5G Core Network Traffic Analysis

Overview

This project provides a comprehensive analysis of 5G Core Network traffic using real-world datasets. The primary goal is to extract, visualize, and summarize key network metrics and communication patterns from packet capture data, supporting research and education in next-generation mobile networks.

The analysis uses the 5G-Core-Networks-Datasets, which contains packet captures and CSV exports from a 5G core network emulation. The project generates summary statistics, protocol distributions, network topology graphs, and time-series visualizations to help understand the behavior and structure of 5G core network traffic.

---

Dataset

We use the 5G-Core-Networks-Datasets, which includes:

• Dataset1.pcapng: A packet capture file containing initial UE registration and over two hours of 5G core network operations.
• Dataset2.csv: A CSV export with selected fields per packet (sequence number, timestamp, source/destination IP, protocol, length, and info).

The dataset captures traffic among various 5G core network functions (AMF, SMF, UPF, etc.) and user equipment, providing a realistic view of network interactions.

---

Methods

Data Extraction

• PCAP Processing: The original implementation in main.py uses pyshark to analyze the .pcapng file directly. However, due to the large size of the dataset, this approach is time-consuming.
• Optimized Analysis: To address performance issues, we use temp.py, which extracts a representative sample from the .pcapng file using tshark and processes the resulting CSV for faster analysis.

Analysis Steps

1. Traffic Pattern Analysis: Compute total packets, unique IPs, protocol distributions, and top talkers.
2. Network Topology Visualization: Build a directed graph of network communications and visualize the most significant connections.
3. Packet Length Analysis: Plot the distribution of packet sizes and analyze by protocol.
4. Time-Series Analysis: Visualize packet activity over time and by protocol.
5. Conversation Patterns: Identify common communication flows and protocol sequences between network functions.
6. Summary Report Generation: Output a detailed summary of findings and visualizations.

All generated plots and the summary report are saved in the output_metrics/ directory.

---

Usage

1. Install Dependencies
   Ensure you have Python 3.x and install required packages:

   pip install pandas matplotlib seaborn networkx pyshark
   

2. Download the Dataset
   Clone or download the 5G-Core-Networks-Datasets repository and place it in the project directory.

3. Run the Analysis
   Execute the optimized analysis script:

   python temp.py
   

   This will generate summary statistics, plots, and a report in output_metrics/ and as 5g_network_analysis_summary.txt.

---

Project Structure

• main.py: Original analysis script (slower, direct .pcapng processing).
• temp.py: Optimized analysis using CSV extraction for faster results.
• output_metrics/: Directory for generated plots and visualizations.
• 5g_network_analysis_summary.txt: Summary report of the analysis.
• 5G-Core-Networks-Datasets/: Directory containing the dataset.

---

MidSem Solution Videos

• Harshit Agrawal (202251053) - drive link
• Malaika Varshney (202251069) - drive link
• Ameya Naik (202251012) - drive link

---

Educator

Dr. Bhupendra Kumar

---

Collaborators

• Harshit Agrawal (202251053)
• Malaika Varshney (202251069)
• Ameya Naik (202251012)