipynb file to split dataset.csv to different sections and generate following files:
data10000.csvdata12000.csvdata14000.csv
These files are used for testing purposes.
Once the sectioned datasets have been generated, run main.ipynb to train the model on the dataset, namely dataset.csv.
The model will be then trained on a particular machine learning algorithm and a .joblib file will be created.
The following files will be created for the respective algorithms:
| Trained algorithm | File generated |
|---|---|
| Logistic Regression | logisticr.joblib |
| Decision Tree Classifier | decisiont.joblib |
| Random Forest Classifier | randomf.joblib |
| Gaussian Naive Bayes | naiveb.joblib |
Start the Graphical User Interface (GUI) by running gui.ipynb and selecting the file to be tested.
This project has been implemented in the Jupyter. Major python libraries involved in faster implementation are listed below:
- numpy
- pandas
- scikit-learn
- tkinter
I would like to take this opportunity to express my gratitude to my project guide Mr. Jainul Abudin and all of my group members Kishor Kumar Hazarika, Ruhul Amin Talukdar and Zakaria Ahmed Laskar. This project would not have been successful without their cooperation and inputs.