Wednesday, December 30, 2020

U.S. C-Band Spectrum Prices Too Exuberant?

As bidding for 280 MHz of U.S. mid-band spectrum in the C-band exceeds expectations, it might be reasonable to wonder if mobile operators risk overpaying for 5G licenses. Given past experience with some spectrum auctions in Europe for 3G licenses, as well as some 4G auctions, it is a reasonable question.


But some will argue the C-band spectrum is strategic for a number of reasons. At a high level, it rarely is possible to run a large mobile business--and control costs or attempt differentiated packaging--without access to licensed spectrum. 


At a more-granular level, U.S. mobile operators have, up to this point, faced clear constraints in commercializing 5G services using the mid-band spectrum, especially in the 3.2 GHz to 4 GHz bands that have proven the best way to supply both 5G coverage and capacity. Most of the mid-band spectrum in those ranges have been licensed to other users in the U.S. market, where that has not been the case in most other places in the world. 


That has meant the U.S. operators (with the exception of T-Mobile, which has a trove of 2.5 GHz spectrum) two relatively unappetizing choices: use the low-band or millimeter wave spectrum. 


The downside: low-band provides good coverage, but no capacity advantages. Millimeter offers huge capacity increases, but at the price of coverage. The former means customers will not see faster speeds 5G promises; the latter means higher infrastructure costs and a slower commercialization timetable. 


In a nutshell, AT&T and Verizon--among others--really need lots more mid-band spectrum to deliver on their 5G promises of higher speeds and coverage, in the near term. T-Mobile will be more than willing to tout the typical scale of its nationwide 5G network. 


source: T-Mobile 


In principle, AT&T and Verizon might overpay for C-band spectrum, but at least so far, that is not an issue. C-band prices remain below the prices paid for AWS-3 spectrum that was vital for 4G, for example. Mobile operators always are wary of overbidding for spectrum rights, recalling the high prices paid for 3G spectrum in Europe, for example, which nearly bankrupted the auction winners. On the other hand, spectrum assets can be quite strategic, as facilities-based competition leadership without adequate spectrum resources is impossible.  


The robust bidding for 280 MHz of C-band spectrum is an indication of the importance mobile operators place on mid-band assets. T-Mobile’s purchase of Sprint likewise was driven in large part by access to the trove of mid-band spectrum Sprint possessed. 


Spectrum prices in the recent CBRS auctions (also of new mid-band spectrum) generated average prices per person (per MHz-POP) of about 21.6 cents. The C-band auction already has average prices of about 79 cents per MHz-POP, higher when including the additional clearing payments to existing licensees. Including those payments, we already have hit about 95 cents per MHz-POP in the C-band auction.


Prices for the A block of frequencies, which it is estimated can be put into commercial use as much as two years earlier than the BC block, are running higher. After the end of round 44 of bidding, A block licenses were at an average of $1.21 per MHz-POP, with a price of $1.54 including the cost of clearing existing users from the spectrum.


Thursday, December 24, 2020

Get Ready for Appellation "Biggest Auction Ever" for C-Band

Get ready for it: the C-band auction will soon be called the “biggest spectrum auction ever,” referring to the amount of proceeds. Keep in mind that the sheer amount of spectrum licensed (280 MHz) vastly exceeds what has been offered in any prior single auction. 


The former “record” for gross proceeds was the 2015 AWS-3 auction of 65 MHz of mid-band spectrum in the 1.7 GHz to 2.2 GHz frequency ranges. The AWS-3 auction raised about $45 billion. 


After 44 bidding rounds, the C-band auction (Federal Communications Commission Auction 107) seems to have already reached $70 billion. But keep in mind that the C-band assets represent capacity at least four times greater than was the case for AWS-3.


Generally speaking, the most-recent spectrum auction is a guideline for expected prices in the current auction, but the comparison with the Citizens Broadband Radio Service auction earlier this year is not necessarily “like compared to like.” Most significantly, the CBRS licenses are not as unrestricted as C-band licenses will be. 


Since CBRS uses a shared spectrum approach, in many areas the existing licensees will have first use of the spectrum. Only where the assets are not required will secondary licensees have rights to use the frequencies. Also, where there is excess capacity, and neither the primary or secondary licensee uses all the spectrum, unlicensed use is permitted. 


So prices for CBRS, all other things being equal, were expected not to reach the levels of the C-band assets, which are unrestricted and exclusive in terms of use. 


Spectrum prices in the recent CBRS auctions (also of new mid-band spectrum) generated average prices per person (per MHz-POP) of about 21.6 cents. The C-band auction already has average prices of about 79 cents per MHz-POP, higher when including the additional clearing payments to existing licensees. Including those payments, we already have hit about 95 cents per MHz-POP in the C-band auction.


