ChEmiNet: Multi-UE Channel Emulator for 5G SA Networks

Academic institutions and small-scale network operators face limited access to 5G network testing and development due to the restrictive and proprietary nature of existing cellular systems. Open-source cellular stacks are already available for software-based testing, but these platforms currently lack capabilities in emulating realistic network conditions for multiple-user scenarios. Hence, this project developed a multi-UE emulator integrated with statistical channel models to allow UMa, UMi, RMa, and indoor multi-user scenarios. A fully virtualized 5G SA network is implemented within a single machine using Open5GS, srsgNB, and srsUE from the srsRAN Project, with GNU Radio Companion to allow multi-user connections. This stack is integrated with 3GPP-compliant, python-based 5G channel models Sionna, PySim5G, and PyWiCh, where their characteristics and behavior under different topology scenarios are described using performance metrics bitrate, SNR, error vector magnitude, and latency. This low-cost and accessible alternative to proprietary testing environments holds the potential to empower a wider range of distributors by serving as a testbed to improve 5G network performance.