In many ways, the principles behind an open, multivendor RAN have been pioneered in the small cell environment. The small cell ecosystem has always supported new and established vendors together, and SCF has developed a range of open interfaces within the RAN, including FAPI (functional API), internal to RAN products, and Network FAPI, which defines a fronthaul connection between radio units and functional baseband units.
The industry is now moving towards these open and disaggregated architectures in every part of the RAN, and many organizations, including SCF and the O-RAN Alliance, are working together to address the challenges of developing and deploying a RAN that is not only multivendor but in many cases virtualized, with some or all of the baseband elements (distributed unit and centralized unit) running on cloud infrastructure. There are many lessons to be learned from the early experiences of open networking in the small cell world, including ways to simplify roll-out, management, and automation. SCF is distilling this experience into operator requirements and blueprints to help lower cost and risk for early adopters (How blueprints unlock the potential of 5G networks.) Many of the early commercial open RANs have been deployed as small cell networks, often for enterprise or private network scenarios. This is because such networks are often greenfield, are usually localized, and do not need very high complexity/cost technologies such as Massive MIMO. All these factors mean small cell networks are often a relatively low-risk environment in which to learn about Open RAN while targeting some immediate new revenues, whether the deployer is a conventional MNO or a new player such as a neutral host or private network operator.
There is no one solution that will fit all requirements. 3GPP has defined a large number of ways to split RAN functions between two or three units (Radio Unit RU, Distributed Unit DU, and Centralized Unit CU), and SCFโs annual survey of deployers shows there is market need for a choice of approaches. 3GPP has standardized split 2 and SCF has pioneered split 6, which places the Layer 1 network functions (the most processor-intensive) on the RU, while O-RAN Alliance is working on split 7.2x, in which some of the Layer 1 functions run in the DU. In the spirit of true and complete openness, SCFโs split 6 work (i.e. (n)FAPI work) strives to ensure that the control plane, user plane, and management plane are all fully open to support multiple hardware and software offerings in each plane.
Our survey of 113 deployers highlighted the need for both the split 6 & 7.2 options (and others such as Split 8 and Split 2). And many roll-outs will still focus on integrated small cells, which in some networks can provide the simplest and most power-efficient approach. 44% of deployers expect to support three or more splits by 2027, across their various networks (whether macro or small cell), according to our survey.
The functional split is not the only choice to be made. Whether to collocate the RU and DU, or DU and CU, is another deployment decision, as is potential integration with edge compute nodes and the type of fronthaul connection. In each case, the options chosen must relate directly to the applications and use cases that are most important to the deployer.
The SCF survey indicated that Split 6, for instance, is the preferred goal in enterprise and indoor networks, while Split 7.2x is more preferred for urban macro RANs.
There is a huge diversity of environments, applications, deployers, and commercial models enabled by dense 5G networks. That inevitably means there must be a diversity of network architectures, but this raises the danger of fragmentation, which in turn would deprive vendors of the scale that they need to enter the market. The solution is to ensure operators or integrators can select multiple different approaches while innovating on a common platform.
This has been a central tenet of SCFโs work on open small cell networks. FAPI takes open interoperability right down to the components on a system-on-chip, while nFAPI implements Split 6-based fronthaul in a way that provides a wide choice of fronthaul transport technologies, including enterprise Ethernet (5G FAPI_PHY API Specification.)ย But this is not a simple Lego set of options โ choosing and integrating them is complex and can add to the cost of deployment. The cost of integration can be particularly serious in a small cell network where TCO is often a critical consideration.
SCF supports several ways to ease this burden, providing tools for assessing the impact of each architecture choice on the chosen use case, deployment blueprints, and support for plugfests and testing. SCF has published reports assessing the impact of some of the key factors that influence performance in different use cases, including fronthaul bandwidth, latency requirements, and how complex the RU and/or DU can be. (See – Options for split RAN: Flexibility within a common framework)
It also provides a ground-breaking tool for making quantitative evaluations of these network options. This is SCF DARTs, a calculator that enables organizations to plan, design, and budget for transport networks for open RANs. (DARTs – An analysis tool for Disaggregated RAN Transport). In this way, operators can use objective tools and data to assess the best options for their use case and implement their chosen architecture within open frameworks to allow for future-proofing and future changes.
Reference designs are also essential for product developers, who have to make their own decisions about which options to prioritize and create product lines that are flexible enough to support multiple options cost-effectively. For instance, there is an emerging demand for multiband enterprise small cells that support 4G and 5G and allow existing LTE networks to be migrated to 5G vRAN over time.
To build operator confidence, the resulting solutions must be proven to be open and interoperable without the deployer having to invest in significant custom testing. Open testing and certification is the key to many successful standards and will be crucial for Open RAN to gain scale. SCF, O-RAN Alliance, and Telecom Infra Project (TIP) have been heavily focused on supporting standard, industry-wide testing processes, including participation in plugfests. Robust certification of compliance with Open RAN standards will be a critical success factor, especially for enterprise small cell networks where deployers will be buying large quantities of cells but will typically not have in-house interoperability testing capabilities.
Collaboration between stakeholders and between all the industry alliances focused on open networks will be important to accelerate progress and share best practices. SCF works with O-RAN Alliance, 3GPP, Open Air Interface (OAI), and many others in technical, commercial, and regulatory dimensions. These cooperations help to drive confidence and a vibrant ecosystem but also educate the whole community of potential deployers. Open systems, coupled with the emergence of open source code and shared spectrum, can enable a host of new deployers, especially in enterprise and smart city spaces where small cells are essential.
These new operators, alongside the established MNOs, can accelerate the availability of high-performance cellular connectivity for every scenario. However, they lack the long years of cellular experience of the MNOs, and so education and confidence building will be as important to the work of industry associations as specifications and certifications.
With a combination of all these efforts, technical and commercial proof points can be quickly established in small cells, and this knowledge will help to foster the adoption of Open RAN in all domains, regardless of use case and architecture.ย That, in turn, will result in significant growth in the deployment of open vRAN architectures, which will account for over half of the small cell installed base by 2028, according to SCFโs latest forecast (Figure 2), with 94% CAGR between 2021 and 2028.