Prices for the A block of frequencies, which it is estimated can be put into commercial use as much as two years earlier than the BC block, are running higher. After the end of round 44 of bidding, A block licenses were at an average of $1.21 per MHz-POP, with a price of $1.54 including the cost of clearing existing users from the spectrum. 


The other relatively-recent precedent, the AWS-3 auctions, also set a record for gross proceeds at the time. But AWS-3 arguably most affected AT&T’s ability to keep pace with Verizon’s 4G spectrum assets.


The C-band auction is more strategic, in that it is, for all providers but T-Mobile, the primary way 5G using mid-band can be provided. And mid-band 5G is important because it provides the “best” blend of coverage and capacity for 5G, as 5G rollouts in other countries illustrate. 


It remains unclear is whether the C-band auction prices will top the AWS-3 auction of mid-band spectrum held in 2015, when prices reached $2.90 per MHz-POP, on average. At least so far, prices do not seem in danger of reaching levels that might be deemed excessive.

Wednesday, December 23, 2020

C-Band Prices are Robust, But Not as High as AWS-3 Mid-Band Prices

Mobile operators always are wary of overbidding for spectrum rights, recalling the high prices paid for 3G spectrum in Europe, for example, which nearly bankrupted the auction winners. On the other hand, spectrum assets can be quite strategic, as facilities-based competition leadership without adequate spectrum resources is impossible.  


The robust bidding for 280 MHz of C-band spectrum is an indication of the importance mobile operators place on mid-band assets. T-Mobile’s purchase of Sprint likewise was driven in large part by access to the trove of mid-band spectrum Sprint possessed. 


Spectrum prices in the recent CBRS auctions (also of new mid-band spectrum) generated average prices per person (per MHz-POP) of about 21.6 cents. The C-band auction already has average prices of about 79 cents per MHz-POP, higher when including the additional clearing payments to existing licensees. Including those payments, we already have hit about 95 cents per MHz-POP in the C-band auction.


Prices for the A block of frequencies, which it is estimated can be put into commercial use as much as two years earlier than the BC block, are running higher. After the end of round 44 of bidding, A block licenses were at an average of $1.21 per MHz-POP, with a price of $1.54 including the cost of clearing existing users from the spectrum. 


The robust bidding for 280 MHz of C-band spectrum is an indication of the importance mobile operators place on mid-band assets. T-Mobile’s purchase of Sprint likewise was driven in large part by access to the trove of mid-band spectrum Sprint possessed. 

 source: Sasha Javid


The point is that it now appears C-band assets are of the strategic type, based on bid prices. T-Mobile arguably has little need for additional mid-band spectrum, but Verizon and AT&T obviously do need mid-band assets badly. Other bidders such as the cable companies and Dish Network have needs, but arguably not to the extent of Verizon and AT&T. 


What remains unclear is whether the C-band auction prices will top the AWS-3 auction of mid-band spectrum held in 2015, when prices reached $2.90 per MHz-POP, on average. At least so far, prices do not seem in danger of reaching levels that might be deemed excessive.

Autonomous Vehicles and Smart Dust are Disruptive Opportunities for Mobile Operators

Autonomous vehicles and balloon-delivered internet are a few disruptive technologies with telecom implications that are “just over the horizon,” according to a compilation prepared by the Imperial College London. 


“Smart dust”--tiny microelectromechanical systems--are another category of just-over-the-horizon technologies with communications service provider implications. 


Imperial College London 


Smart dust and autonomous vehicles--including unmanned aerial vehicles used for freight delivery--might be classified as internet of things use cases, and all will use untethered or mobile networks for command and control. 


60-GHz Makes a Bug a Feature

In 2001 the Federal Communications Commission created a new unlicensed band of seven GigaHertz worth of spectrum between 57 and 64 GHz, available for wireless communications.


The 60-GHz wireless band was likely chosen for unlicensed use at least in part because of the high signal attenuation. As this graph shows, signal attenuation caused by oxygen absorption becomes much higher in the region around 60 GHz than at the lower and higher frequencies around the 60-GHz peak. 


source: Electronic Design 


On the other hand, high signal attenuation is less of an issue for indoor or short-range communications such as Wi-Fi, which makes 60-GHz signals feasible for local signal distribution. 


Still, point-to-point wireless systems operating at 60 GHz have been used for many years by the intelligence community for high security communications and by the military for satellite-to-satellite communications, in large part because the high oxygen attenuation also means greater security.


