A New Cellular Carrier?

One of the most interesting aspects of the proposed merger of Sprint and T-Mobile is that the agreement now includes selling some of Sprint’s spectrum to Dish Networks to enable them to become a 5G cellular provider. This arrangement is part of the compromise required by the Department of Justice to preserve industry competition when the major wireless carriers shrink from four to three.

This agreement would have Dish Networks paying $5 billion for the spectrum assets, which complement the spectrum already owned by Dish. The agreement also includes an MVNO agreement between Dish Networks and T-Mobile that would let Dish enter the cellular market immediately before having to build any network. As part of that arrangement, Dish would purchase Boost Mobile from T-Mobile for $1.4 billion, providing them with an immediate base of cellular customers.

Dish already owns spectrum valued at several billion dollars. The company has been under pressure from the FCC to deploy that spectrum, and Dish recently began building a nationwide narrowband network to support IoT sensors. The company admits they are not happy with the IoT sensor business plan but didn’t want to lose their spectrum. Perhaps the best aspect of this deal from Dish’s perspective is that they are being given a new time clock to use existing spectrum in a more profitable way.

This deal has plenty of critics who don’t believe that Dish can turn into a viable competitor. This includes numerous consumer groups as well as a group of state Attorney Generals who have filed to block the merger. The merger is far from a done deal and is going to court, although it has crossed the major hurdles of getting DOJ approval and informal approval from the FCC.

Dish Chairman Charlie Ergen says the company is ready to become the fourth facility-based cellular carrier in the market. He thinks that launching with a new 5G network will provide some advantages over carriers that will be upgrading older networks. The company faces some significant challenges such as gaining access to tower space in crowded markets. The other cellular carriers have also been busy and have invested significant amounts of capital in building fiber to support cellular small cell sites.

The challenge of building a new nationwide cellular network from scratch is intimidating. As a satellite provider, the company does not already operate an extensive landline network. The logistics of hiring the needed talent and constructing the core network infrastructure is a major challenge. A few years ago Dish had estimated the cost to build a nationwide cellular network at $10 billion. The company says they have already released an RFI and an RFP to start the process of hiring contractors to build the new network.

Ergen says the company could build the core network in 2020 and could construct a network to cover 70% of the homes in the country by 2023. As far as being competitive, Dish says they would enter the market with ‘disruptive’ pricing to capture market share.

Dish needs something like this if it is to survive. The company lost over 1.1 million satellite TV customers last year, a little over 10% of its customer base. It looks like cord cutting is accelerating this year and one has to wonder how long they will remain as a viable business.

Interestingly, Dish won’t be the only new competitor in the cellular market. Comcast recently spent over $1.7 billion on spectrum. The company has been reselling cellular service and offering low-price broadband as part of its bundle for the last few years. The company reporting hitting 1.2 million cellular customers at the beginning of this year. While Comcast is not likely to tackle building a nationwide network, they could become a formidable competitor in the urban markets where they are already the cable provider. Other cellular companies like Charter and Altice are considering a similar path.

Cellular Broadband Speeds – 2019

Opensignal recently released their latest report on worldwide cellular data speeds. The company examined over 139 billion cellphone connections in 87 countries in creating this latest report.

South Korea continues to have the fastest cellular coverage in the world with an average download speed of 52.4 Mbps. Norway is second at 48.2 Mbps and Canada third at 42.5 Mbps. The US was far down the list in 30th place with an average download speed of 21.3 Mbps. Our other neighbor Mexico had an average download speed of 14.9 Mbps. At the bottom of the list are Iraq (1.6 Mbps), Algeria (2.1 Mbps) and Nepal (4.4 Mbps). Note that these average speeds represent all types of cellular data connections including 2G and 3G.

Cellular broadband speeds have been improving raoidly in most countries. For instance, in the 2017 report, Opensignal showed South Korea at 37.5 Mbps and Norway at 34.8 Mbps. The US in 2017 was in 36th place at only 12.5 Mbps.

