I’ve noted in a prior blog that the Next Generation Mobile Networks Alliance (NGMN) has offered some views on 5G infrastructure. They’ve now released a position paper on 6G, and while we’re surely at least half-a-decade from much 6G relevance, I’ve noted in past blogs that I’m hoping that 6G will correct some of the process errors that led the 3GPP, the NFV ISG, and the O-RAN alliance to sub-optimal thinking. The paper may not answer that question decisively but it may at least give us a read on whether things are starting off right…or wrong. As might well be fitting given how early we are in the 6G cycle, the paper is short and a bit general. Still, if 5G and NFV went off the rails because of early process errors, maybe we can spot troubling signs now and even propose solutions.
The “Overarching Statement” at the start of the paper frames the perception of 6G. “6G will build on, and extend beyond, our existing 5G ecosystem to foster new innovations which deliver value to customers and simplify network operation.” Its goal is “delivering compelling new services and capabilities for customers whilst maintaining essential offerings such as voice.” I have to admit this is a disappointing goal set for me, because I’d contend that it’s pretty much what was said about 5G.
Of 54 mobile operators who commented on their 5G rollout, 48 said it was “successful” but only 9 said it had demonstrated any potential for incremental revenue. Of the 33 who said that it offered that potential, all believed that the revenue potential would come from a new class of applications, and 31 said that that class of applications were best characterized as real-time event-driven or IoT. What features facilitated that application class? “Latency” and “network slicing”. Of the 31 real-time proponents, 17 said they had specific IoT features, but all of that group listed developer programs as the prime example. Twelve indicated they offered IoT connectivity and registration/management services, though I couldn’t find validation for the claim on three of that group’s websites.
Only 4 of the operators (remember that nine said that they’d gotten incremental revenue from 5G) said that developer programs had yielded applications that were revenue-positive for them and all but one of this group said that they were “working hard” to expand that number, but admitted that most of the developer benefits came along quickly, and that growth in support for revenue-generating development had tapered off.
Of the 54 5G operators, none of them suggested that they had actually developed and deployed any specific 5G applications. None indicated that they offered the resources needed to run these applications, only the resources needed to connect them. They valued cloud provider partnerships higher than their own developer programs (by a 3:1 margin) because they believed that “IoT application growth depends on cloud provider support.”
Given all of this, it’s disappointing to find that there’s really nothing in the NGMN paper that references 6G features that would address any of the things that have held back 5G in generating new service revenues. The application targets for 6G are given as “joint sensing and communications, AI, extended AR/VR, enhanced positioning etc.” and again I’d contend that these are really targets of 5G as well, and that in fact if we had to wait until the likely 6G date five years or more down the line, other strategies would either have developed to support all of these, or the opportunity they represented would have proven inadequate to justify investment.
Then there’s the fact that the “Guiding Principles” of 6G initiatives are directed mostly at not disrupting the 5G investment or 5G services. 6G can’t “inherently trigger a hardware refresh of 5G RAN infrastructure.” It can’t “result in degraded performance” for 5G customers, it must be backward-compatible and interoperable with 5G, too. If you add all this on, it’s hard not to conclude that the vision of 6G is to capitalize on 5G successes, but the problem is that those 5G successes haven’t moved the profitability ball much for telcos. You could argue that 5G represented a fairly significant technical shift in wireless infrastructure, but it’s hard to see how that would be the case for 6G given those “Guiding Principles”.
There’s already been flak in the media on the point of preserving 5G infrastructure, but telcos are probably right in opposing major new equipment requirements for 6G. 5G was a major investment that didn’t generate much in terms of incremental revenue in return. To ask telcos to push another wave of hardware changes without compensating revenue surely lined up is to ask them to risk increasing their (unpopular) demand for subsidization from Big Tech. But is expecting 6G to move the ball without any changes to infrastructure realistic? Especially when the priority in service terms seems to be “stay the course” with no service-specific details to support the new applications the paper cites?
I think the problem here is pretty simple. With the telcos, and with telco-driven organizations and standards, we’re getting a vision of future service needs from a group who doesn’t understand at their core that future services are really about software and the cloud, and doesn’t really play in either area. How can you facilitate things like AI or extended AR/VR without the ability to develop applications there? At the very minimum you’d need to understand what communications needs were generated by those applications, and those needs are more a result of the implementation of the applications than the description. What does “IoT” need? Truth is, we don’t know, so we fall back on platitudes like “latency” and “security”.
When I ask telcos why they believed that they needed to define 5G services far in advance of any mass of applications that would justify them, they point out that their service and infrastructure planning cycle is necessarily long. One said “We can’t wait until somebody starts deploying a metaverse and fails on network features to start thinking about what network features we needed.” True, but it’s also unrealistic to think about features for applications that you don’t understand, then compound that by starting to think years before any realistic work is being done on them.
The metaverse, AR/VR, AI, and IoT really do require service support, but they need more than the basic services to connect elements that are being deployed, which in at least some form are already available. They need to get incrementally vital service elements, build a market, and generate the top-level revenue that telco plans for 6G expects will result in a downhill revenue trickle to their own bottom lines. Meta, Microsoft, Google, and others are still working out whether any of those things that 6G proposes to connect would be profitable for them to build. If they aren’t, then there’s no 6G opportunity. Even if they are, the question is whether those who create the applications would promote a 6G-based network service set that, because it was more profitable to operators, would necessarily be more expensive for them and for the users of the applications.
In all, this seems to me to cry out for the “facilitating services” approach that AT&T advocated, but doesn’t appear to have done much to promote. Are there things that, if offered by operators in an efficient way and leveraging feature-deployment resource pools, would be cheaper for operators to sell to a cloud provider or application provider than it would be for those types to build their own solutions? I think that’s the question that has to be answered, not only for 6G but for 5G and for the future of telcos overall.