60 GHz satellite-to-satellite communications prevent eavesdropping because the earth’s atmosphere acts like a shield. 


Also, highly-focused 60-GHz signals have another advantage for terrestrial communications. The beams can be focused tightly enough that there is little interference with signals at other frequencies, on other paths. That is part of the thinking behind the Terragraph platform, used either for signal trunking or signal distribution (end user access).  


Also, despite the high path loss at higher frequencies, the right antenna design and use of advanced chip-level signal processing can drastically reduce signal loss compared with sub-6-GHz bands. It is possible that millimeter wave signals up to 300 GHz can have signal loss of 10 dB/km.


That should enable mobile use for small cells covering up to about a kilometer, or fixed access at up to about 10 kilometers. 


The other point worth making is that we now are about 20 years removed from the decision to allocate the 60-GHz band for unlicensed use, with commercial deployments yet to scale. It sometimes takes much longer than expected for a new resource to be deployed commercially.


U.S. 4G Might be Faster than 5G, for the Moment

U.S. 5G “available everywhere” speeds lag speeds for most 5G networks deployed elsewhere, for reasons directly related to the spectrum assets available at launch. Simply put, most other 5G networks globally were able to use mid-band spectrum. 


U.S. operators for the most part had to rely on low-band spectrum, spectrum sharing or spot deployment of millimeter wave spectrum in the most-dense areas. That is why the Citizens Broadband Radio Service and current C-band auctions are so important: they will supply lots of mid-band spectrum for 5G. 


At the moment, on some U.S. networks, 4G might be faster than 5G. On other networks 5G speeds will be comparable to 4G. The reasons have everything to do with the spectrum assets presently deployed. 


Up to this point, T-Mobile 5G has used low-band spectrum for coverage, as have AT&T 5G networks. That inherently means lower speeds than will be possible when millimeter wave spectrum is more widely deployed and as mid-band spectrum becomes the primary coverage layer spectrum.


Verizon, which has very little spectrum available to support 5G, has used dynamic spectrum sharing, allowing its 4G network to support 5G devices. It helps, but is not a panacea. Keep in mind that Verizon, with the most customers, and the least, spectrum-per-customer  will find even this bridge strategy difficult. 


The main takeaway is that, for U.S. mobile operators, the coming mid-band capabilities will dramatically boost 5G speeds. As that happens 5G performance will change dramatically, for most potential users. For some use cases, millimeter wave assets will radically boost speeds. 


The robust bidding for 280 MHz of C-band spectrum is an indication of the importance mobile operators place on mid-band assets. T-Mobile’s purchase of Sprint likewise was driven in large part by access to the trove of mid-band spectrum Sprint possessed. 


Spectrum prices in the recent CBRS auctions (also of new mid-band spectrum) generated average prices per person (per MHz-POP) of about 21.6 cents. The C-band auction already has average prices of about 79 cents per MHz-POP, higher when including the additional clearing payments to existing licensees. Including those payments, we already have hit about 95 cents per MHz-POP in the C-band auction.


Prices for the A block of frequencies, which it is estimated can be put into commercial use as much as two years earlier than the BC block, are running higher. After the end of round 44 of bidding, A block licenses were at an average of $1.21 per MHz-POP, with a price of $1.54 including the cost of clearing existing users from the spectrum. 

 source: Sasha Javid


The point is that, as with the early days of any new mobile platform, performance and coverage will constantly shift. That arguably is especially true in the U.S. market, where the crucial mid-band spectrum is just now being awarded.


Tuesday, December 22, 2020

By 2025, 53% of World Population Will be Able to Buy 5G Service

By 2025, about 53 percent of the world’s population will be able to buy 5G service, according to Bankr. Network coverage will grow by 254 percent on a cumulative basis between 2020 and 2025,  reaching about 4.1 billion people. 


In 2020, about 15 percent of the world population, or 1.17 billion people, will be able to buy 5G service (though not all will, by any means).

source: Bankr

CBRS Infrastructure Supports Mobile Broadband and Fixed Wireless Today, Private Networks by 2025

The Citizens Broadband Radio Service infrastructure market is dominated by fixed wireless access and mobile broadband use cases in 2020, with the private wireless segment representing a mid-single-digit share, says Kyung Mun, Mobile Experts senior analyst.


That should change by 2025 as private wireless reaches almost half of the CBRS infrastructure and devices market.

source: Mobile Experts, Fierce Wireless

Monday, December 21, 2020

5G Apparently is Being Deployed Faster than 4G Had Been Rolled Out

Despite some fears, it appears 5G is being deployed faster than 4G had been. 

Three UK Customer Average Data Consumption Now Exceeds 15 GB

Three UK says its customers now average 15 gigabytes of mobile data consumption each month, Three UK says. Three UK customers have in the past consumed about three times as much mobile data as other U.K. mobile customers.


