Early 5G in U.K. Now Offering Speeds 4X to Nearly 7X 4G Speed

Early mid-band 5G in the United Kingdom shows speeds twice to more than six times as fast as 4G, on U.K. networks.

In early tests, EE 5G downstream speeds in the fall of 2019 were running at 185.7 Mbps, while EE’s 4G offered  47.4 Mbps. So EE 5G was about four times faster than EE’s 4G. 

Vodafone 5G median downstream speed was 112.2Mbps. Median downstream 4G speed was 57.1 Mbps. So 5G on Vodafone’s networks is about twice as fast as 4G, according to RootMetrics tests in Birmingham between August and September 2019.

source: RootMetrics

Mid-band spectrum is touted for 5G as it offers better coverage than millimeter wave spectrum but higher bandwidth than low band frequencies. The latest Speedtest data on U.K. 5G show the results. Compared to “typical” speeds on all other U.K. mobile networks taken as a whole, 5G provides downstream speeds about 450 percent to 618 percent faster, according to Speedtest. 

Keep in mind: this does not mean 5G is that much faster than 4G. 5G is that much faster than the average of downstream data on all other networks in use. On the other hand, 5G arguably is just about that much faster than 4G. 

source: Speedtest

On EE, it is conceivable that 5G is about 400 percent faster than 4G. On 02, 5G might be 760 percent faster than 4G. 

source: Speedtest

5G Chicken-and-Egg Problems Not New for Telcos

Telcos and cable TV companies are used to solving chicken-and-egg problems. It is a problem they always face when building a new network or introducing a new service requiring customer purchase of new edge devices, and will be true of many 5G services.

Recall two old examples: facsimile machines or phone services are more valuable as the number of people using those devices and services grow plentiful. It’s a network effect, where the value of the network grows with the number of users. 

source: Strategic Toolkits

Chicken and egg strategy problems occur whenever the value proposition to two separate groups is dependent on adoption by the other. Developers will not have great incentives to create apps using 5G network capabilities until the network is substantially built out. Neither will potential users emerge until the apps and services are available. 

That is doubly true when network effects exist, where few suppliers mean few potential customers, but incentive to increase supply is quite low because there are so few customers.

Every next-generation communications network represents that sort of problem, and connectivity providers must always bet on the chickens. The network must be built; the investments made, before customers can buy. 

Some might note that there are a few time-tested ways to create value in the early going. One tactic used by connectivity providers is the “slow and steady” approach, building incrementally, on a geographic basis, neighborhood by neighborhood, and marketing to those neighborhoods as each new segment of the network is built.

Mass media marketing tends to be problematic in such cases, as touting the new service or app actually creates problems when customers are not yet able to buy. A historically useful tactic is neighborhood marketing, rather than reliance on “mass” techniques such as television or radio advertising. 

Many business services specialists use the same approach, building specialized networks in business parks or downtown cores, or building out to specific buildings, and then marketing only to those locations. Multi-city service providers often start in one city first, then expand to other areas. 

The “jumpstart” approach, useful when competitors are eyeing the same market, relies on shortcuts, such as leveraging another partner that already has a big customer base. It might seem odd that Google once used Netscape or Yahoo to gain entry into the search engine business, but that is what Google did. 

Etsy found that the people most likely to buy products on Etsy were those who also sold products on the site. This essentially turns the problem of satisfying two sets of marketplace participants into a single set, essentially. 

Likewise, Google created the Open Handset Alliance as a way of competing with Apple’s IoS ecosystem. 

A more expensive option is to subsidize consumption, as Uber did with discounted rider offers and driver payments. 

Service providers often partner when entering new markets, to build scale faster. That can mean use of channel partners, alliances with application providers, content or device suppliers to build scale faster, using existing customer bases to promote new apps and services. 

In other cases, some unorthodox way must be found to fulfill early transactions, before full-fledged systems can be built. 

Amazon made and fulfilled its own sales until it had enough volume to create the Amazon Marketplace, where it now acts to assist third parties in making their own sales. Airbnb used Craigslist to build its volume. 

Connectivity providers often rent or lease before building their own facilities, or use makeshift systems before at-scale in-house solutions can be created. 

Another approach might be to build one sort of business, adding to that initial business with other services or products, once the first is established. Some might point to Square, which started out offering low-cost credit card processing services for small businesses, but now operates several different businesses related to payroll, point-of-sale systems, e-commerce platforms and business loans. 

Yelp arguably did the same, beginning with information listings, and then gradually building an advertising business once scale was obtained. 

