Deployment, and Monitoring of a Private 5G Standalone Core Network Using Open5GS**

Introduction

As part of my Bachelor’s degree in Electrical and Electronic Engineering (Telecommunications) from University of Tripoli, I developed a graduation project focused on building a private 5G Standalone (SA) core network.

The objective of this project was to design, deploy, and analyze a fully functional 5G core using open-source tools, while demonstrating key concepts such as control and user plane separation, scalability, and real-time performance monitoring.

This project was successfully completed with a final grade of 92% (Excellent).

---

Project Objectives

The main goals of this project were:

• Build a virtualized 5G Standalone core network
• Implement distributed User Plane Functions (UPFs)
• Simulate real-world user behavior using multiple UEs
• Measure key performance indicators (KPIs)
• Visualize network performance in real time

---

System Architecture

The system was designed using a 5 Virtual Machine (VM) architecture, where each component plays a specific role:

• VM1: Open5GS Core (AMF, SMF, NRF, AUSF, UDM, PCF)
• VM2: UPF1
• VM3: UPF2
• VM4: UERANSIM (gNB + UE simulation)
• VM5: Monitoring (Prometheus + Grafana)

This architecture enables separation between control plane and user plane, improving scalability and flexibility.

---

Implementation

The project was implemented in structured stages:

1. Planning and Design

• Defined KPIs (throughput, latency, jitter, packet loss)
• Designed network interfaces (N2, N3, N4, N6)
• Planned IP addressing and topology

2. Environment Setup

• Created and configured 5 virtual machines
• Assigned static IPs and verified connectivity

3. Control Plane Deployment

• Installed Open5GS
• Configured AMF and SMF
• Registered network functions with NRF

4. User Plane Configuration

• Deployed two UPFs for load balancing
• Configured GTP-U, PFCP, and routing
• Set up OGSTUN interface for UE IP allocation

5. RAN and UE Simulation

• Used UERANSIM to simulate gNB and UEs
• Configured subscriber database (500 users)
• Verified successful UE registration

6. PDU Session Establishment

• Established data sessions between UE and network
• Verified IP allocation and traffic flow

7. Traffic Testing

• Used iperf3 to generate TCP and UDP traffic
• Measured throughput, latency, jitter, and packet loss

8. Monitoring and Visualization

• Deployed Prometheus and Grafana
• Built dashboards for real-time KPI monitoring

---

Results and Analysis

The project produced several important results:

🔹 Throughput Performance

• Achieved up to ~994 Mbps using dual UPFs
• Reduced to ~704 Mbps when only one UPF was active

🔹 Security Algorithm Impact

• Standard algorithms: ~11 seconds registration time
• High-performance algorithms (NIA2/NEA2): ~6 seconds

🔹 Network Slicing

Two slices were implemented:

• eMBB: High throughput
• URLLC: Low latency

Results showed:

• Improved throughput for eMBB
• Reduced latency for URLLC

---

Technologies Used

• Open5GS
• UERANSIM
• Prometheus
• Grafana
• Linux (Ubuntu)
• VMware ESXi
• Networking protocols (TCP/IP, GTP-U, PFCP)

---

Key Learnings

Through this project, I gained practical experience in:

• 5G core network architecture
• Virtualized network deployment
• Network performance analysis
• Monitoring and observability tools
• Troubleshooting complex distributed systems

---

Conclusion

This project successfully demonstrated the design and deployment of a private 5G Standalone core network with distributed user plane architecture.

It highlights the importance of scalability, performance optimization, and real-time monitoring in modern telecommunications systems.

The experience gained from this project provides a strong foundation for working with next-generation mobile networks and cloud-based telecom infrastructure.