Overprovisioning and Faster Speeds are Sort of Related Customer Responses

Disappointing 5G speeds are a common complaint one hears, and the complaint often is justified, especially when mobile service providers use low-band spectrum for coverage and do not have plenty of mid-band spectrum. 

In such cases, as was the situation for early 5G deployment in much of the United States, actual experienced speeds for 5G were about the same as for 4G, and in some cases 4G might have been faster. 

But development within each mobile generation continues while each generation is in its prime, with the result that typical speeds tend to increase for each mobile network generation.

That clearly was the case for 3G and 4G, and technologists already are working on a 5.5G standard that will boost 5G downstream speeds by an order of magnitude. 

That has been the experience in the 3G and 4G eras as well: the mid-cycle update has boosted performance by as much as an order of magnitude. 

Network Generation	Typical Speeds (Down)	Deployment Date
3G	384 kbps - 7.2 Mbps	2001-2010
3.5G	7.2 Mbps - 21 Mbps	2008-2012
4G	10 Mbps - 100 Mbps	2010-present
4G-Advanced	100 Mbps - 1 Gbps	2014-present
5G	1 Gbps - 10 Gbps	2019-present
5.5G	10 Gbps - 100 Gbps	2025-2030 (proposed)

None of that is going to impress many observers, any more than faster home broadband speeds tend to be associated with specific new use cases or applications. Faster is better, in part to support multi-device and multi-household accounts; to account for congestion during the day or week; and even, for some, as a way of compensating for performance hits imposed by use of virtual private networks. 

The actual improvement in single device, single user experience might not be so great, except as an insurance policy to maintain adequate throughput or latency performance at peak load periods. 

So, yes, we are likely in a situation where raw increases in expected speeds, or lower latency performance, is desirable for all sorts of reasons beyond the minimum performance levels to support a single user, on a single device, using a single app. 

As was the case in earlier years, faster speeds are essentially forms of overprovisioning, as customers once purchased usage buckets larger than they expected to require, simply to avoid surprise overage charges. 

In the same way, faster access plans are a way to overprovision capacity so performance by any single user, on any single device, running any particular app, is protected from congestion during peak hours of network usage.

Bad Storylines With a Long History

A well-known and frequent storyline is that the United States is lagging behind on some key measure of communications or computing infrastructure. That was said to be the case for mobile phone adoption, text messaging, home broadband speeds or coverage. 

Some still make that argument about U.S. home broadband, despite much data to the contrary. 

One frequently hears that U.S. home broadband is expensive or slow, for example. But in 2020, about 80 percent of U.S. households could buy gigabit per second service if they choose, looking only at coverage by cable TV networks. And speeds since then have started to ramp up into multi-gigabit ranges. 

By the end of 2025, for example, one major internet service provider, Comcast, will have enabled multi-gigabit speeds for 50 million homes, with plans to upgrade to symmetrical 10-Gbps service. Other ISPs likewise are boosting speeds into multi-gigabit ranges.  

"Compared to what?", one is tempted to ask. One frequently hears that home broadband also is expensive. Again, the question is “compared to what?”  Recent studies suggest that prices have been declining, even when not normalized for currency differences, inflation or cost of living differences between countries.  

That “lagging” storyline has been used about U.S. 5G. The point is that the storyline often is correct for a while, but never correct over the longer term. Just as important, any recitation of infrastructure performance is relatively non-correlated with the benefits such advanced platforms are always said to support: economic development; productivity growth; educational outcomes; innovation or jobs growth. 

We simply cannot measure the actual causal effects of better broadband. 

But the “U.S. is behind” storyline has been used often over the last several decades. Indeed, where it comes to plain old voice service, the U.S. is falling behind meme never went away.

In the past, it has been argued that the United States was behind, or falling behind, for use of mobile phones, smartphones, text messaging, broadband coverage, fiber to home, broadband speed or broadband price. 

In the case of mobile phone usage, smartphone usage, text message usage, broadband coverage or speed, as well as broadband prices, the “behind” storyline has proven incorrect, over time. 

Some even have argued the United States was falling behind in spectrum auctions. That clearly also has proven wrong. What such observations often miss is a highly dynamic environment, where apparently lagging U.S. metrics quickly are closed.

6G Is Not a "Race"

Many observers argue that 6G mobile networks are a “race,” as was claimed for 5G as well. That arguably is true from the standpoint of a “race” between infrastructure providers to sell networks. We can measure market share for mobile infra sold by Huawei, Ericsson and Nokia, for example. 

The other claimed “race” asserts that nations deploying 6G early gain an advantage over countries that deploy a bit later. That second sense of “a race” makes less sense. 6G, as was 5G and 4G before it, will be deployed earliest in countries that already are developed, in terms of computing and communications usage; economic productivity and output. 

In other words, early 5G and 6G is a correlation with already-advanced computing, communications and economic output. High degrees of wealth, income and economic activity are correlated with early 5G deployment. It will be the same with 6G. 