One possible explanation for that difference is that Three UK customers are more likely than customers of other mobile networks to rely on mobile-only access. The likely reason for that behavior is Three’s marketing emphasis on large usage buckets and unlimited plans.   


According to 2018 data from Ericsson, that would be anomalous, as the Western Europe average was about 6.1 GB per month. 


source: Ericsson 


That compares with a U.S. average of about 8.6 GB data consumption, according to Ericsson. The difference might lie in the higher amount of Wi-Fi offload in the U.S. market or work from home policies that kept people indoors more than usual, and lead to increased fixed network data consumption.  


Saturday, December 19, 2020

Verizon Expected to Lead Spending in C-Band Auction

Many expect the U.S. C-band spectrum auction will result in about $47 billion in bids, with Verizon expected to spend most heavily, as it has the greatest need for additional spectrum. Verizon has the most customers but the least bandwidth of all three national networks, meaning Verizon’s networks are more highly loaded than its competitors, in terms of potential bandwidth per customer


After 33 rounds of bidding, a total of about $34 billion already has been offered for the 280 MHz of prime mid-band spectrum favored for 5G because it is a good compromise of coverage and capacity features. 


Also, the new 280 MHz of capacity boosts total U.S. low-band and mid-band mobile spectrum by nearly 50 percent. It is consequential. 


Some might argue it is possible total bids will be higher than $47 billion. Analyst Sasha Javid tracks FCC auction prices closely. His analysis of price trends after 33 rounds (C-band auction is the orange line) suggests demand has been stronger in the early going than in past auctions. 

source: Sasha Javid


Some might argue that if that demand curve remains higher for longer, prices will rise. Bidding for “A block” licenses that will be cleared earliest suggests bidders value the ability to put that spectrum to work earliest. It is possible that B block and C block licenses will not be cleared until perhaps 2023. 


If so, prices for A block licenses should be higher than for B and C allocations, all other things being equal. 


Bloomberg reporters Scott Moritz and Todd Shields suggest that Verizon will wind up spending  about $20 billion; AT&T about $9 billion, with lesser amounts committed by T-Mobile ($6 billion). 


Dish Network might bid $3 billion, while Comcast and Charter, bidding in a consortium, might be willing to spend $5 billion. 


As this analysis by S&P Global shows, T-Mobile has, by far, the greatest present trove of mid-band spectrum. 

source: S&P Global

Thursday, December 17, 2020

Early Thinking on "Next G" after 5G


Here's the optimistic view of what could happen in the 6G era, and what capabilities a 6G network might feature. 

It might be thought of as an early imagining of capabilities a 6G network requires to support new "futuristic" use cases and apps. 

AT&T Does Not Use CBRS in Mobile Network; Verizon Does

Mobile operators (though not all) and cable operators (the biggest) were big winners of Citizens Broadband Radio Service 3.5 GHz mid-band spectrum in recent U.S. spectrum auctions. How that spectrum will be used might vary, however. 


source: futurithmic


AT&T does not actually use Citizens Broadband Radio Service (3.5 GHz) in its core mobile network. 


“The way we use CBRS today is mainly in part of our fixed wireless in rural America, as part of our CAF-II program,” says Igal Elbaz, AT&T SVP. But we didn't build CBRS on our standard macro,” he says.


AT&T also uses CBRS for private networks, though. 


Verizon, which has been capacity challenged, compared to T-Mobile and AT&T, does use CBRS in its core mobile network. Prior to the CBRS auction, Verizon had roughly the same amount of low-band and mid-band spectrum as did upstart Dish Network. 


source: Wells Fargo, LightReading 


It is not yet completely clear how Comcast will deploy CBRS. In principle, Comcast, Charter Communications or Cox Communications could use CBRS to construct local radio access networks of their own. They might federate their assets to create a larger footprint, as they did with their Wi-Fi services. 


What is unclear is whether fixed wireless services will be an early use case, as Charter seems to suggest will happen, or whether CBRS winds up being more useful as a way of shifting mobile traffic onto its own network, and off the wholesale Verizon network. That obviously would provide cost savings for the cable operators. 


Of course, longer term, CBRS might well wind up used in a variety of ways: supporting private networks, fixed wireless and in the core mobile network. As always, use cases will match the particular requirements for each operator. Verizon needs lots more mid-band spectrum, so its use in the core network makes sense.


AT&T has less need for using CBRS in its mobile network, in part because it is expected to bid extensively in the C-band auctions of mid-band spectrum as an alternative to adding CBRS spectrum. 


