Private 5G and Edge Computing are Sometimes Virtually Synonymous

AT&T Private 5G Edge--now under development--allows enterprises to more easily set up internet of things edge computing networks based on Microsoft Azure multi-access edge computing. In such cases, private 5G is functionaly indistinguishable from MEC.

“A feature we are working on for AT&T Private 5G Edge is the ability to roam off these private networks but still stay connected via AT&T’s U.S. public mobile network,” AT&T says. 

For example, a hospital might use its private network to precisely track ventilators, wheelchairs, and other critical items in its building. But if a ventilator gets loaned to another hospital, that feature ensures the “roaming” machine always remains accounted for even outside the private network, AT&T suggests. 

source: Microsoft 

AT&T also runs its 5G core using Azure, so in a sense, AT&T Private 5G Edge will be running on an AT&T 5G core network that uses Azure cloud computing infrastructure. 

The new offering is designed to help enterprises deploy private wireless networks rapidly across radio spectrums, including Citizens Broadband Radio Service spectrum, licensed or unlicensed, and uses the public network to support a private 5G deployment.

The new offer from AT&T and Microsoft creates private 5G networks using public network resources. Logically, the network is private. Physically, the private network is built using public network resources. 

ABI Research believes “5G network” capital investment, by about 2036, could rival mobile operator capex on the “public” network. Keep in mind that forecast includes what enterprises spend on infrastructure to create their own private 5G networks as well as what mobile operators spend on public 5G. 

That spending includes support of private and enterprise networks “on the premises” and “indoors” instead of being spent on the outdoors infrastructure. What is not yet clear is how much of that investment will create private 5G networks using public network resources, and how much is invested building private 5G networks on a “do it yourself” basis. 

source: ABI Research, Enterprise IoT Insights 

It is one more example of how virtualized public networks create new opportunities for use cases and applications. 

 source: Janakiram & Associates, Forbes. 

Our terminology and thinking evolves as edge computing evolves and we might, in some cases, be overthinking the concept. 

Not all edge computing requires use of the public networks. “Micro edge” computing does all the computation directly on a device microprocessor. “Mini edge” computing also runs autonomously from the public network, but at the device level. 

“Medium edge” uses a cluster of edge machines, while “heavy edge” runs gear in a data center rack. In addition to MEC, which supplies computing someplace outside the enterprise premises, but within a regional location, some refer to “cloud edge,” which is the running of distributed functions at a remote location beyond the metro region.

Access Platforms for IoT are Not "One Size Fits All"

No single access network technology is “best” for every application or use case, and that remains true for internet of things connectivity as well. 

Some use cases only require short-range or  local connectivity; others require wide area support. Some use cases require small amounts of bandwidth, with infrequent updating; others require video bandwidth and almost-continual updating. 

source: Ericsson 

Some applications will have key latency performance requirements; others will not. Power consumption often will be a key constraint, but not always. Device cost must be quite low in some cases, less a concern in other instances. 

For such reasons, mobile platforms are not “best” for every business case. Specialized networks, sometimes using unlicensed spectrum, will be sufficient in many instances. Local connectivity will do for as much as 80 percent of use cases. 

Though mobile service provider interest in internet of things connections is driven in substantial part by the sheer number of projected devices to be supported, incremental connections might not be as large an opportunity as some hope. 

IoT Analytics expects the global number of connected IoT devices to grow nine percent,  to 12.3 billion active endpoints. By 2025, there will likely be more than 27 billion IoT connections, the firm estimates. 

source: IoT Analytics

But there are lots of nuances. Perhaps 22 million of the projected 2025 connections of 27 billion will use some short-range technology, not the mobile or other specialized networks, or fixed networks. 

source: IoT Analytics

In other words, up to 80 percent of IoT connections will use local or personal area networks, not the mobile--or some other network--for access. So new revenue will be an issue, even as huge forecasts are easy to find. 

Will the 80/20 Rule Apply in Telecom?

Most of us are familiar with the 80/20 rule, which suggests that roughly 80 percent of value or outcomes are generated by about 20 percent of actions. Formally, it is the Pareto theorem. 

Pareto applies to most aspects of the connectivity, data center or computing businesses. It even applies to revenue generated by mobile cell sites. Half of mobile revenue is driven from traffic on about 10 percent of sites. Fully 80 percent of revenue is driven by activity on just 30 percent of cell sites. 

source: Medium 

 

Pareto also seems to apply broadly to global connectivity provider revenue and profits as well. Annual global connectivity provider revenue has been estimated at about $1.5 trillion for 2021 and 2022 (including video entertainment subscriptions). 

source: IDC

But as much as $820 billion to $1.1 trillion in revenue is earned from mobile services. Being conservative, assume mobile revenue globally is $820 billion, while total revenue is $1.5 trillion. That implies mobile represents 55 percent of total revenue. 

