OK, I admit that as a long-time technologist in IT, networking, and network applications, I have a tech bias. Maybe that influences how I parse the coverage of tech in the media, but it sure seems to me that we’re spending as much or more on hyping 5G than we are fixing its problems. In fact, I don’t know whether we’re even trying to fix them, but that might be OK because there may be no fix possible.
I mentioned in a prior blog that, as 5G deployments wind down and the budgets are used up, we could expect to see an attempt to define a hype-successor to keep reporters and editors busy. 6G is the obvious candidate, but since consensus says that it won’t be around for at least five or six years, the realization may be too far out to build any excitement in the present. Some are hoping “5.5G” or 5G Advanced will fill the PR void that 5G maturity is creating, but is there anything to that?
The problem I have with 5G Advanced is that it advances further down a path that, for the original 5G, didn’t pay off. What is it supposed to do? The biggest claims are that it will boost connection speeds and connect more IoT devices. Well, 5G was supposed to do both those things, and yet the average 5G user can’t see any difference in connection speeds, nor can they find any new IoT devices demanding connection. I think it’s safe to say that 5G Advanced is advancing the hype, extending the PR life, but I don’t think it’s moving the ball. Because of this, I wonder whether we’re leaving what we might call the Era of Standards Myths and entering the Era of Media Myths with regard to what’s driving our tech “advances”.
Yesterday, I blogged about the transition from transactional to event-based networking and applications, and this transition is IMHO fundamental to any major change in network requirements associated with (to quote an article), “connecting everything in our lives”. If you think about it, we’re not trying to connect all the things in our lives, we’re trying to connect them to some life-managing application set. Where is that? The mission of connectivity in our Internet age is about connecting to something that hosts a valuable experience, not connecting all the pieces of “everything” to each other. A network can fulfill a connectivity mission, for sure, but it can’t connect what doesn’t exist.
What 5G Advanced is doing, and what 6G will surely do if somebody doesn’t get smart, is push network capability further beyond the applications that would justify it. If I build a rowboat and try to launch it in Death Valley, only to find it doesn’t serve any purpose there, building a yacht instead isn’t the answer. We need to be thinking about where the water is.
Which is at the application level. Applications these days start with a platform, a set of tools that allow developers to build things that can be easily integrated with hosting and network in one direction, and user missions and experiences in the other. I think (speaking perhaps as a superannuated software architect) that we likely have most or all of the elements we need for the platform that would build us toward a 5G Advanced or even 6G network. We apparently don’t know we have them, because we’re not thinking about what the platform has to do, and thus must support to be successful. It is not a network problem, and so network technology, standards, and hype will not solve it, but….
…we probably can anticipate what “the network” that supports the platform solution and the applications that build on it would be. Today, business networking is really transaction networking, as I noted. Transactions are created in somewhat enduring bilateral relationships that, let’s face it, look a lot like calls. If we are serious about connecting everything in our lives, we surely can’t assume that we’re going to establish enduring relationships with them all. But what does a service that supports transient events look like?
In one sense, we know the answer at the higher level at least. It looks like a datagram service, which is after all what IP networks are really based on. It’s TCP, roughly Level 4 in an ISO Model sense, that provides session capability. However, TCP sessions also provide other stuff that we might value in our hypothetical event network, other stuff that we won’t get in IP datagrams.
One such thing is error detection and correction. It’s hard to believe that any real-time event-driven application would function properly in the face of errors in the events or the control packets that were generated as a result of the events. Without TCP there is no error correction. So, should an xG service (where “x” is whatever number you like) provide forward error correction?
Another thing is flow control. IoT devices aren’t exactly powerhouses from the perspective of handling traffic. How can we communicate their ability to successfully receive and process something without creating a multi-step message exchange that would be a latency problem and also burden the device to maintain status on multiple partner applications or elements?
Then there’s security. What do you suppose the chances that the “everything in our lives” is made up of things that are exclusively ours and inaccessible to others? The chance that all the others are benevolent is minimal. How do you secure many lives’ full of simple devices? Encryption is possible at the network level, and proper registration of devices is a good start, but a better start would be to do something like register the hashcode of current software and check the device against it at registration. In any event, there are useful things a network could do here.
I’m not trying to say that any of my examples are compelling justifications for more “G’s” in our mobile networks, or that these represent the only or even best network missions. They’re examples of how a network standard could embrace future missions without requiring that all the technologies needed to create those missions be in place. I’m sure earnest effort would put together a full list, and probably less effort would be involved than that already expended touting missions for 5G that had little chance of developing.
This doesn’t solve the hype problem, though. If we were to prep the network for the future, what’s the chance that we wouldn’t start seeing stories about how the network was creating that future right now. What’s the solution? I think that laying out how an event-driven future application platform, and a mission set to justify it, would actually look is the best strategy. Lay out the puzzle pieces we have, identify from that what’s needed to complete the picture, and hope the hype develops around solutions to the puzzle and not just some random and easily hyped alternate reality.
Whether we succeed in framing that future in a complete user-to-network way, the network features I’ve cited would at least prepare things, and might even boost the chances something will be done to complete the picture at the application level. That’s way better than a repeat of the useless claims 5G has already made.
