Mobile edge computing has quickly become a strategic imperative for network operators, but little has been said about how carriers will charge for access to the edge.

Amid all the clamoring to deploy mobile edge computing capabilities — striking partnerships with cloud providers and initiating trials to foster potential use cases — the perceived revenue awaiting operators at the end of those efforts is still largely unproven.

Mobile edge computing, in many ways, requires operators to adopt entirely new business models. Carriers and their potential customers are still working through how those services might be packaged and sold.

Moreover, because the edge presents a significant and long-desired opportunity for operators to reach more enterprises, a mere repackaging of existing services will not suffice. “Part of the challenge is that the use cases are still sort of being thought up or developed,” said Phil Marshall, chief research officer at Tolaga Research.

It’s not just early days for edge computing, it also presents operators with a complex and complicated ecosystem that requires specialized hardware depending on the requirements of each specialized use case, Marshall explained.

“You’re now dealing with real estate in facilities that are highly distributed, and because it’s distributed you tend to be building edge infrastructure to support the application,” he said. A cloud radio access network (RAN) implementation is distinct from a hardware standpoint, relative to an artificial intelligence (AI) application, for example.

Edge computing as a business requires carriers to consider if they want to sell an instance, similar to how hyperscalers frame their offerings, or devise infrastructure, facilities, or specific applications as a service, Marshall said.

“We end up with a kind of matrix where our [as-a-service] offerings can be sold either through wholesale or retail channels,” he said. “Operators might wholesale their solution to cloud providers who then bundle the edge capabilities into their cloud offerings. It might also host an automotive edge solution with a wholesale relationship with a car company, for example.”

Nokia Software CTO Ron Haberman also likened the potential business models for mobile edge computing to those pursued by the likes of Amazon Web Services (AWS), Microsoft Azure, and Google Cloud.

“Edge computing, for me, is just one other form of resource and the resource is priced as any other resource, namely compute, memory, disk, and now bandwidth with 5G is going to form into a charge for a particular [service-level agreement],” he said. “As these applications that utilize [edge computing] need those types of resources it would be very likely priced at the rate for that particular SLA just like you today consume resources from public cloud where you basically pay for the element or the machine type.”

As edge computing takes hold, operators will also be determined to not repeat previous mistakes in the cloud. “They don’t want to do what they did in the past with cloud when they tried to compete with the cloud providers and failed,” Marshall said.

Dave Bolan, research director at Dell’Oro Group, expects operators to charge for access to edge computing based on SLAs wherein multiple variations around guaranteed bandwidth, latency, jitter, packet loss, and enhanced security can all be delineated by price.

“I believe the pricing models today from the cloud providers is the best effort,” he said. “The purpose of edge computing is to provide better performance via SLA that comes at a price based on the parameters selected.”

Bolan isn’t too concerned about operators losing control of their respective networks because carriers own the edge sites and can charge accordingly for access to those capabilities. “The biggest advantage is providing wireless connectivity that other edge providers will not be able to offer,” he said.

This requires a change in mindset for mobile operators, according to Marshall. “No one player can control the edge and it is inevitable that others (non-operators) are going to extract value from the networks,” he said. “Operators must shift their attention away from being in control of their assets to one where they can focus on how their assets can best serve customers.”

Marshall is also confident that cloud providers will be good tenants or partners with mobile operators at the edge. “Operators have no choice but [to partner] with cloud providers,” he said. “I think the risks are greater if they don’t.”