The Open RAN conversation has evolved from a narrow focus on the nuts and bolts of making a multi-vendor radio system work, to operationalizing these systems and layering in app-based automation processes. To the automation point, this is where the Open RAN RIC, or RAN Intelligent Controller, comes into play; the RIC comes in two flavors: non-real-time and near-real-time.
In a panel session at the recent Open RAN Global Forum, available on demand here, Core Analysis Founder and CEO Patrick Lopez described the non-real-time RIC as “an evolution of EMS, OSS, as well as the earlier implementations of self-organizing networks…The near-real-time RIC is really today a collection of functionalities that are embedded within the radio units, between the distributed unit or the centralized unit. And together the non-real-time RIC and near-real-time RIC promise to provide more control as well as some level of innovation within the RAN ecosystem.”
The RIC enables automation, service differentiation
So how are operators thinking about using the RIC and to what end? BT Distinguished Engineer Richard MacKenzie called the platform “the basic enabler for the Open RAN architecture. If we’re truly going to coordinate a multivendor end-to-end [system], then we need to have a common controller. So just to enable multi-vendor, the RIC is important…it’s an enabler for automation.”
MacKenzie continued: “And in the longer term, especially if we integrate it with AI engines, then we can do some serious optimization. But from an operator perspective, one of the other exciting opportunities is simply the opportunity to differentiate from our competitors. So by piecing together the right combination of xApps and rApps, we can actually build new services quickly and easily and it enables us to differentiate from other service providers more effectively.”
Managing complexity and gaining spectral efficiency with the RIC
Given the diversity of spectrum and RAN infrastructure necessary to deliver the right blend of 5G coverage and capacity, it’s complicated from a management perspective. The RIC can help here, according to US Cellular EVP and CTO Mike Irizarry.
“Because the networks are getting more complicated, not less, multiple layers, multiple frequencies, classic designs, some of the new disaggregated designs, and you can’t manage that complexity the old ways with processes that are rigid and more people. You need to apply data to AI and ML enabled automation to run the networks. So we’re very excited about the prospects of that capability.”
Speaking to the need to maximize spectrum investments—US Cellular spent around $2 billion on mid-band—Irizarry said, “We’re very excited about the possibility of these AI-enabled algorithms to further improve spectrum efficiency and improved power efficiency, lower our costs.”
Assurance from lab to live and balancing openness with standardizaiton
Along with the rise of Open RAN, there’s a renewed appreciation for the test and measurement and service assurance domains. Said simply, the more vendors put into a RAN, the more tests need to be done ahead of deployment, and the idea of introduction RAN applications via the RIC is another dimension that spotlights the need for reliable, hardened, automated test processes.
Viavi Regional CTO for EMEA Chris Murphy said the RIC is “a complex thing to test because you generally don’t have a network, a live network, which is representative of many scenarios ready to go. So having an emulation… and being able to make complex scenarios is a very powerful tool for being able to make sure that the system you’ve set up, that the RIC is doing what it should do, the apps it’s supporting do what they’re supposed to do and that they work together.”
On the nascent xApp and rApp space, Murphy said Viavi is working on its own applications and working with third parties. “The key here is that we’re asking the radio networks to deliver a wide variety of services, and we’re seeing metaverse services on the horizon which are going to have very strict quality of service requirements in order to do that efficiently. And with RIC efficiency and energy efficiency, we’re going to need to have apps which are able to be put together to deliver those mix of services that the operators have selected.”
Beyond the familiar machinations of 3GPP, Open RAN is also guided by the work of the O-RAN Alliance, and there’s increasingly overlap between the scopes of those two organizations. Another level of alignment will likely be necessary as RIC applications proliferate. Fujitsu’s Head of Product Planning, Network Automation, Joe Krystofik discussed the balance between standardization and scalability in the context of supporting openness but recognizing the need to foster replicability in the context of applying AI/ML models drawing on network data.
“When we take that to the RIC specifically, its challenging to address the use cases and applications without having a standardized approach on how you deliver the data pipeline for ML adoption. So if you’re approaching ML use cases and applications and you don’t have a regimen, structure and discipline to how you deliver data from the network in a unified manner, it becomes very challenging to reach the end game or the vision of what an Open RAN can contribute to the network and applications over the top. So we recognize that the open standardization, the open community is very important in facilitating and driving that within the RAN network and the RIC specifically.”