Earlier this year Opensignal released their detailed report about the state of mobile broadband in the United States. This report looks at speeds by carrier and also by major metropolitan area. The US cellular carriers have made big strides just since 2017. The following table compares download speeds for 4G LTE by US carrier for 2017 and 2019.

2019 2017
Download Latency Download Latency
AT&T 17.8 Mbps 57.8 ms 12.9 Mbps 63.8 ms
Sprint 13.9 Mbps 70.0 ms 9.8 Mbps 70.1 ms
T-Mobile 21.1 Mbps 60.6 ms 17.5 Mbps 62.8 ms
Verizon 20.9 Mbps 62.6 ms 14.9 Mbps 67.3 ms

Speeds are up across the board. Sprint increased speeds over the two years by 40%. Latency for 4G is still relatively high. For comparison, fiber-to-the-home networks have latency in the range of 10 ms and coaxial cable networks have latency between 25 – 40 ms. The poor latency in cellular networks is one of the reasons why browsing the web on a cellphone seems so slow. (the other reason is that cellphone browsers focus on graphics rather than speed).

Cellular upload speeds are still slow. In the 2019 tests, the average upload speeds were AT&T (4.6 Mbps), Sprint (2.4 Mbps), T-Mobile (6.7 Mbps) and Verizon (7.0 Mbps).

Speeds vary widely by carrier and city. The fastest cellular broadband market identified in the 2019 tests was T-Mobile in Grand Rapids, Michigan with an average 4G speed of 38.3 Mbps. The fastest upload speed was provided by Verizon in New York City at 12.5 Mbps. Speeds vary by market for several reasons. First, the carriers don’t deploy the same spectrum everywhere in the US, so some markets have less spectrum than others. Markets vary in speed due to the state of upgrades – at any given time cell sites are at different levels of software and hardware upgrades. Finally, markets also vary by cell tower density and markets that serve more customers for each tower are likely to be slower.

Many people routinely take speed tests for their home landline broadband connection. If you’ve not taken a cellular speed test it’s an interesting experience. I’ve always found that speeds vary significantly with each speed test, even when run back-to-back As I was writing this blog I took several speed tests that varied in download speeds between 12 Mbps and 23 Mbps (I use AT&T). My upload speeds also varied with a top speed of 3 Mbps, and one test that couldn’t maintain the upload connection and measured 0.1 Mbps on the test. While landlines broadband connections maintain a steady connection to an ISP, a cellphone establishes a new connection every time you try to download and speeds can vary depending upon the cell site and the channel your phone connects to and the overall traffic at the cell site at the time of connection. Cellular speeds can also be affected by temperature, precipitation and all of those factors that make wireless coverage a bit squirrelly.

It’s going to be a few years until we see any impact on the speed test results from 5G. As you can see by comparing to other countries, the US still has a long way to go to bring 4G networks up to snuff. One of the most interesting aspects of 5G is that speed tests might lose some of their importance. With frequency slicing, a cell site will size a data channel to meet a specific customer need. Somebody downloading a large software update should be assigned a bigger data channel with 5G than somebody who’s just keeping up with sports scores. It will be interesting to see how Opensignal accounts for data slicing.

Are Broadband Investments Increasing?

The largest ISPs and their lobbying arm USTelecom are still claiming that the level of industry capital spending has improved as a direct result of the end of Title II regulation. In a recent blog they argue that capital spending was up in 2018 due to the end of regulation – something they describe as a “forward-looking regulatory framework”. In reality, the new regulatory regime is now zero regulation since the FCC stripped themselves of the ability to change ISP behavior for broadband products and practices.

The big ISPs used this same argument for years leading up to deregulation. They claimed that ISPs held back on investments since they were hesitant to invest in a regulatory-heavy environment. This argument never held water for a few reasons. First, the FCC barely ever regulated broadband companies. Since the advent of DSL and cable modems in the late 1990s, each subsequent FCC has largely been hands-off with the ISP industry.

