DDoS Attack Detection for 5G Network Slice using CNN
This project provides a solution for detecting Distributed Denial of Service (DDoS) attacks in 5G network slice simulated data using a Convolutional Neural Network (CNN). The implemented model achieves a high accuracy of more than 99% in accurately identifying DDoS attacks.
Data exploration, cleaning, transformation, and other data mining steps are followed here. Data set is also generated by our team for this project which is around 10 Million rows in total. The dataset used for training and testing the DDoS detection model is available at IEEE DataPort - DoS/DDoS Attack Dataset for 5G Network Slicing. You can download the dataset and upload the .csv files to your Google Drive for easy access in the Colab environment.
Each of the details of this model is described in the Google Colab file
The notebook provides detailed results, including accuracy metrics, precision, recall, and F1-score, allowing to assess the effectiveness of the DDoS detection model. In eash of the cases the result is more than 99%. Heatmap of the detectoin of attack and normal traffic is given below:
