TwinRAN: Twinning the 5G RAN in Azure Cloud

TL;DR

TwinRAN presents a cloud-based digital twin of the 5G RAN built on Azure DT and powered by the O-RAN architecture. By maintaining two concurrent twin graphs (intercell and intracell) and a purpose-built xApp to feed telemetry, it enables non-invasive, vendor-agnostic, multipurpose network management. The work provides a thorough evaluation of Azure DT latency and demonstrates three concrete use cases—data warehousing for ML, precise positioning for handovers, and what-if scenario exploration—along with a cost analysis. Despite promising feasibility, the study notes current cloud-pricing and latency limitations that constrain continuous, large-scale deployment, suggesting targeted, time-bounded use as a practical path forward.

Abstract

The proliferation of 5G technology necessitates advanced network management strategies to ensure optimal performance and reliability. Digital Twin (DT)s have emerged as a promising paradigm for modeling and simulating complex systems like the 5G Radio Access Network (RAN). In this paper, we present TwinRAN, a DT of the 5G RAN built leveraging the Azure DT platform. TwinRAN is built on top of the Open RAN (O-RAN) architecture and is agnostic to the vendor of the underlying equipment. We demonstrate three applications using TwinRAN and evaluate the required resources and their performance for a network with 800 users and eight gNBs. We first evaluate the performance and limitations of the Azure DT platform, measuring the latency under different conditions. The results from this evaluation allow us to optimize TwinRAN for the DT platform it uses. Then, we present the system's architectural design, emphasizing its components and interactions. We propose that two types of twin graphs be simultaneously maintained on the cloud: one for intercell operations, keeping a broad overview of all the cells in the network, and another where each cell is spawned in a separate Azure DT instance for more granular operation and monitoring of intracell tasks. We evaluate the performance and operating costs of TwinRAN for each of the three applications. The TwinRAN DT in the cloud can keep track of its physical twin within a few hundred milliseconds, extending its utility to many 5G network management tasks, some of which are shown in this paper. The novel framework for building and maintaining a DT of the 5G RAN presented in this paper offers network operators enhanced capabilities, empowering efficient deployments and management.

Figures (6)

• Figure 1: Sequence Diagram of the messaging in TwinRAN: The TwinRAN xApp uses the O-RAN indication messages from the Near-RT RIC and updates the DT hosted on the Azure cloud. The desired KPMs have to be subscribed to by the xApp via the Near-RT RIC.
• Figure 2: Digital Twins on the Azure Cloud: In TwinRAN, the inputs to the DT instances hosted in a live execution environment are via the TwinRAN xApp. For every instance, the Twin graphs keep the twins and their relationship updated. The output of the DTs is then connected to other Azure cloud services for post-processing or recording.
• Figure 3: Measurement setup for the lag and query latency of Azure DT. The update messages are timestamped at three points. We vary model sizes and update sizes to ascertain the behavior of Azure DT.
• Figure 4: Service times, lag, and query latency measurements for different DT models and update sizes. The service times scale with not only the size of the model but also the size of the update message. The measurements' mean values exhibit a linear trend, and the distributions have a long tail.
• Figure 5: Two different twin-graphs are maintained : (a)TwinRAN-multi: a single instance keeps track of the entire network. One DTDL model defines one cell. The neighboring cells are connected by relationships, mainly defining the amount of interference caused by the neighbor. (b) TwinRAN-cell, where one instance is spawned for every cell. The gNB connects to each UE via a relationship.