Mobile Generation Key Performance Indicators

KPIs	4G	5G	6G
Peak data rate /device	1 Gbps	10 Gbps	1 Tbps
latency	100 ms	1 ms	0.1 ms
Max. spectral efficiency	15 bps/Hz	30 bps/Hz	100 bps/Hz
Energy efficiency	< 1000x relative to 5G	1000x relative to 4G	>10x relative to 5G
Connection density	2000 devices / km2	1millon devices /km2	>10millon devices/km2
Coverage percent	< 70 %	80 %	>99 %
Positioning precision	Meters precision (50 m)	Meters precision (20 m)	Centimeter precision
End-to-end reliability	99.9 %	99.999 %	99.9999 %
Receiver sensitivity	Around −100dBm	Around −120dBm	< −130dBm
Mobility support	350 km/h	500 km/h	≥1000 km/h
Satellite integration	No	No	Fully
AI	No	Partial	Fully
Autonomous vehicle	No	Partial	Fully
Extended Reality	No	Partial	Fully
Haptic Communication	No	Partial	Fully
THz communication	No	limited	Widely
Service level	Video	VR, AR	Tactile
Architecture	MIMO	Massive MIMO	Intelligent surface
Max. frequency	6 GHz	90 GHz	10 THz

source: Science Direct

In other words, early deployment of 5G or 6G does not actually provide quantifiable advantages where it comes to economic growth. That growth already exists. 

The converse might be argued: were a developed or developing nation to refuse to deploy the next G network--something that never has happened--we might see an eventual degradation of economic activity.

We cannot test the thesis because it never has happened. 

That is not to deny the new “lead” applications or “new” use cases that have developed with each next-generation mobile network. But we might also claim that for each 10X increase in fixed network home broadband as well.

Over time, new use cases emerge; new apps are possible. Over time, as computing, storage and communication costs fall, while capabilities increase, developers and entrepreneurs are able to create new use cases and products. In that broad sense, yes, it does matter that networks and computing capabilities advance. 

But such developments always happen in concrete circumstances, with existing capabilities already in place: human capital, rule of law, capital availability, geography, infrastructure and other network effects. 

A next-generation network--in and of itself--does not create new possibilities if the other foundations are not already present.

T-Mobile Passes Informal Test of New Product Success

One informal test of new product relevance at a tier-one connectivity service provider such as AT&T or Verizon is that a new product has to offer potential revenue of at least $1 billion a year to be worth pursuing. 

By that informal test fixed wireless already is rapidly approaching that point at Verizon and already has surpassed the threshold at T-Mobile. 

In February 2023 Verizon claimed 1.5 million fixed wireless subscriptions, with an average annual revenue contribution of about $500 per account (about $42 per month, per account). That implies $750,000 in revenues.

T-Mobile reported some 2.6 million fixed wireless accounts. Assuming the same monthly revenue, that implies T-Mobile is generating $1.3 billion annually from fixed wireless services. 

Neither of those firms, nor AT&T, Comcast or Charter Communications yet report revenue or subscriptions for new services from internet of things connections, edge computing or private networks. That suggests revenue for those products is not close to $1 billion per year in new revenue.

Granted, consumer 5G subscriptions are a "new product." But as each new 5G account mostly cannibalizes and replaces an existing 4G subscription, 5G is properly viewed as the upgraded replacement for an existing service, and not a "new service or product."

Fixed wireless might be properly viewed as a new service because it allows T-Mobile to enter a business--home broadband--in which it has had zero percent market share. Fixed wireless also is a new service for Verizon in allowing it to provide home broadband to perhaps 80 percent of U.S. homes where Verizon has no fixed network.

Though Perhaps 80% of "Rest of World" 5G Investment Remains Unfilled, U.S . and China have Passed the Peak

Revenue troubles at Ericsson, with growth rates mired at the one-percent to two-percent range, and generally mirrored at Nokia, but the global legacy telecom supplier chain has shrunk for at least a couple of decades. 

At least some of the trouble at the suppliers is that capital investment in new mobile networks follows a clear “S” pattern. 

Spending always ramps up at the onset of construction. It’s a brand new network, after all. At some point, when perhaps half of the population is covered, spending starts to slow, before tapering back down to a “maintenance” mode. 

source: Dell'Oro Group 

Looking back, AT&T spent heavily to build out its 3G network in the early 2000s, with spending peaking about 2006 and then declining. That’s normal. The same pattern held for 4G. 

AT&T spent $17.2 billion on capex for its 4G network in 2010, growing to about $26 billion by 2012. But that was the peak. By 2013 AT&T's capex fell to $23 billion, reflecting the fact that a substantial portion of the network had already been built. 

China Mobile spent $20 billion in capex in 2010 and peaking at about $26.5 billion in 2012. By 2013, China Mobile's capex had fallen to $23.4 billion. Again, the spikey pattern is dictated by the build schedule when creating a new network. 

Granted, the rest of world investments in 5G will continue. But much of the North American and China networks are substantially completed. 

The other issue for Ericsson and Nokia has been the emergence of Huawei and other suppliers including ZTE and others such as Samsung, plus specialists of all sorts that pose gross revenue and profit margin pressures for the legacy suppliers.

Market share is an issue. In 2021, for example, Ericsson had 23.5 percent share. Nokia had

22 percent. Newer supplier Huawei had 22 percent share, while ZTE captured about seven percent. Between them, Samsung, Cisco and Ciena had a combined market share of around 17 percent. 

Also, mobile service providers are always working on ways to reduce capex and increase the efficiency of their investments. Some of that also is at work. But competition and the maturation of newbuild activity likely explain much of the flattish revenue performance.