The one area where the last FCC added some regulations was with net neutrality. According to USTelecom that was crippling regulation. In reality, the CEO of every big telco and cable company has publicly stated that they could live with the basic principles of net neutrality. The one area of regulation that has always worried the big ISPs is some kind of price regulation. That’s really not been needed in the past, but all of the big companies look into the future and realize that the time will come when they will probably raise broadband rates every year. We are now seeing the beginnings of that trend, which is probably why USTelecom keeps beating this particular dead horse to death – the ISPs are petrified of rate regulation of any kind.

The argument that the big ISPs held back on investment due to heavy regulation has never had any semblance to reality. The fact is that the big ISPs make investments for the same reasons as any large corporation – to increase revenues, to reduce operating costs, or to protect markets.

As an example, AT&T has been required to build fiber past 12.5 million passings as part of the settlement reached that allowed them to buy DirecTV. AT&T grabbed that mandate with gusto and has been aggressively building fiber for the past several years and selling fiber broadband. Both AT&T and Verizon have also been building fiber to cut transport expense to cell sites – they are building where that transport is too costly, or where they know they want to install small cell sites. The large cable companies all spent capital on DOCSIS 3.1 for the last few years to boost broadband speeds to protect and nurture their growing monopoly of urban broadband. All of these investment decisions were made for strategic business reasons that didn’t consider the difference between light regulation and no regulation. Any big ISP that says they will forego a strategic investment due to regulation would probably see their stock price tumble.

As a numbers guy, I always become instantly suspicious of deceptive graphs. Consider the graph included in the latest USTelecom blog. It shows the levels of industry capital investments made between 2014 and 2018. The graph makes the swings of investment by year look big due to the graphing trick of starting the bottom of the graph at $66 billion instead of at zero. The fact is that 2018 capital investments are less than 3% higher than the investments made in 2014. This is an industry where the aggregate level of annual investment varies by only a few percent per year – the argument that the ISPs have been unleashed due to the end of Title II regulation is laughable and the numbers don’t show it.

There are always stories every year that can explain the annual fluctuation in industry spending. Here are just a few things that made an significant impact on the aggregate spending in the past few years:

  • Sprint had a cash crunch a few years ago and drastically cut capital spending. One of the primary reasons for the higher 2018 spending is that Sprint spent almost $2 billion more in 2018 than the year before as they try to catch up on neglected projects.
  • AT&T spent $2 billion in 2018 for FirstNet, the nationwide public safety network. But AT&T is not spending their own money – that project is being funded by the federal government and ought to be removed from these charts.
  • Another $3 billion of AT&T’s spending in 2018 was to beef up the 4G network in Mexico. I’m not sure how including that spending in the numbers has any relevance to US regulation.
  • AT&T has been on a tear building fiber for the past four years – but they announced last month that the big construction push is over, and they will see lower capital spending in future years. AT&T has the largest capital budget in the industry and spent 30% of the industry wide $75 billion in 2018 – how will USTelecom paint the picture next year after a sizable decrease in AT&T spending?

The fact that USTelecom keeps harping on this talking point means they must fear some return to regulation. We are seeing Congress seriously considering new consumer privacy rules that would restrict the ability of ISPs to monetize customer data. We know it’s likely that if the Democrats take back the White House and the Senate that net neutrality and the regulation of broadband will be reinstated. For now, the big ISPs have clearly and completely won the regulatory battle and broadband is as close to deregulated as any industry can be. Sticking with this false narrative can only mean that the big ISPs think their win is temporary.

Clearing Mid-range Spectrum

The FCC is in the process of trying to free up mid-range spectrum for 5G. They just opened a Notice of Proposed Rulemaking looking at 2.5 GHz spectrum, located in the contiguous block between 2495 and 2690 MHz. Overall this is the largest contiguous block of mid-range spectrum. Over half of the spectrum sits idle today, particularly in rural America. The history of this spectrum demonstrates the complications involved in trying to reposition spectrum for broadband and cellular use.

The frequency was first assigned by the FCC in 1963 when it was made available to school systems to transmit educational TV between multiple schools. The spectrum band was called Instructional Television Fixed Service (ITFS). The band was divided into twenty channels and could transmit a TV signal up to about 35 miles. I grew up in a school system that used the technology and from elementary school onward we had a number of classes taught on the TV. Implementing the technology was expensive and much of the spectrum was never claimed.

