We’re in yet another cycle of 5G hype so I need to explain the problem with 5G without getting into all the technical details.

I’m writing this as the 5G hype cycle reaches a new crescendo.

What is most striking is how few of those touting 5G (or “private 5G”) is how little technical detail there is beyond saying it is the next generation of cellular protocols. Yes, one can look at volumes of specifications with myriad options and flavors but no sense of 5G as such.That alone should make one suspicious—why does one have to eschew wires to get 5G? But the lack of a definition is a feature—it allows extravagant claims without the inconvenient limits of science and engineering.

The challenge I have in explaining the problem with 5G is the claims of how capacity and low latency seem plausible and wonderful and there is a grain of truth for those looking for confirmation.

In traditional telephony, there was a wire dedicated to each phone line whether it was in use or not. The situation improved once you got to the nearest central office and calls could be assigned to any available pair of wires. There was still a limit on the total number of calls. Phones calls were expensive because, in effect, you were renting a dedicated wire for the duration of the call (even if you weren’t saying anything).

Everything changed with the advent of VoIP in the 1990s. Now all those calls could be handled without dedicating facilities. The same copper wire that handled one call could handle all the calls at once. The scarcity created by dedicating wires disappeared.

This created a problem—phone calls were now apps and the value they created no longer went to fund the network. In effect, the anchor tenant that paid for facilities left the building.

The race was on to find a new tenant—one that required dedicated facilities. More to the point, one that required the low latency guarantees that could only be made if a provider controlled the entire path.

This is why the loss of voice traffic was so devastating—it was the premier example (or, for your old folks, the poster child), of an application requiring such guarantees. Remote control for surgery and driving are prime examples. The other big one is real-time remote video gaming—that thing that cost Facebook, I mean meta, billions of dollars.

5G emphasizes wireless because the same is over for wired (or fiber) connections because we know we don’t need a dedicated path for each call. But the cellular world is still proffering the old model even though, there too, all the calls use VoIP technology over a common data network.

It’s similar to the period in time when wired phone companies charged for virtual phone lines over broadband priced as if they still needed dedicated wires. This is the case because, unlike your home in which you could just buy Wi-Fi and do it yourself, we still use the old economic model because we’ve done such a good job emulating the past that we’re not forced to rethink the premises even to the point of accepting the idea that we’re using up capacity for each call!

OK, maybe we’re not running out of capacity for cellular calls but what about those billions and billions of devices? What about them? They generally use local connectivity and are served well by Wi-Fi.

There is a point—we do need to provide more wireless capacity and there is technology—MIMO—which greatly increases the capacity. It’s also possible to massively scale MIMO to share facilities. The choice is or should be an engineering decision. It’s easy to add capacity—you can just buy more Wi-Fi capacity without the complexity and never-ending costs of renting a shared base station.

MIMO itself has nothing necessarily to do with 5G. The problem is that researchers misrepresent their research as being about 5G thus misleading those making public policy.

In a normal market, if the anchor tenant (voice) has left the facility, the cost to those who remain would go up by a factor of 100!

It hasn’t because the legacy phone companies have been granted total control of the commons—wired and wireless! They then get additional largess in the guise of broadband subsidies because their business model no longer makes sense!

The alternative is a public packet infrastructure (https://rmf.vc/CIPPI) which allows us to take advantage of the abundance that VoIP tapped into.

It gets weirder with 5G because none of the applications cited make sense. At the very least you can’t guarantee against all failures (even if you say they are acts of God). So, we need to implement protocols that handle latency no matter what the promises are. It was OK with voice calls because people could handle the interruptions but if you absolutely require low latency lack the resilience needed for vital applications. And the Mars rover has shown that we know how to do remote driving even with a two-hour delay!

It’s easy to see the appeal of the 5G story. It plays into a simplistic idea of progress as incremental improvement when the reality is that real change is about discovery rather than simple FBC (Faster Better Cheaper).

It’s also easy to understand the focus on wireless—the last refuge of telecommunications. It’s similar to when the carrier saw their voice business threatened so they took an existing technology and repackaged it as ADSL to sell Interactive TV. And then that failed, they repackaged it as DSL with the Internet as a service that, no surprise, was supposed to need a dedicated, rented, pipe.

Those implementing new protocols and services see past this. Why pay a premium for 5G when you can use the shared Internet at no additional cost? Phone users don’t see any difference between phones with and without 5G.

There is some branding of point-to-point MIMO connections as 5G Home Internet but that’s just a branding of MIMO and is otherwise the same as using a cable.

You don’t need to be very technical to see through the story—you just need to step back and take your blinders off.