In a previous No Jitter post, I shared my thoughts on how the massive increase in licensed and unlicensed radio spectrum
in the U.S and many other parts of the world will lead to a fresh wave of wireless innovation. However, it’s important we don’t forget the other major driver for innovation here: entrepreneurs.
Over the past year or more I have been in touch with a number of startups proposing new and innovative ways of improving cost and performance in cellular networks, easing migration to 5G, and pushing cellular technology into new areas like private 5G networks. The challenge will be getting the cellular providers to embrace the new entrants and their ideas.
As cellular carriers have fixated on rolling out initial 5G services to an increasing percentage of customers, a second battle, to determine how they should go about delivering on that promise, is brewing. With Huawei effectively barred from the domestic market, U.S. mobile operators depend on three primary infrastructure providers: Ericsson, Nokia, and Samsung. The startups, with their innovative ideas for building the network, are gearing up to challenge these incumbents for a share of a potential $200+ billion opportunity.
The big question going forward is: Will the upstarts be able to convince the mobile operators that the cost and flexibility benefits of these new options will be worth the coordination effort and the risk of going with a new supplier? This is shaping up as a classic battle between daring new architecture ideas and decision making in the slow-moving, risk-averse culture of utilities.
Free the RAN!
The RAN is far and away the carriers’ biggest investment, representing between 70% and 80% of capital expenditures, and so it is an obvious area of focus for them. In addition, since the RAN touches the customer, it is the primary determining factor in how good an experience they’re going to have with a service. In a conservative environment, that translates into, “Don’t screw up.”
The primary organizer around the Open RAN is the O-RAN Alliance
, a worldwide community of mobile network operators, vendors, and research and academic institutions operating in the RAN industry. The O-RAN Alliance focuses on disaggregating hardware and software that characterize current monolithic hardware-based RAN solutions to deliver options that not only are cheaper, but also more flexible and adaptable.
These initiatives are coming at the same time that the incumbent suppliers, led by Ericsson, are pushing a RAN upgrade called Centralized RAN
(C-RAN). The two ideas are not mutually exclusive, but having multiple balls in the air doesn’t make the decisions any easier.
What Are the Stakes?
For the moment, O-RAN is the single most recognized battleground between incumbents and upstarts. However, the new technology crusade touches a whole panoply of interconnected ideas that include private 5G networks, small cells, millimeter wave services, and distributed radio systems. The one thing all of these loosely related topics have in common is that 5G brought them all to the fore.
The 3GPP 5G standards opened lots of possibilities for how cellular networks could deliver higher data rates and new services. Those options could incorporate millimeter wave (i.e., greater than 20 GHz) frequency bands, introduce services with varying degrees of quality of service (QoS) through “network slicing,” and include the idea of offering private 5G networks
to enterprises. Since many of the startups are involved in two or more of these initiatives, keeping the different elements separate is getting difficult.
I’ve previously written about millimeter wave
and private 5G, and O-RAN is a topic that spans both of them. O-RAN promises to provide a more flexible, cost-effective, and forward-looking option for the most expensive element in the cellular infrastructure. Traditionally, operators stuck with a single supplier for all the RAN, mobile switching center (MSC, or central office), and core network equipment in all or major regions of their networks.
While the mobile industry involves a lot of high-tech stuff, it is important to remember that the mobile operators are essentially utilities. Utilities are profit-driven entities that must do very difficult jobs (like delivering reliable public radio communication services) on an enormous scale. Nationwide cellular networks utilize upwards of 200,000 cell sites, and while the physical tower may be rented from Crown Castle or American Tower, all of the technology and support needed to deliver the mobile communication service comes from the operator.
Utilities like big contracts to buy lots of stuff from a single vendor so they have that essential one throat to choke. The bonds between the operators and their primary suppliers go back decades, and the technology path is mapped out by the 3GPP (with lots of input from Ericsson, Nokia, Qualcomm, and the like).
In hindsight, this is almost like the old Bell System with its telephone operating companies, Bell Labs, and Western Electric. Of course, you probably remember where that got us.
And Then There’s C-RAN
Along with the idea of opening up the market via O-RAN adoption, the carriers are getting the Ericsson-led C-RAN sales pitch. However, C-RAN appears to have a fundamental inconsistency with another key element in the 5G plans, that being the idea of deploying thousands of millimeter wave small cells to serve high-usage areas we refer to as hotspots.
The biggest development in traditional base station equipment has been a reshuffling of functions and a move to fiber. The first step was to replace the RF-carrying coaxial cable link from the base of the tower to the antenna with a digital fiber connection. At the top of the tower, a device called a remote radio head (RRH) generates the radio signal. The remainder of the base station equipment, called the baseband unit (BBU), then requires less power and air conditioning.
That digital fiber link from the BBU up to the RRH uses a standard called the Common Public Radio Interface
(CPRI); importantly, it has a range up to 40 kilometers. The idea with C-RAN is to haul all of that remaining base station equipment (i.e., the BBUs) back to the MSC/central office, and use fiber links carrying CPRI to extend to all of the different cell sites. So, it’s the BBUs that represent the “centralized” part.
But it gets better.