At the same time, one can note that fixed network data revenues were about $400 billiion in 2020, while voice contributed about $170 billion, for a total of about $570 billion, or 38 percent of total revenue. 

source: N-IX 

Globally, most people using the internet do so using the mobile network. Most people in developed regions have access to both fixed and mobile modes. The percentage of people using fixed internet access alone is almost too small to measure. 

.source: Omdia 

The “mobile-only” pattern of internet access has been in place for close to a decade, as mobile internet usage began to spike upwards since 2010, to the point that half of all the world’s people were using mobile internet by the end of 2019, according to GSMA figures. 

The shift of subscriptions from fixed to mobile happened about 2002. 

source: ITU 

source: ITU

At the same time, one can note that fixed network data revenues were about $400 billiion in 2020, while voice contributed about $170 billion, for a total of about $570 billion. 

On the other hand, capital investment for the fixed network was about 68 percent of total in 2021. So you see the pattern: nearly 70 percent of capex to generate 38 percent of revenue. Conversely, 30 percent of capex spent on the mobile networks generates 55 percent of total revenue. 

 

.source: Omdia 

Though not an idealized Pareto distribution, the distribution of revenues and capital investment is beginning to approach the Pareto distribution. .

Revenue, Profit, Investment Issues Dominate Service Provider Strategy in 2022

Much service provider strategy in mobile and fixed network domains is fueled by revenue and profit issues: slow revenue growth; declining legacy revenues; low returns on invested capital with high demands for investment in 5G and home broadband  and constrained free cash flow. 

Probably nobody will be surprised that S&P Global Market Intelligence expects faster connectivity service provider revenue growth in Asia and Latin America and low-single-digit growth in the United States  and Europe in 2022. That has been the trend for a decade or so. 

There are some clear implications, including a shift to non-traditional funding of access networks. 

“We believe telcos worldwide will continue considering alternate means to fund their capex, including network and spectrum sharing (some of which are arguably regulatory driven). A new trend is the possible  trend to create joint ventures to fund access networks, as Virgin Media O2 is considering. 

Investment capital also will be raised by the  sale of non-core legacy, or lower-scaled telecom operations and media assets; and monetization of tower and fiber networks, S&P says. That trend has been underway for a decade. 

In large part, such asset sales are intended to raise capital either for debt reduction or deployment of capital elsewhere in the business. They are advantageous because high capital investment for both fixed network broadband and 5G, plus shareholder payouts will likely constrain discretionary free cash flow available for debt repayment in 2022, S&P says. 

Globally, secular industry declines from legacy products, coupled with significant competition from cable broadband and lost subsidy revenue, will likely constrain fixed network topline revenue growth and profitability. In fact, S&P Global Market Intelligence expects a high-single-digit revenue decline in 2022.

Shrinking and low returns on invested capital remain an issue as well. “The low return on capital across rated telcos has generally been declining, to less than six percent in 2021 from a bit more than seven percent  in 2011, S&P notes. 

That has fueled a search for new revenue sources. Many operators (in Europe) are diversifying from voice and connectivity services to value-added digital services covering a wide range of IT-related, cyber security, IoT, or cloud-based services, S&P notes. 

That move “up the stack” or “across the value chain” is driven by a search for new growth drivers as legacy services become lower-margin products with little--if any--growth potential. 

Also, many operators are seeking to bolster revenue by investing more heavily in fiber-to-home facilities. But monetization of fixed broadband upgrades has been unequal between regions. 

There has been higher revenue growth in the United States than Europe or Latin America, for example. 

5G monetization is uneven between regions as well. Service providers in Asia-Pacific have seen a lift in average revenue per unit (ARPU) from 5G, largely from moving consumers to higher-priced service plans. 

U.S. mobile service provider revenue also grew in 2021, largely by moving customers to higher-priced plans. 

“We expect U.S. wireless service revenue to have increased 3.5 percent to four percent  in 2021, and slowing to around 2% in 2022,” S&P says. 

source: S&P Global Market Intelligence 

Gains are more muted in other regions such as Europe, S&P says. 

In the U.S. market, a key trend is higher investment in fiber-to-home by telcos, which will limit cable operator subscriber growth and market share gains. 

“We expect U.S. telco capital spending to increase to 13 percent to 15 percent in 2022 as carriers deploy spectrum licenses acquired in recent auctions and for FTTH builds,” says S&P. “We expect increased spending to remain elevated over the next couple of years,” largely to support the 5G mid-band networks. 

source: S&P Global Market Intelligence

“We estimate that U.S. telco FTTH coverage will be around 35 percent  in 2022, up from 31 percent in 2021,” the firm says. “We expect FTTH to cover 50 percent to 55 percent of U.S. households by 2028.

“For 2022 and 2023, we expect wireline capex to increase to 10 percent to 15 percent annually, reflecting the accelerated investments in fiber,” S&P says. 