Another tactic is to focus on getting high-value users or customers onboard early, to attract others. OpenTable and Uber often are cited as having used that tactic.

5G is just one more instance of connectivity providers having to solve a chicken-and-egg problem.

5G Chickens Come Before Eggs

Big new markets always face a "chicken and egg" problem: investment is difficult when there are few potential customers, yet people have no incentives to buy when the service is not available. 

That is doubly true when network effects exist, where few suppliers mean few potential customers, but incentive to increase supply is quite low because there are so few customers.

Every next-generation communications network represents that sort of problem, and connectivity providers must always bet on the chickens.

The network must be built; the investments made, before customers can buy. That also means mobile operators have high incentives to get revenue-generating customers on that new network, as quickly as possible. That especially is important if the benefits to customers are not obvious.

You might find that an odd statement. Is not the whole point of 5G to supply order of magnitude faster speeds, to name one key example of benefit? The key word is "benefit."

So look at "benefits" from use of low-band, mid-band and millimeter wave spectrum. In many cases, low-band 5G will, in fact, not offer speeds much different--if different at all--from 4G. So there is little obvious consumer benefit.

Where mid-band or millimeter wave spectrum is used, speeds will be much higher, on the order of twice to three times as fast, on average. The problem? There are few consumer phone apps and use cases that can take advantage of such speed improvements. 

So supplier push rather than customer demand is going to drive early 5G subscriptions, since the experience advantages are going to be quite intangible in most cases. In many markets where 5G is launched using low-band spectrum, speed will not, in fact, be much different than fast 4G.

In other markets, where mid-band or millimeter wave spectrum also is used, consumer use cases will not be able to take advantage of the additional speed, with a few exceptions such as big file downloads, virtual reality or augmented reality. But how much time do consumers spend downloading big files? How many use AR or VR already?

And yet, 5G will be adopted. We do not know whether adoption will be faster, slower or at the same pace as 4G, but 5G will be purchased. One big reason is that 5G--though having eventual advantages for most consumers--has bigger advantages for mobile operators. 

Even if consumers might not experience much benefit at first, mobile operators will. 5G features lower cost per bit than 4G, which helps prop up the business model when more capacity has to be supplied at a relatively fixed price. 

Over the long term, consumers can only spend so much money on their communication services, so wallets budge only slowly, if at all. 

5G also has network features helpful either for cost containment or new service creation. 5G networks will be virtualized, which creates the ability to turn up virtual private networks (network slicing) rather easily, compared to legacy methods. 

In the radio network, virtualization allows mobile operators to contain radio network costs, since equipment from different suppliers increasingly will be mixed and matched with core networks. 

Longer term, 5G latency performance opens up potential space for a role in edge computing networks, vertical market applications and partnerships. 

And, as always, for some suppliers in each market, 5G represents a chance to preserve or upset existing market shares, if competitors cannot easily keep pace. 

Hence the paradox: Though 5G is, by design, supposed to have performance advantages over 4G for mobile service provides and customers, the early advantages will rarely be sufficient to drive consumer demand. Instead, supplier push will be at work. 

For customers, the advantages include faster top speeds and lower latency.  For mobile operators, the advantages include faster top speeds, lower latency, virtualized networks, which promise new features and lower cost, higher device density (supporting lots of sensors and internet of things devices) and lower cost per bit. 

That list is a clue to why 5G will be adopted, perhaps even more rapidly than 4G: it benefits mobile operators more than consumers. In fact, an argument might be made that 5G benefits mobile operators almost to the exclusion of consumers, at first. In the early going, using low-band spectrum, 5G might not be noticeably faster than 4G. 

When mid-band and millimeter wave spectrum is used, speeds will be much faster than 4G, but in ways that do not actually benefit the use cases most people have for their phones. Big file downloads will be faster, but what percentage of time do people spend downloading big files? 

At first, the early adopter desire to "get it first," where the driver is as much status as anything else, will be the driver. But there will be indirect drivers as well, such as the value of "5G comes with your new service plan." In many cases, 5G will be a "nice to have" attribute of a service plan, but it is the service plan that drives the switch. 

Perhaps that will change over time, as new use cases develop. But it might also be the case that 5G gets adopted because it provides value for mobile operators, who will create incentives to adopt 5G, even if the actual experience advantages might be hard to demonstrate. 

This can be seen in recent Opensignal measurements of speeds on new 5G networks. Where early tests of U.S. 5G rely exclusively on millimeter wave spectrum, the U.S. has the highest speed. 

