Sim2Field: End-to-End Development of AI RANs for 6G

TL;DR

The paper addresses the gap between simulation-based AI wireless research and field performance by introducing Sim2Field, a Digital Twin–driven methodology for site-specific training of AI PHY components. It combines ray-traced channels, domain randomization, and receiver impairment modeling to generate realistic training data, mitigating the reality gap. An end-to-end prototype on the ARC-OTA platform demonstrates real-time AI uplink channel estimation integrated into NR PHY via CUDA-graph accelerated cuBB, achieving substantial throughput gains over conventional MMSE in both channel emulation and OTA experiments. The results support the practical potential of AI in the PHY layer and underscore the importance of realistic, field-oriented training data for reliable deployment in 5G/6G networks.

Abstract

Following state-of-the-art research results, which showed the potential for significant performance gains by applying AI/ML techniques in the cellular Radio Access Network (RAN), the wireless industry is now broadly pushing for the adoption of AI in 5G and future 6G technology. Despite this enthusiasm, AI-based wireless systems still remain largely untested in the field. Common simulation methods for generating datasets for AI model training suffer from "reality gap" and, as a result, the performance of these simulation-trained models may not carry over to practical cellular systems. Additionally, the cost and complexity of developing high-performance proof-of-concept implementations present major hurdles for evaluating AI wireless systems in the field. In this work, we introduce a methodology which aims to address the challenges of bringing AI to real networks. We discuss how detailed Digital Twin simulations may be employed for training site-specific AI Physical (PHY) layer functions. We further present a powerful testbed for AI-RAN research and demonstrate how it enables rapid prototyping, field testing and data collection. Finally, we evaluate an AI channel estimation algorithm over-the-air with a commercial UE, demonstrating that real-world throughput gains of up to 40% are achievable by incorporating AI in the physical layer.

Figures (8)

• Figure 1: Platform for rapid AI wireless innovation.
• Figure 2: Sim2Field training data generation procedure.
• Figure 3: Example ResNet CE model architecture.
• Figure 4: Aerial ARC-OTA testbed.
• Figure 5: AI-CE cuBB (CUDA) implementation.