Still, says S&P, “we forecast total U.S. wireline revenue will decline five percent to seven percent in 2022.”

S&P does not believe 5G will drive high revenue growth. “The wireless industry is mature with limited growth opportunities in the traditional retail market.” Though internet of things offers upside, those “IoT opportunities are likely several years away.”

In Europe, the firm  expects revenue growth will be modest, although up from 2021 levels as roaming and equipment sales continue to recover and because of the gradual benefit of accretive fixed-line broadband upgrades. 

Most telecom operators in the Asia-Pacific region  will maintain steady operating performance in 2022, fueled largely by customer data plan spending.

In Latin America, higher growth will be fueled by data demand and 4G, the firm says. 

Many operators are therefore diversifying--mainly through partnership or mergers and acquisitions--from more traditional voice and connectivity services to value-added digital services covering a wide range of IT-related, cyber security, IoT, or cloud-based services. This strategy also seeks to find alternative paths for growth while traditional services are becoming more and more utility-like. 

Replacing TCP is Becoming More Common

Transport protocols such as TCP/IP are the foundation of computing and communications. But transport protocols change from time to time. So Quic is viewed as a replacement for TCP that also provides reduced latency performance. 

source: NetApp 

The “layers” design of data communications ensures that TCP functions and IP functions can be updated or changed without upsetting the whole transport system. IP obtains and defines the address of the application or device the data must be sent to. 

TCP is responsible for transporting and routing data through the network architecture and ensuring it gets delivered to the destination application or device that IP has defined. But other protocols can be used in place of TCP, such as UDP. 

It might be quite fanciful at this point to speculate on a replacement for TCP/IP as the global standard for communications. It is not speculation that QUIC can replace TCP. 

Other proposals have been floated, but have not gotten traction. Most such proposals involve modifying TCP, though at least one proposal suggests replacing IP.  

Technology Value is What the User or Customer Says it Is

Customer value might be defined as “the perception of what a product or service is worth to a customer versus the possible alternatives.” When new technologies are created, developments have some idea of why they believe an innovation will provide value. 

Sometimes, customers disagree. They see value, but not where developers believed value would exist. Unintended uses, in other words, happen. Technologies and products are used in ways developers did not foresee. Using Xbox controllers to fly unmanned aerial vehicles provides one example.   

Text messaging was a byproduct of instituting Signaling System 7. TCP/IP became the global telecommunications next-generation protocol when it was originally intended to support military computer communications under conditions of uncertainty. 

Consider multi-access edge computing. The stated value is ultra-low latency performance. But connectivity service provider executives say their own belief is that MEC’s greatest value--as a means of moving workloads to the edge,  will be a reduction of bandwidth use or cost. 

But a recent survey conducted by Heavy Reading--94 percent of whom were connectivity providers--suggests the top motivator for moving workloads to the edge is to “reduce bandwidth use or cost.” “Improved application performance” was fourth on the list, ranked in terms of “top motivators.”

source: Heavy Reading 

To be sure, multiple values exist. But it is worth noting that the intended value as developed by suppliers  is not always the “perceived value” seen by end users or customers. 

------------------------------------

For as long as I can remember, calls have been issued that connectivity service providers are missing out on revenue opportunities from small and mid-sized businesses. 

Looking at 5G revenues, for example, BearingPoint and Omdia point out that up to 99 percent of all businesses globally are SMEs, though connectivity providers focus on enterprises. 

source: BearingPoint, Omdia 

That is an “untapped” opportunity, they argue. That is misleading. 

Omdia’s own data show that 63 percent of mobile operators believe large enterprises will generate the “most” 5G revenue. 

Omdia’s own data also suggest that 54 percent of service providers believe SMEs will generate the “most” revenue. That is hardly neglect, as 32 percent of respondents believe consumers will generate the “most” revenue from 5G. 

Keep in mind that mobility has, since the 2G era, been built on consumer revenues, not revenues from business. 

source: Omdia, BearingPoint 

But there are other reasons why smaller businesses are not specifically targeted. The cost of sales and buying behavior is virtually indistinguishable from the behavior of consumer customers. 

In the U.S. market, for example, 83 percent of all businesses are “micro” sized, having no more than nine employees. 

“Small” firms with 10 to 99 employees represent  15 percent of all businesses, while “medium” organizations with 100 to 499 employees represent just two percent of entities. 

If enterprise is targeted directly with field sales, then “micro” (83 percent of business entities) are marketed through the mass market channels. “Small and medium” organizations tend to be marketed to by partner and channel entities. Think of the role played by resellers and system integrators and distributors in the computing hardware business. 

source: CompTIA

So some of us would argue that mobile operators are not neglecting SMEs. They sell using mass market techniques tot he 83 percent of customers that behave like consumers when evaluating and buying 5G services. 