In 1972 the FCC recognized the underuse of the spectrum and allowed commercial operators to use the bands of 2150 to 2162 MHz on an unlicensed basis for pay-TV transmissions to rooftop antennas. The spectrum could only carry a few TV channels and in the 1970s was used in many markets to transmit the early version of HBO and Nickelodeon. This spectrum band was known as Multipoint Distribution Service (MDS) and also was good for about 35 miles.

Reacting to pressure from cellular companies, the FCC reallocated eight additional channels of the spectrum for commercial use. Added to the MDS spectrum this became known as Multichannel Multipoint Distribution Service (MMDS). At the time this displaced a few school systems and anybody using the spectrum had to pay to move a school system to another workable channel. This spectrum was granted upon request to operators for specific markets.

In 1991 the FCC changed the rules for MMDS and allowed the channels to be used to transmit commercial TV signals. In 1995 any unused MMDS spectrum was sold under one of the first FCC auctions, which was the first to divide service areas into the geographic areas known as Basic Trading Areas (or BTAs) that are still used today. Before this auction, the spectrum was awarded in 35-mile circles called Geographic Service Areas (GSAs). The existing GSAs were left in place and the spectrum sold at auction had to work around existing GSAs.

The FCC started getting pressure from wireless companies to allow for the two-way transmission of data in the frequency (up to now it had been all one-way delivery to a customer site). In 2005 the FCC changed the rules and renamed the block of spectrum as Broadband Radio Service (BRS). This added significant value to licenses since the spectrum could now be repositioned for cellular usage.

At this point, Clearwire entered the picture and saw the value of the spectrum. They offered to buy or lease the spectrum from school systems at prices far lower than market value and were able to amass the right to use a huge amount of the spectrum nationwide. Clearwire never activated much of the spectrum and was in danger of losing the rights to use it. In 2013 Sprint purchased Clearwire, and Sprint is the only cellular company using the spectrum band today.

Today the spectrum band has all sorts of users. There are still school districts using the spectrum to transmit cable TV. There are still license holders who never stopped using the 35-mile GSA areas. There are still MMDS license holders who found a commercial use for the spectrum. And Sprint holds much of the spectrum not held by these other parties.

The FCC is wrestling in the NPRM with how to undo the history of the spectrum to make it more valuable to the industry. Education advocates are still hoping to play in the space since much of the spectrum sits idle in rural America (as is true with a lot of cellular and other mid-range spectrum). The other cellular carriers would like to see chunks of the spectrum sold at auction. Other existing license holders are fighting to extract the biggest value out of any change of control of the spectrum.

The challenge for repositioning this spectrum is complicated because the deployment of the spectrum differs widely today by market. The FCC is struggling to find an easy set of rules to put the genie back in the bottle and start over again. In terms of value for 5G, this spectrum sits in a sweet spot in terms of coverage characteristics. Using the spectrum for cellular data is probably the best use of the spectrum, but the FCC has to step carefully to do this in such a way as to not end up in court cases for years disputing any order. Reallocating spectrum is probably the most difficult thing the FCC does and it’s not hard to see why when you look at the history of this particular block of spectrum and realize that every block of spectrum has a similar messy past.

ISPs Are Violating the Old Net Neutrality Rules

It’s been just over a year since the FCC repealed net neutrality. The FCC’s case is being appealed and oral arguments are underway in the appeal as I write this blog. One would have to assume that until that appeal is finished that the big ISPs will be on their best behavior. Even so, the press has covered a number of ISP actions during the last year that would have violated net neutrality if the old rules were still in place.

It’s not surprising that the cellular carriers were the first ones to violate the old net neutrality rules. This is the most competitive part of the industry and the cellular carriers are not going to miss any opportunity to gain a marketing edge.