C-RAN goes hand in hand with the idea of network function virtualization (NVF). Now, rather than having line after line of BBU equipment sitting in the MSC, that functionality can be built to run on standard virtual machines (VMs). So rather than a single-purpose lump of hardware costing tens of thousands of dollars sitting at each cell site, the BBU becomes an instance running on a VM at the MSC (or maybe in the cloud). This sounds a little like the migration we went through on the way to IP PBX systems.
The obvious downside of this is that while C-RAN has a clear advantage when dealing with macro-cell sites, this could be cumbersome when deploying the thousands of small cells envisioned for a 5G rollout. So now we are seeing any number of solutions, many from my previously mentioned upstarts, to organize, integrate, and manage the RAN backhaul infrastructure designed with small cells in mind.
The bottom line is that we are seeing a wide range of ideas from both incumbent suppliers and startups regarding how to configure the RAN and backhaul to provide better, more flexible, and ultimately cheaper ways of supporting these various requirements. The reality is that with the scale of the cellular networks, it’s not likely that carriers will universally adopt any one particular vision. Rather, cellular engineers will have a range of solutions they can use to address specific challenges they encounter in different parts of their service areas.
And the Crowd Cheered… Well, Sort Of
Mobile operators will write the next chapter of this story. The big three — Verizon, AT&T, and T-Mobile — will take the lead, but Dish Networks, if it actually delivers on its promise to build a fourth nationwide 5G cellular network, may come into play, too.
While I have been less than convinced by Dish’s vision that it can build a nationwide network for $10 billion, the company’s website
is replete with leading-edge cellular buzzwords like O-RAN, network slicing, versatile spectrum, open source, open interface, and so on. Also, Dish is currently testing
some of these ideas in its network buildout in conjunction with Mavenir.
The established carriers appear to be taking a more measured approach, keeping their options open. Surprisingly, the most vocal opponent to the O-RAN idea has been Neville Ray, president of technology for T-Mobile. At a BCG and New Street Research 5G Conference
last month, he assessed the O-RAN option by saying, “it’s not ready for prime time for us,” though he allowed, it "kind of makes sense" for a greenfield operator.
This is surprising given that T-Mobile and the newly acquired Sprint were the two carriers that we traditionally found to be most open to new technology ideas like T-Mobile’s embrace of Unlicensed Mobile Access (UMA) and Sprint’s Mobile Integration that utilized equipment from Tango Networks.
I’m afraid that reading the tea leaves on this one won’t yield much insight, and we’ll probably have to wait to see what shows up in the field to see how all this develops. In the meantime, we’re monitoring press releases to get a feel for which way the wind is blowing.
Conclusion: Big Ideas Are More Important Than Little Ideas
In the big picture, all of these new options will be little more than technical curiosities unless the operators start opening up their networks and focusing on delivering meaningful new services. To review, the mobile operators sell: 1) mobile telephony, 2) best-effort mobile Internet access, and 3) very basic texting service (with ubiquitous access). Their most recent service introduction was in the last century.
The operators keep telling us we’re going to get massive-scale IoT-focused services with 5G. But what happened to the Cat M1 and NB-IoT services
they rolled out a few years back? Then there’s the elusive ultra-reliable low latency communications (URLLC) to power our autonomous vehicles. It appears that delivering on these long-held promises will require
a significant core upgrade they reference as stand-alone (SA) deployment. That term describes a 5G implementation using 5G New Radio along with a 5G Core. The other option is a non-stand-alone (NSA) implementation that uses 5G NR with an LTE Enhanced Packet Core. Virtually all current 5G implementations are of the NSA variety.
In the meantime, why don’t any of the cellular carriers offer something like a trunk connection to business telephone/UC&C/whatever you want to call those business communication systems? Why can’t we get QoS-capable data services with SLA guarantees on cellular data networks? Why do we have to rely on smartphone apps to understand the location and status of our mobile devices, when the network already knows? There might be privacy and security issues, but address them.
In the end, the impact of these RAN changes on enterprises may be negligible. A RAN that’s cheaper to run will help hold down the cost of basic cellular service — yippee. On the other hand, mobile executives who quit listening to their own marketing BS and recognize they are operating on a much-outdated, utility-driven business model could potentially affect some real change.
Let’s face it, the operators are sitting on some of the most phenomenal advances in radio technology we have ever seen, and billions of dollars’ worth of licensed radio channels to run them over. And with all that, they still have the same three services (albeit with some performance enhancements) we had at the turn of the century?
A little history: First Western Union put the Pony Express out of business, then AT&T crushed Western Union with its telephone (and hardball business practices). AT&T fell to MCI (and slew of other factors), and now we don’t have a monopoly, but a cooperative cellular oligopoly of similarly organized utilities, in charge of deciding what happens next.
Maybe it’s just me, but that doesn’t sound like a prescription for change. What the operators must come to grips with is that things like O-RAN are minor tactical issues. The big strategic issue is reimagining their business so as not to become the New Bell System.
This post is written on behalf of BCStrategies, an industry resource for enterprises, vendors, system integrators, and anyone interested in the growing business communications arena. A supplier of objective information on business communications, BCStrategies is supported by an alliance of leading communication industry advisors, analysts, and consultants who have worked in the various segments of the dynamic business communications market.