They use channel partners to sell to the mid-market. And they have dedicated field and inside sales teams to sell to enterprises. 

One can argue that more internal resources should be devoted to direct sales efforts for SMEs. But cost of sales is an issue. In many cases, the financial return from higher sales and marketing cost expended on most “small or medium” customer accounts would generate a zero to negative financial return, compared to using channel partners.

All of us have to define our terms: what is “small;” what is “medium” and what does that mean for marketing, sales and requirements? 

Also, are we talking about “new revenue sources” or “total revenue generated by segment?” It might, for example, prove to be the case that much “new use case” revenue is generated by enterprises, as they will be the entities deploying large internet of things and sensor networks. 

Those are non-phone revenues. But phone revenues might still be led by consumer users, as historically has been the case. 

The issue is “which are we talking about: new use case revenues or total or segment revenues?

Either way, though, it might be hard to make the argument that most of the return is going to come from SMEs, as compared to consumer or enterprise revenues. 

For Whom is 5G an SMB Opportunity?

Very frequently, some mobile or fixed network operators have greater or lesser revenue growth opportunities based on their existing market shares, asset and liability positions. A firm with zero to low exposure to a revenue opportunity often has a much-higher chance of fast growth, compared to legacy suppliers that lead a market.

Since many connectivity markets are essentially zero-sum games, a win by one contestant is a loss for another supplier. And leaders have the most to lose; attackers the most to gain.

In the U.S. market, for example, cable operators with zero exposure to fixed voice, mobility services and business internet access have been able to take market share from the incumbents. Conversely, telcos who historically have led those markets have lost share to the attackers.

T-Mobile and Verizon stand to take share from cable operators (who are the market leaders) using new fixed wireless platforms. T-Mobile historically has had zero share of the home broadband market, while Verizon has been limited by its geographic footprint, covering perhaps 20 percent or so of all U.S. home locations with its fixed network.

So the "SMB" segment is an opportunity for some competitors; less so for the leaders. What might be questionable are assertions that mobile operators are "neglecting" the SMB segment where it comes to 5G.

For as long as I can remember, calls have been issued that connectivity service providers are missing out on revenue opportunities from small and mid-sized businesses. 

Looking at 5G revenues, for example, BearingPoint and Omdia point out that up to 99 percent of all businesses globally are SMEs, though connectivity providers focus on enterprises. 

source: BearingPoint, Omdia 

That is an “untapped” opportunity, they argue. That is misleading. 

Omdia’s own data show that 63 percent of mobile operators believe large enterprises will generate the “most” 5G revenue. 

Omdia’s own data also suggest that 54 percent of service providers believe SMEs will generate the “most” revenue. That is hardly neglect, as 32 percent of respondents believe consumers will generate the “most” revenue from 5G. 

Keep in mind that mobility has, since the 2G era, been built on consumer revenues, not revenues from business. 

source: Omdia, BearingPoint 

But there are other reasons why smaller businesses are not specifically targeted. The cost of sales and buying behavior is virtually indistinguishable from the behavior of consumer customers. 

In the U.S. market, for example, 83 percent of all businesses are “micro” sized, having no more than nine employees. 

“Small” firms with 10 to 99 employees represent  15 percent of all businesses, while “medium” organizations with 100 to 499 employees represent just two percent of entities. 

If enterprise is targeted directly with field sales, then “micro” (83 percent of business entities) are marketed through the mass market channels. “Small and medium” organizations tend to be marketed to by partner and channel entities. Think of the role played by resellers and system integrators and distributors in the computing hardware business. 

source: CompTIA

So some of us would argue that mobile operators are not neglecting SMEs. They sell using mass market techniques tot he 83 percent of customers that behave like consumers when evaluating and buying 5G services. 

They use channel partners to sell to the mid-market. And they have dedicated field and inside sales teams to sell to enterprises. 

One can argue that more internal resources should be devoted to direct sales efforts for SMEs. But cost of sales is an issue. In many cases, the financial return from higher sales and marketing cost expended on most “small or medium” customer accounts would generate a zero to negative financial return, compared to using channel partners.

All of us have to define our terms: what is “small;” what is “medium” and what does that mean for marketing, sales and requirements? 

Also, are we talking about “new revenue sources” or “total revenue generated by segment?” It might, for example, prove to be the case that much “new use case” revenue is generated by enterprises, as they will be the entities deploying large internet of things and sensor networks. 

Those are non-phone revenues. But phone revenues might still be led by consumer users, as historically has been the case. 

The issue is “which are we talking about: new use case revenues or total or segment revenues?

Either way, though, it might be hard to make the argument that most of the return is going to come from SMBs, as compared to consumer or enterprise revenues.

Beyond that, some competitors do have an oportunity, especially if they have historically had zero to low exposure to SMB customer revenues. Cable operators have been in that position, as has T-Mobile.