Aside from big file downloads, the experience advantages are almost impossible to demonstrate. Faster might be better, but actual consumer smartphone apps cannot actually take advantage of the faster speeds, yet, since virtual reality and augmented reality are not yet widely used. 

Conversely, not that in some markets where low-band 5G has been launched, the speeds are almost identical to 4G, so there is no actual experience advantage. 

source: Opensignal

Compared to existing 4G, 5G in early days has doubled to nearly tripled real-world speeds in some cases, but had almost no impact on speed in a few cases. Again, the choice of spectrum, or availability, really do matter. 

Where low-band spectrum is the 5G choice, it sometimes does not life speeds very much. Where millimeter or mid-band spectrum is used, 5G speed advantages are clear, but app performance and user experience do not change much, with the exception of big file downloads.

source: Opensignal

Should Consumers Buy 5G Now?

A growing number of mobile consumers will have the option of buying 5G phones and 5G service in coming months. Whether they decide they should do so is the big question.

If I had to guess right now, I’d predict that much early consumer adoption of mobile 5G will be driven indirectly, the result of choices made about service plans and high-end phones, not 5G service specifically. 

There are, of course, value drivers unrelated to technical performance, namely the early adopter desire to first use a new product with the best specs. 

In fact, in some markets, where 5G is based on low-band spectrum, the speed advantage of 5G might be quite muted and hardly better than 4G.

In the United Kingdom and Spain, where 5G initially was launched using low-band spectrum, there was only a slight improvement in experienced internet access speeds. Under those conditions, it is hard to say what the direct value proposition might be. 

source: Opensignal

In markets where 5G initially is launched using mid-band spectrum, consumers might have a relatively straightforward value proposition: slightly higher cost, significantly-higher bandwidth, even if the use cases remain few. 

The point is that, in all markets, most consumers might struggle to find a direct value proposition unrelated to status or bragging rights, in the early going, as there are very-few consumer mobile applications or use cases that benefit from near-gigabit-per-second speeds, or even average speeds in the 100-Mbps range. 

Consumers who do lots of downloading on their mobiles are about the only users who will--at first--be able to take advantage of higher 5G speeds on mid-band or millimeter wave networks. 

Ignoring for the moment the few use cases where 5G speeds in the 100-Mbps to near-gigabit range are relevant, indirect drivers are likely going to be more important than direct value propositions, for most customers. 

The consumer direct value proposition might be low-band 5G, which offers slight or possibly even no real advantage over 4G; mid-band 5G, which offers quantifiable speed advantages, but few experience advantages; or high-band 5G, which offers huge speed advantages, but few experience benefits over 4G. 

Beyond that, some customers, including those in the U.S. market, will have to choose between 5G phones supporting low-band 5G or millimeter wave 5G. The early devices will not support both low-band and high-band. 

All of that suggests to me that indirect value will drive most 5G subscriptions. Customers will get 5G as a byproduct of their service plan or device choices.

AT&T to Launch Consumer 5G in First Half of 2020

AT&T plans to launch consumer 5G services in the first half of 2020, after a year of selling only to businesses. For mass market customers, AT&T plans to rely on low-band spectrum, using millimeter wave high-band spectrum for capacity reinforcement at large venues. 

In the coming weeks, consumer 5G service will launch over low-band spectrum in the Indianapolis, Pittsburgh, Providence, R.I., Rochester, N.Y., and San Diego market areas. AT&T plans to follow with  Boston, Las Vegas, Milwaukee, New York City, San Francisco, Birmingham, Ala., Bridgeport, Conn., Buffalo, N.Y., Louisville, Ky., and San Jose, Calif. 

source: AT&T

AT&T initially is focusing on coverage, rather than capacity, so although consumers in the new markets will see 5G speed boosts over 4G levels, they will not see the highest speeds available on the millimeter wave networks. 

Still, availability is likely to outstrip subscribership for some time, as past experience suggests fewer than half of existing customers able to upgrade to 5G will do so, about five to six years after launch in any market. 

In the 4G era, seven years after launch, about 38 percent of potential customers actually had signed up. Some believe that percentage could be higher in the 5G era, but there also are arguments that adoption could be slower, in part because of the controlled roll-out.

source: GSMA Intelligence

To be sure, low-band 5G should offer some speed advantages over 4G. But not always. Early evidence from some markets using low-band spectrum suggests users might not perceive much of a difference. 

In the United Kingdom and Spain, where 5G initially was launched using low-band spectrum, there was only a slight improvement in experienced internet access speeds. 

source: Opensignal

To the extent that consumers become aware of the differences between 5G and 4G, and when low-band spectrum does not offer startling speed advantages, it is not so clear that demand will be robust. 

