As operators and vendors delve deeper into the development and testing of Open Radio Access Networks (Open RAN), the RAN Intelligent Controller, or RIC, is one of the main elements of focus.
“The main purpose of RIC is to automate and optimize the RAN at scale, with the end goal of reducing the mobile operator’s total cost of ownership,” explained Owen O’Donnell, head of marketing for TerraVM at Viavi Solutions, in a session at this year’s Open RAN Global Forum.
There are two types of RIC, he added: The near-real-time RIC (nRT RIC) and non-real-time RIC (non-RT RIC). They both draw their heritage from self-organizing network (SON) features, O’Donnell said. In SON, there was centralized SON, or C-SON, that resided on servers deep in the network and took feeds from many radios to establish a RAN-wide view on which to base changes, which were then fed back to the RAN for optimization purposes. There was also distributed SON, or D-SON, which resided at the edge in base station software and only focused on the specific base station functions. Neither aspect of SON really took off, O’Donnell explained—partly because of proprietary implementations that tied them to specific manufacturers, but also because SON didn’t act quickly enough to be effective.
“The time for things to take effect was quite long, 15 minutes,” he said. “During that time, an awful lot can happen in the RAN, so it wasn’t actually proven to be too effective.”
RIC, he says is essentially the “rebranding of SON” for next-generation wireless systems, without the barriers of the proprietary ties to specific base station manufacturers, and also with the ability to take closed-loop actions far faster.
“It’s no coincidence that RIC is shaping up just as Open RAN technology takes off,” he added. “The RIC itself is sometimes called the next-generation SON, and it brings to the market so much more than SON.”
There are several differences between the nRT RIC and the non-RT RIC.
The nRT RIC manages events and resources that require very fast response times down to 10 milliseconds, O’Donnell said. It uses closed-loop measurements such as data analytics and input from the non-RT RIC, plus artificial intelligence and machine-learning algorithms “to optimize traffic, to optimize mobility, to optimize load to reduce energy consumption and other improvements across multiple radios,” he added. xApps, which are specialized applications for automation and optimization of the RAN, are hosted on the near-real-time RIC and optimize radio spectrum efficiency. This RIC tends to be located closed to the edge of the network (much like D-SON before it) in order to assist with low-latency transactions.
In contrast, the non-RT RIC, which builds upon the C-SON concept, responds to “less urgent messages” on a timescale of 1 second or more. It operates deeper within the network, from within the Service Management Orchestration or SMO platform. rApps are specialized microservices operating on the nonRT RIC for optimization and automation; the non-RT RIC communicates with the nRT RIC’s xApps to provide policy-based guidance to assist in RAN optimization.
O’Donnell said that Viavi Solutions is seeing RIC vendors, research bodies and universities, and third-party start-ups develop xApps and rApps, which he says offer a way for network operators to differentiate their network’s performance. The xApp vendors themselves, he said, are likely to want to prove their xApps’ ability to make percentage-improvements on specific network KPIs. He sees an opportunity for a new kind of integration lab for xApp benchmarking and certification.
There are hundreds of potential use cases for the RICs, O’Donnell said, and operators are currently interested in traffic steering, network slicing and energy use optimization, among others. While initial interest is focused on low-risk and “quick win” use cases for implementations, he says he expects this to evolve as RICs become more commonly deployed and the standards mature, to capabilities such as signaling storm detection or RIC security-based applications.
“Operator spend billion of dollars on 5G spectrum, and then millions more on new radio units and base stations and small cells and core networks and transmission hubs, and then more again on services to get the systems up and running and tuned and working,” he said. “So the total cost to an operator for 5G is billions of dollar. So how can the operator make the return on investment?” The traditional path is to attract subscribers and encourage them to use costly services and apps that consume a lot of data. The other part of that equation is maximizing the efficiency of how the Radio Access Network runs. With the capabilities of the nRT RIC and nonRT RIC, O’Donnell says, operators can run the RAN “as close to maximum efficiency as possible … getting every last call and data session out of the available spectrum. … This is where RIC comes in.”
O’Donnell’s full session and additional on-demand content from Open RAN Global Forum are available here.