The first virtual RANs (vRANs) are live and providing good service to customers. The focus now is on how the technology ecosystem can innovate to deliver the system performance and operational capability needed to extend open vRAN into more deployment scenarios over the next few years.

To deploy and operate open vRANs at commercial scale – for example, across a city, a region or nationwide – involves an end-to-end redesign of the classic RAN operations models. It also requires a new generation of vRAN software and silicon to meet the demands of large national networks with millions of customers.

A new Heavy Reading white paper – Scaling Open RAN Deployment & Operations – discusses how to take best practices from traditional RAN and cloud infrastructure operations to create a scalable deployment model for a disaggregated RAN. The paper discusses site deployment, bare metal server hardware and the requirement for pre-integrated "workload-ready" infrastructure stacks that can be rapidly deployed across tens of thousands of cell sites.

Individual mobile operators are among the largest network providers in any given country. A larger European operator, for example, typically runs 20,000+ macro cell sites and several thousand outdoor small cells. In the US, national operators typically run 80,000+ macros and tens of thousands of outdoor small cells.

Each cell site hosts multiple pieces of active networking equipment that must be provisioned and managed remotely. RAN deployment tools and workflows must therefore operate at large scale. Moreover, in open RAN, instead of discrete hardware appliances from a single vendor, automation tools must manage diverse multi-vendor hardware, software infrastructure and RAN functions. This is no small task, and until now, it has been an under-addressed challenge for open vRAN.

An important part of the solution is the creation of declarative blueprints to prepare infrastructure nodes based on commercial-off-the-shelf servers – which includes Layer 1 accelerator cards – and cloud software for RAN application software. By adopting a blueprint approach, with pre-integration and testing built-in to the model, operators can optimize the deployment and commissioning workflow.

Infrastructure lifecycle automation for vRANs should run from Day 0 through the commissioning and acceptance phases to ongoing monitoring and updating of the node. Software lifecycle management best practices can, relatively speaking, easily be applied to the RAN. For server hardware, however, the requirements of the RAN and the highly distributed deployment model introduce new challenges. For example, updating the firmware and BIOS settings of large numbers of remote commercial-off-the-shelf servers deployed in hard-to-access locations is not currently in the standard RAN operating model.

The payoff from disaggregation is the opportunity to introduce state-of-the-art cloud operating models and automation into the multi-vendor vRAN. This is a fundamental enabler of open RAN.

To download Heavy Reading's Scaling Open RAN Deployment & Operations white paper, click here.

— Gabriel Brown, Senior Principal Analyst, Mobile Networks & 5G, Heavy Reading

This blog is sponsored by Dell Technologies.