Open-Source 5G Digital Twin: Virtualizing Core and RAN for Network Replication

This paper presents a proposed methodology for creating digital twins of 5G networks using open-source 5G core and Radio Access Network (RAN) implementations. Our research will demonstrate how virtualized network functions can be deployed to create accurate digital replicas of operational 5G infrastructures. We will evaluate three leading open-source 5G core implementations—Open5GS, free5GC, and OpenAir-Interface—assessing their performance characteristics and suitability for digital twin applications. The proposed architecture will specifically replicate a distributed multi-campus research network, featuring multiple gNodeB units at each location connected through IP transport networks to Compact Mobility Units (CMUs) that bundle core network functions. Our implementation will enable bi-directional synchronization between physical networks and their digital counterparts, facilitating testing, optimization, and AI/ML model training without disrupting operational environments. Performance benchmarks and implementation guidelines provided in this paper will contribute to the growing field of network digital twins, offering a costeffective approach for creating high-fidelity network replicas using open-source tools.