AT&T is openly advertising that cellular customers can stream the company’s DirecTV Now product without it counting against monthly data caps. Meanwhile, all of the competing video services like Sling TV, Paystation Vue, YouTube TV, Netflix or Amazon Prime count against AT&T data caps – and video can quickly kill a monthly data plan download allotment. AT&T’s behavior is almost a pure textbook example of why net neutrality rules were put into place – to stop ISPs from putting competitor’s products at an automatic disadvantage. AT&T is the biggest cellular provider in the country and this creates a huge advantage for DirecTV Now. All of the major cellular carriers are doing something similar in allowing some video to not count against the monthly data cap, but AT&T is the only one pushing their own video product.

In November a large study of 100,000 cellphone users by Northeastern University and the University of Massachusetts showed that Sprint was throttling Skype. This is not something that the carrier announced, but it’s a clear case of pushing web traffic to the ‘Internet slow lane’. We can only speculate why Sprint would do this, but regardless of their motivation this is clearly a violation of net neutrality.

This same study showed numerous incidents where all of the major cellular carriers throttled video services at times. YouTube was the number one target of throttling, followed by Netflix, Amazon Prime, and the NBC Sports app. This throttling wasn’t as widespread as Sprint’s throttling of Skype, but the carriers must have algorithms in their network that throttles specific video traffic when cell sites get busy. In contrast to the big carriers, the smaller independent cellular carrier C.Spire had almost no instances of differentiation among video streams.

Practices that might violate net neutrality were not limited to cellular carriers. For example, Verizon FiOS recently began giving free Netflix for a year to new broadband customers. AT&T also started giving out free HBO to new customers last year. This practice is more subtle than the cellular carrier practice of blocking or throttling content. One of the purposes of net neutrality was for ISPs to not discriminate against web traffic. By giving away free video services the landline broadband companies are promoting specific web services over competitors.

This doesn’t sound harmful, but the discussions in the net neutrality order warned about a future where the biggest ISPs would partner with a handful of big web services like Facebook or Netflix to the detriment of all smaller and start-up web services. A new video service will have a much harder time gaining customers if the biggest ISPs are giving away their competitors for free.

There are probably more bad practices going on that we don’t know about. We wouldn’t have known about the cellular throttling of services without the big study. A lot of discrimination can be done through the network routing practices of the ISPs, which are hard to prove. For example, I’ve been seeing a growing number of complaints from consumers recently who are having trouble with streaming video services. If you recall, net neutrality first gained traction when it became known that the big ISPs like Comcast were blatantly interfering with Netflix streaming. There is nothing today to stop the big ISPs from implementing network practices that degrade certain kinds of traffic. There is also nothing stopping them from demanding payments from web services like Netflix so that their product is delivered cleanly.

Interestingly, most of the big ISPs made a public pledge to not violate the spirit of net neutrality even if the rules were abolished. That seems to be a hollow promise that was to soothe the public that worried about the end if net neutrality. The FCC implemented net neutrality to protect the open Internet. The biggest ISPs have virtual monopolies in most markets and public opinion is rarely going to change an ISP behavior if the ISP decides that the monetary gain is worth the public unhappiness. Broadband customers don’t have a lot of options to change providers and Cable broadband is becoming a near-monopoly in urban areas. There is no way for a consumer to avoid the bad practices of the cellular companies if they all engage in the same bad practices.

There is at least some chance that the courts will overturn the FCC repeal of net neutrality, but that seems unlikely to me. If the ISPs win in court and start blocking traffic and discriminating against web traffic it does seem likely that some future FCC or Congress will reinstitute net neutrality and starts the fight all over again. Regardless of the court’s decision, I think we are a long way from hearing the last about net neutrality.

2.5 GHz – Spectrum for Homework

As part of the effort to free up mid-band spectrum, the FCC is taking a fresh look at the 2.5 GHz spectrum band. This band of spectrum is divided into 33 channels; the lower 16 channels are designated as EBS (Educational Broadband Service) with the remainder as BRS (Broadcast Radio Service).

