Open RAN xApps look stillbornOpen RAN xApps look stillborn

When Steve Jobs showed off the first iPhone in 2007, many spectators would have been intrigued by the mosaic of small square icons on the touchscreen. These clickable software applications – apps, as they are now widely known – became as big a hit and global phenomenon as the gadget itself. An app store stocked 500 of them at launch the following year, allowing third-party software developers to produce their own iPhone apps and split revenues from them with Apple. Today, it is a Brobdingnagian library of more than 3.8 million apps generating over $1 trillion in billings and sales each year, according to one estimate.

The operators of mobile networks hoped to replicate the approach and its obvious success by opening those networks to third parties. Just as an app could be downloaded onto a smartphone, so one designed for the network could be hosted at a basestation site or another suitable node. Apps were discussed that would boost capacity, power down mobile sites when neighborhoods fell silent, carry out predictive maintenance and more.

The open radio access network (RAN) seemed to hold the key. The concept started out as an attempt to pair a radio from one vendor with a baseband unit from another via an open interface. Since those early days, however, it has spread to encompass all the various links between different RAN components. A new platform dubbed the RAN intelligent controller (RIC) would host the apps. All networks had to do was ensure the appropriate interfaces were opened to the RIC.

Unfortunately, it has not gone to plan. Those network apps come in two varieties according to the speed at which they operate. For tasks that can be done in more than a second, there is the rApp, hosted on what the industry calls a "non-real-time" RIC. Anything that demands a faster response, down to as little as 10 milliseconds, needs an xApp and a "near real-time" RIC. The rApps opportunity has taken shape as big vendors offer support for the idea and hospitality to third parties. But the xApps market looks about as active as Santa Claus on Boxing Day.

Open RAN, kind of

That is largely because some of the biggest vendors have left the closed sign up on the front door. Accommodating an xApp requires the use of an open RAN interface called E2. And, so far, Nokia appears to be the only big vendor based outside China to have swung behind it, a fact the Finnish company was keen to highlight in a trial with AT&T last year, about nine months before the US telco giant famously dumped Nokia to build an open RAN with Ericsson instead.

Ericsson, though, has been very clear about its unwillingness to accommodate xApps. "The functionality of xApps is provided by the RAN itself," a spokesperson previously told Light Reading. "At this point, we don't see benefits of xApps on top of what our RAN is doing." Telco advocates would disagree. And Ericsson's critics believe the real issue is Swedish concern that innovation brought by third-party xApps hurts sales of newer and more expensive radios.

Yet Ericsson is by no means the exception. Samsung, number three on the list of vendors outside China, is equally dismissive of xApps, and for similar reasons. "I think a lot of the capabilities that the operators are looking for around that level of very time-sensitive information, we are working with operators to achieve through our vCU [virtual central unit] that exists today," said Alok Shah, vice president of strategy, business development and marketing, for Samsung Electronics America. "So I don't know that I see, in the near term at least, a path for near real-time RIC as a standalone network element."

There is also not much happening outside these vendors in the market for near real-time RICs. While VMware previously advertised one, its status is unclear and Broadcom's takeover of VMware a year ago is a possible threat to this activity, according to an industry source. Amid signs Broadcom has been ratcheting up VMware's prices, Boost Mobile warned during an interview with Light Reading in late November that suppliers could easily be replaced. Just this week, the operator announced a deal with Wind River, one of VMware's main competitors.

Another near real-time RIC was built by Juniper Networks and became a big talking point several years ago. But there has been relative silence since HPE announced a $14 billion takeover of Juniper at the start of this year. Even if HPE remains fully committed to the Juniper RIC, the deal seems to have been held up for a surprisingly long time by a regulatory review. Prospective clients may be wary of signing contracts until the future looks more certain.

Telco frustration

Does any of this really matter? One view is that xApps may be unnecessary if rApps thrive. "When I test this in the lab, I get almost the same performance out of rApps and expect that with some more tuning I will be at xApp performance," said Anders Vestergren, Ericsson's head of SA network management, during a previous interview.

But others disagree, and at least one prominent xApp is unavailable in rApp form. Based in California, Cohere Technologies offers a capacity-boosting Universal Spectrum Multiplier product that today has only two routes into a commercial network – by direct integration with a vendor's basestation technology or via an xApp. It is a favorite with major telcos and regularly touted by Vodafone as an example of innovation in network software.

There is accordingly some frustration within Vodafone about the lack of xApps progress, although Paco Pignatelli, Vodafone's head of open RAN, appreciates why stakeholders are concerned. "We see the different views in the industry about xApps and how feasible they are, because you have multiple applications touching the same component of the radio in real time," he told Light Reading at the recent FYUZ event in Dublin. "Who controls, who polices, all that? That is the question. And some suppliers are taking a clear view on that – they see difficult implementations, even though the idea was good."

Ignoring Cohere, Ericsson may also have a point about the limited value of most xApps. Its latest strategy is to market itself as the supplier of "programmable" network equipment whose 5G Advanced software features – all provided by Ericsson – seemingly play the same roles that xApps would be targeting.

"Our vision is that we will solve whatever problem we can solve in the RAN, because that is where you access the real-time information," said Sibel Tombaz, the head of Ericsson's cloud and purpose-built 5G radio access network (RAN) product line, when she recently spoke with Light Reading. "You cannot wait for seconds, minutes or hours to be able to have a closed loop."

A recently published report from SNS Telecom & IT predicts the market for service management and orchestration (SMO) plus rApps and xApps will generate sales of $700 million in 2027, up from just $50 million this year. But this would have to happen in the context of a RAN market that has been in decline for two years, with overall revenues dropping 11% in 2023, to $40 billion, and expected to fall another 10% to 15% in 2024, according to Omdia, a Light Reading sister company.

The xApps sector accounts for as little as $100 million of the $700 million forecast by SNS Telecom & IT, with the other $600 million in 2027 coming from SMO technology and rApps. Yet by solving problems through its own RAN software, directly installed on its baseband units, Ericsson could be limiting the opportunity for not only xApps but also rApps. If the same features are baked into the RAN, why would anyone buy them elsewhere?