Sure and steady wins the race: why vRAN is an evolution, not a revolution
Disaggregated RAN – or vRAN – is quickly gaining momentum among OEMs and others in the telecoms industry, and has arguably reached critical mass. Nowhere was this clearer to the VIAVI team than during the recent TIP Summit in Amsterdam, the participants at which included a significant percentage of first-time attendees.
It was also interesting to see an increasingly diverse representation, with service and technology providers, systems integrators, start-ups, investors, analysts and other ecosystem partners all coming together to discuss developments in (among other things) virtualised RAN functions.
The interest and investment in vRAN may be accelerating, but robust models of success have been somewhat slower to emerge.
Mobile network operators, including a growing number of Tier 1s, are demanding support for disaggregated RAN as a mandatory element for new deployment activities. Take big-hitter Vodafone, which recently put its entire European RAN footprint out to tender, comprising 100K cell sites across 14 markets. The catch? The operator is only accepting solutions built on disaggregated RAN components and will only consider proposals from vendors which adhere to Open RAN models.
Facebook likes vRAN
It’s not just the operator community which is moving to vRAN, either. Venture capitalists and investors in vertical sectors are helping to bring disaggregated RAN/vRAN-based solutions to the market. Among these is Facebook, whose interest in a technical telecoms topic like vRAN will come as a surprise to many of the social media platform’s users.
Consider that Facebook’s business is dependent on health of the mobile network ecosystem, and it believes the health of the ecosystem is dependent on developing innovative approaches like vRAN. Facebook staked out its vRAN territory back in 2016 when it launched the Telecom Infra Project with a number of partners, with the aim of disaggregating the components of network infrastructure.
Even with the might of Facebook behind it, the challenge is by no means a small one, especially due to the nature of the scenarios that stakeholders are aiming to find solutions for: high traffic, high reliability and high complexity models. Facebook is but one example, with many other parties coming across similar stumbling blocks.
Unfortunately, there’s little prior groundwork which could help overcome challenges. Historically, interworking on RAN interfaces has involved scheduling, baseband, and IQ traffic between equipment vendors, a very difficult topic which has seen very limited success.
Japan: setting the standard?
A notable exception is the work of NTT DoCoMo. In March 2016, the Japanese mobile operator announced the first commercial deployment of a multi-vendor, interoperable virtualised EPC (Evolved Packet Core). Many people attribute the success of the CPRI interworking project to the strict control and management that NTT DoCoMo applied, particularly in the areas of development and testing.
We’ve also seen more recent successes, including the launch of Rakuten in June this year. Rakuten Mobile, the mobile network subsidiary of Rakuten Group, along with integrator NEC Corporation, created the first 5G open vRAN architecture in Japan, setting an example for others to follow.
What’s behind this success?
In part, it is due to operators in a green field scenario being able to work within tight specifications and a short timespan to deliver a network which meets the quality expectations of Japan. An additional driving force was the fact that both Rakuten and investor (and participant) TechMahindra took ownership stakes in equipment vendor (and participant!) Altiostar, and had a very significant and rather obvious stake in the project’s success.
The disaggregated vRAN ecosystem has still a lot to prove. Despite this, legacy vendors aren’t shying away from the challenge, and attempting to surmount it with innovations of their own. Proof points may be scarce, but they do exist: service providers such as Rakuten and Telefonica have deployed vRAN at scale, and shared the lessons learned with the rest of the TIP community.
Running the vRAN race
The question of when will we see robust vRAN architectures widely deployed and delivering ROI is a tricky one. Yet we do know what can be done to accelerate development and help reach that point (whenever it happens to be).
First, engineers must feel confident and secure in their ability to fail in the lab. Learning lessons in this environment is crucial and allows engineering teams to identify and rectify issues while protecting the users of the network from any fallout.
Second, using a continuous integration/continuous delivery (CI/CD) approach is also important. Facebook noted that while CI/CD was easier in its closed system, the additional complexity this introduced for the telco community highlighted the need for telcos to take greater responsibility for end-to-end integration and operation of the network. Employing a CI/CD approach will allow engineers to the develop and deploy small, incremental changes to the network.
The move to vRAN will take time and cross-industry effort, but sure and steady will win the race. When it comes to vRAN, think evolution, not revolution.