Also, to the extent that good 4G service supports all the apps and use cases most consumers have, it is not so clear what compelling advantage 5G might offer, except indirectly. The actual reason for switching to 5G could come with a change of service plans or purchase of a new high-end device. 

AT&T, for example, seems to be offering consumer 5G access when customers buy its higher-cost unlimited usage plans. access to 5G service in the below plans at no extra cost when we roll out 5G in the coming weeks.

The AT&T Unlimited Elite plan provides 30 gigabytes (GB) of mobile hotspot data per line, 100 GB per line of unthrottled data, HBO and HD streaming all for $50 a month per line on four lines after autopay and paperless bill discounts.

Much More Data, Same Price: The Enduring Telecom Challenge

There are many reasons why communications networks--especially those serving consumer customers--must constantly drive capital investment and operating costs down. But data consumption and ability to pay are the two biggest reasons. The former is unlimited; the latter sharply limited. 

Data consumption simply keeps growing, Cisco Visual Networking Index always shows. 

source: Cisco

source: Science Direct

Consumer spending is quite limited. As data from the United Kingdom shows, consumer spending on mobility, mobile data or communications services does not change much from year to year. Data from the U.S. market shows similar trends. 

The big takeaway is that consumer budgets tend to be highly fixed. 

So the big business challenge for mobile operators is finding ways to increase the supply of bandwidth, for roughly the same retail prices, while maintaining profit margins. 

And 5G, though touted for enabling many new use cases and applications, actually is one solution to the need for lower capex and opex. 

To be sure, 4G can keep supplying additional capacity, up to a point. A few technologies, better radios and small cell architectures are key to that effort.

Among the technology improvements, many would point to MIMO radios, 256 QAM for more intense modulation and carrier aggregation (for 60 MHz bandwidth or more) as ways to get to gigabit 4G.

The 3GPP standards for 4G support aggregating up to 32 carriers for LTE. Not only does that create bigger channels that support more bandwidth, and therefore higher speeds, but doing so also increases efficiency, eliminating bandwidth that otherwise would be wasted for guard bands. 

Operators are deploying massive MIMO antenna configurations with up to 128 antenna elements (64 transmit and 64 receive), which increases bandwidth by essentially creating multiple parallel channels where a single-antenna system has but one channel. 

Radios also are using more heavily sectorized radios, which supply more effective bandwidth by reducing interference. Where traditionally cell tower radios have used three sectors, 4G cell sites now use six sectors or nine sectors.

source: Commscope

Denser networks using smaller cells also are part of the solution, creating more effective bandwidth by more intensively reusing existing spectrum. 

So one might ask why not simply use 4G and add more new spectrum. Part of the issue is that the 4G standard only includes frequencies up to 5 GHz, not any frequencies above that range. Standards-compliant network elements and platforms therefore cannot be purchased to support spectrum above 5 GHz. 

The point is that 4G can continue to increase bandwidth, using a number of known technologies and architectures, but only up to a point. We cannot use spectrum above 5 GHz for 4G networks, by design. 

For that reason, LTE capacity will typically saturate for mobile operators from around 2020. And that is why some mobile operators already are moving to add 5G capacity. 

source: Nokia

The cost of bandwidth--not just its supply--is another reason for the planned switch to 5G. As users continue to increase their bandwidth consumption, their propensity to pay does not. 

In fact, the fundamental reality is that more bandwidth must be supplied at roughly the same retail price, all other things being equal.

So 5G makes sense because it offers a cost per gigabyte at least 10 times lower than 4G, to start, and using a range of technologies and architectures might conceivably get to cost per gigabyte 100 times lower than 4G. That will especially be possible using millimeter wave spectrum. 

source: Mobile Experts

Early Millimeter Deployment Might Remind You of Early 1G

This map of Verizon’s 5G deployment in downtown Denver illustrates the use of 5G as an overlay on 4G for capacity reinforcement. The map also illustrates a feature of millimeter wave spectrum when used to support 5G or other networks: it does not propagate as well as mid-band or low-band frequencies.

Note the coverage basically outdoors, along streets and sidewalks. For some of you, that might harken back to the early days of analog, first-generation cellular service, which essentially was a "call from your car" service. That being the case, it made sense to build first along major streets and highways.

Note that 4G coverage is shown as pervasive. That illustrates the importance of non-millimeter wave frequencies to supply coverage, especially indoors. 

source: Verizon