Real-Time Network Intrusion Detection Dataset: Campus Traffic and Behavioral Features

Description

This dataset is a collection used during my final year undergraduate project

Steps to reproduce

Download the Dataset Files Download the raw capture CSV files, generate_dataset.py, and requirements.txt to your local machine. Set up the Directory Structure Create a directory named "data/raw/" relative to where you place the Python script, and move the four primary capture files into it. Your folder structure must look exactly like this: generate_dataset.py requirements.txt data/raw/ultra_normal_92k.csv data/raw/ultra_attack_90k.csv data/raw/human_traffic_wireshark.csv data/raw/comsit_traffic.csv (Note: campus_attack_capture.csv is used for separate adversarial testing experiments and is not required to generate the primary engineered dataset). Install Dependencies Open a terminal in the directory containing generate_dataset.py and install the required Python packages (Python 3.9+ is recommended): pip install -r requirements.txt Generate the Engineered Dataset Run the self-contained reproducibility script: python generate_dataset.py Verify Output The script will read the four raw CSVs, apply a rolling-window behavioral feature extraction pipeline (extracting 12 features such as Inter-Arrival Time, packet sizes, and flow durations), and output the final merged dataset to: data/processed/engineered_realtime_dataset.csv (approx. 1.6M rows).

Institutions

• University of Ilorin

  Kwara State, Ilorin

Categories

Licence