A 5G NR Digital Twin: Ray-Traced Channel Emulation and UL-TDoA Positioning in OpenAirInterface

OpenAirInterface (OAI) provides a complete, standard-compliant 5G protocol stack, while ray tracers such as NVIDIA Sionna RT produce physically consistent, site-specific channel impulse responses (CIRs). Coupling these two technologies paves the way for digital twins in wireless communications, enabling realistic simulations that evaluate real-world system performance without physical over-the-air deployment. This work proposes a novel channel-injection architecture that substantially reduces the CIR update period, allowing the system to track the channel's dynamic evolution instead of treating it as frozen between refreshes. The engine redesigns the Open Wireless Digital Twin (OWDT) refresh mechanism around bounded, blocking, thread-safe queues with ownership-transfer semantics and fully separated transmit/receive states. This architecture advances the CIR trajectory in lockstep with the radio slot clock, ensuring race-free execution by construction and enabling simulation speeds constrained only by the underlying hardware capabilities. Validation is performed over OAI's rfsimulator using commodity hardware. CIRs are generated offline from a 3D urban scene and convolved per slot inside the OAI gNB through this engine. Furthermore, the architecture extends to positioning applications: a single UE transmission simultaneously reaches four fixed anchors, each characterized by its own ray-traced link. The resulting Sounding Reference Signal (SRS) measurements are transported via NRPPa and resolved into a position estimate by the OAI Location Management Function (LMF). The modifications applied to both the OAI RAN and LMF components enable the standard uplink time difference of arrival (UL-TDoA) chain to run end-to-end, returning accurate location estimates. The final result is a functional, portable digital twin for 5G NR UL-TDoA positioning that can be executed entirely on a standard consumer-grade laptop.