Yesterday, I noted that private 5G offered the best way for telcos and their infrastructure vendors to drive the next wave of IT and networking, real-time applications, forward. They don’t develop new service needs directly, but they do facilitate the expansion of current real-time applications to a state where new services could be of value.
Opportunity forces, then, are a real factor in how service planning for 6G should be done. Is it being done, though, or are telcos obsessing on the fear that the mistakes of 5G, which generated minimal (if any) real ROI for telcos, would be repeated in 6G? It’s good to avoid repeating mistakes, but to do that you need to do something that isn’t one. I’m still not sure telcos get that, and recent announcements seem to support my concern.
The NGMN Alliance (Next-Generation Mobile Networks Alliance) is, self-described, “a global, operator-driven organization, established by leading international mobile network operators (MNOs). As a global alliance of operators, vendors, and academia, NGMN provides industry guidance to enable innovative, sustainable and affordable next-generation mobile network infrastructure.” It’s warning the industry, in a couple of publications representing operators’ views, not to let 6G repeat the mistakes of 5G. That’s hardly surprising to me, given that of the 88 operators I’ve chatted with, 85 say that 6G cannot follow 5G in its failure to address new opportunities in a realistic way, but at the same time demanding a major infrastructure upgrade at significant cost. But will the viewpoint really matter if the group doesn’t embrace a better, specific, alternative?
There is a value to upgrading mobile standards, and a cost. The cost has to be carried by the telcos for service infrastructure and consumers (and telcos, with handset subsidies) for devices. The former are the mobile network operators or MNOs, who are the founders of the NGMN. The value goes first to those who the costs are paid to, and also in at least a justification sense to those who have to bear the cost. Over the last couple decades, the mobile infrastructure vendors and handset vendors have been proponents of change, radical change. The telcos have gone along because, up through 4G/LTE at least, they saw a service revenue value flowing to them. With 5G that changed.
None of the telcos I’ve chatted with will admit that they drove 5G hype on their own. They say that it was media/analysts who did that, likely pushed by a combination of click-appetite and vendor influence. I tend to agree with that; I saw telcos accepting the market vision of others, which was usually the case, and fearing that if they didn’t follow the herd, they’d end up empowering competitors. But none of the telcos today believe that the 5G market vision of the past was valid, and their concern is that 6G will end up repeating that vision failure, which resulted (for 5G) in a major infrastructure spending boom with minimal financial gain to MNOs.
So the statements of the NGMN Alliance fixes that? I’d guess that such an expectation would sound naive to you, and your skepticism would be at least somewhat justified. Yes, the NGMN can discourage vendors from making 6G into another boondoggle, but the 3GPP is a separate body, and telcos don’t supply most of the resources for its work. The most likely way for vendors to assert pressure would not be outright defiance, but a kind of spec creep. Gee, Telco, if we just diddled this or that section a little, it would open up massive revenue opportunity. The diddling could well end when revolutionary fervor has totally displaced evolutionary caution.
The problem here, the problem with a “no-forklifting” goal, is that almost any useful evolution of standards, meaning any that generate any value at all, will tend to impact handsets, radios, and mobility management to some degree. Any useful suggestion would likely live near, or over, the boundary between evolution and revolution. And telcos (almost all of my 88 for example) believe they need some new service revenue opportunity. Can you propose a change that would justify users spending more, but that would have no impact on handset or infrastructure? I’m not at all sure I could even uncover such a thing, unless….
…unless we presumed a jump up the stack to features that were on the network rather than in it. We all know that the Internet shifted the goal of user connectivity from user-to-user to user-to-experience. What gets the user committed is not the route, but the destination. That’s particularly true for broadband, which isn’t useful for text or voice calling. The whole telco mantra of disintermediation came about because telcos first were happy to find a way of getting consumer data applications going because it justified the growing capacity available in the network, then resentful that the suppliers of those applications were making better profits. Telcos, if they want OTT-like margins, need to offer OTT-like services. But they don’t want to do that, so in a way the whole pressure on mobile standards to evolve rather than revolutionize is consistent with their attitude on services; in effect they want a service upgrade in comfortable services that generates revenue, without raising their costs. No forklifts in cost, only in revenue.
The Internet created the need for the progression of G’s, the need for consumer broadband. Absent it, offering users hundreds of megabits or gigabits of connectivity would just waste bits on voice and text. You can’t drive changes in service requirements without changing the thing you’re serving. Some new mission set would have to drive things.
And that’s the problem with the no-forklift movement for 6G. If you don’t define and promote and ultimately validate some new missions in 6G, you cannot do anything that’s really worthy of any new standards at all. Instead, you should simply work on lowering costs until everything vanishes to a point. But to validate a new mission, telcos would need to consider the points I made in my blog yesterday; to focus on connectivity alone in a real-time future means focusing on something the cloud providers and others will use to disintermediate you (again), and perhaps also work to minimize even the need for the services you want to sell.
Low latency and high availability are service features, but their value depends on how you organize event processing. The more you do local to the event, the more higher layers of processing are immunized against the need for these service features. If you’re already hosting transactional front-end processing, as the cloud providers are, would you not want to push low-latency needs down to the local level? If you make chips, phones, computers, would you not want to do the same? It’s very possible, even likely, that the decision to push critical control loop applications to the local level offers the best short-term path of evolution and the best long-term outcome.
The fact is that we don’t know the “ideal” balance of event processing, partly because every stakeholder’s “ideal” is optimized to their own profits, and partly because the way that applications evolve could influence just how much new services are needed, versus how much new local technology and chips are needed. We don’t even know what impacts the shift in event processing location would impact the overall business case for enterprises or value proposition for consumers. This is a complex moving target, and for that reason it may be difficult for organizations like the NGMN Alliance to say “no forklift” without impacting the future opportunities of their members. Better to try to come to terms with where things could go before setting boundaries on what the future services should look like.