T-Mobile has such extensive mid-band capacity holdings that it is not expected to be active in acquiring additional mid-band assets. 


Tuesday, December 15, 2020

Iliad's Free Launches 5G at No Extra Cost

Iliad’s Free is launching 5G and offering the only plan available in France including 5G at no extra cost. Freebox subscribers also have unlimited mobile internet access.

source: Iliad


Free says it has the France’s largest 5G network, with 5,255 active sites covering almost 40 percent of the French population. 


source: Iliad


Free uses both 700-MHz frequencies and 3.5-GHz spectrum for capacity, with speeds as much as three times as fast as provided by 4G. 


South Korea and the 5G "Race"


Many seem to believe 5G is a "race" to see who can deploy it first. That never makes sense to me. Countries do not compete against each other based on the qualities of their mobile networks but on a range of other advantages or disadvantages. 

Ultimately, every country uses every generation of mobile technology. Some adopt faster; some slower. But overall advantages and disadvantages--economic, social, educational, technological--are fundamental. Use of mobile technology is only derivative. 

5G is more nearly a "race" between rival mobile service providers within a single country, who can gain or lose market share, customers, revenue and profit if they lag. 

For infrastructure suppliers, 5G is a race in the sense that market share always is an issue for contestants in any industry. Whether "leadership" in infrastructure supply is important is a matter of debate. 

In the internet era, some countries, firms and regions "lead" globally. So 5G arguably falls into the framework of leadership in other areas deemed to be important economically. Still, in and of itself, the availability of 5G might not matter very much in the near term.

The belief often seems to be that, longer term, early adopters are able to establish compelling new use cases and value, and reap the economic advantages of those use cases. That seems largely misplaced. 

It is not clear that the ability to use personal computers and PC apps has had decisive advantage, in and of itself. It is more nearly the case that economies and nations that were, for other reasons, able to innovate in the internet era. 

Some argue it is "experience" that matters for customers. Others might say "applications" are what matters. Yet others might say it is the value that can be created from 5G that ultimately matters (economic growth, for example). 

To be sure, 5G winners and losers will emerge. But winning countries are likely to be those who were winning for other reasons, though winning mobile operators will gain market share against rivals who are slow to adopt, in countries where consumers see compelling value. 




Monday, December 14, 2020

5G Growing 4X Faster than 4G?

Many observers would not be at all surprised if 5G subscriptions grew faster than did 4G, in the first five years. In part, that estimate would be in accord with mobile service adoption trends. For  each digital generation, subscription growth happened at a faster rate than for the prior generation. 


According to 5G Americas, there were 225 million 5G subscribers activated between the third quarter of 2019 and Q3 2020. That is four times the rate for 4G, 5G Americas argues. 


source: 5G Americas 


Most of that growth probably happened in China. 


source: 5G Americas 


As of December 2020, there were 229 million 5G subscriptions globally, which represents a 66 percent increase over last quarter, says 5G Americas.  


5G adoption rates generally have been expected to match or exceed 4G adoption rates.  


source: 5G Americas 



Still, 5G adoption rates might well vary by country, as such rates have occurred in the past.

Friday, December 11, 2020

Thinking about 6G Has Begun

6G remains a decade away. But it is safe to predict that 6G is coming, and will arrive in about 10 years. The mobile industry does so about every decade and 6G will not be an exception. 


source: Samsung

 

As prior digital generations have done, 6G will will address capacity by adding new spectrum. Each next-generation mobile network has done so. 


That means 6G will incorporate ways to commercially deploy even higher-frequency millimeter wave spectrum than used in the past. And, as has proven to be the case with millimeter wave frequencies generally, as well as beam-forming antennas and other sophisticated radio and modulation techniques, the application of cheap signal processing will make possible commercial use of frequencies that in the analog and early digital eras had been too costly for commercial communications use. 


Also, the 6G standard is likely to deepen the already prevalent trend towards heterogeneous access methods, allowing people and devices to use any available lawful resource for access purposes.

 source: Samsung 


It is fair to expect even better latency performance, which might seemingly reach limits as access link latency approaches single-digit levels in 5G, leaving little additional room for improvement. 


But it might not be unfair to think about the concept of negative latency, particularly as artificial intelligence becomes a commonplace means of anticipating issues and resolving them before users actually are aware potential problems exist. 


Better energy consumption on the part of both the network and devices should be expected. Operating costs, a drive towards carbon neutrality and need for devices that operate without batteries and the attendant operating cost implications also are likely.


Is Sora an "iPhone Moment?"

Sora is OpenAI’s new cutting-edge and possibly disruptive AI model that can generate realistic videos based on textual descriptions.  Perhap...