The EBS band was first granted to educational institutions in 1963 under the designation ITFS (Instructional Television Fixed Service) and was used to transmit educational videos within school systems. It became clear that many schools were not using the spectrum and the FCC gave schools the authority to lease excess capacity on the spectrum for commercial use. In urban markets the spectrum was leased to networks like HBO, Showtime and the Movie Channel which used the spectrum to delivery content after the end of the school day. In the late 1990s the spectrum was combined with MMDS in an attempt to create a wireless cable TV product, but this use of the spectrum never gained commercial traction.

In 1998 the FCC allowed cellular companies to use the leased spectrum for the new 3G cellular. In 1998 the FCC also stopped issuing new licenses for the spectrum band. Companies like Craig McCaw’s Clearwire leased the spectrum to deliver competitive cellular service in many urban areas. In 2005 the FCC cemented this use to allow the spectrum to be used for two-way mobile and fixed data.

Today the technology has improved to the point where the spectrum could help to solve the homework gap in much of rural America. The spectrum can be used in small rural towns to create hot spots that are tied directly to school servers. The spectrum can also be beamed for about 6 miles from tall towers to reach remote students. The spectrum has nearly the same operating characteristics as the nearby 2.4 GHz WiFi band, meaning that long-distance connections require line-of-sight, so the spectrum is more useful is areas with wide-open vistas than in places like Appalachia.

A group of educational organizations including the Catholic Technology Network, the National EBS Association, the Wireless Communications Association International and the Hispanic Information and Telecommunications Network petitioned the FCC to expand the EBS network and to grant new EBS licenses to fully cover the country. The FCC has been considering a plan that would strengthen the educational use of the spectrum and which would also auction the rest of the spectrum for use as wireless broadband.

The use of the spectrum for rural educational uses could be transformational. Rural students could get a small dish at their homes, like is done with the fixed wireless deployed by WISPs. Students would them have a direct connection to the school systems servers for doing homework. Interestingly, this would not provide a home with regular Internet access, other than what might be granted by schools for links needed for doing homework.

The disposition of the spectrum band is complicated by the fact that Sprint holds much of the spectrum under long-term lease. Sprint holds licenses to use more than 150 MHz of the spectrum in the top 100 markets in the country, which currently provides them with enough spectrum to simultaneously support both 4G LTE and 5G. The speculation is that the FCC is working on a plan to free up some of this spectrum as a condition to the merger of Sprint and T-Mobile.

This is the only current spectrum band where the FCC is envisioning different urban and rural uses, with rural parts of the country able to use the spectrum to connect to students while in urban areas the spectrum is used to support 5G. This divided use was only made possible by the historic educational component of the spectrum. If the FCC tries to give all of this spectrum to the cellular carriers they’d have to reclaim the 2,200 licenses already given to school systems – something they are politically unwilling to tackle.

However, this solution points to a wider solution for rural residential broadband. The FCC could order the same type of rural/urban bifurcation for many other bands of spectrum that are used primarily in urban settings. We need to find creative ways to use idle spectrum, and this spectrum bank provides a roadmap that ought to be applied to other swaths of spectrum.

Freeing the spectrum for full use by rural education offers big potential, but also creates challenges for rural school systems which will have to find the money to build and deploy wireless networks for homework. But solving the rural homework gap is compelling and I’m sure many school districts will tackle the issue with gusto.

5G Cellular for Home Broadband?

Sprint and T-Mobile just filed a lengthy document at the FCC that describes the benefits of allowing the two companies to merge. This kind of filing is required for any merger that needs FCC approval. The FCC immediately opened a docket on the merger and anybody that opposes the merger can make counterarguments to any of the claims made by the two companies.

The two companies decided to highlight a claim that the combined Sprint and T-Mobile will be able to roll out a 5G network that can compete with home broadband. They claim that by 2024 they could gain as much as a 7% total market penetration, making them the fourth biggest ISP in the country.

The filing claims that their 5G network will provide a low-latency broadband product with speeds in excess of 100 Mbps within a ‘few years’. They claim that customers will be able to drop their landline broadband connection and tether their home network to their unlimited cellular data plan instead. Their filing claims that the this will only be possible with a merger. I see a lot of holes that can be poked into this claim:

Will it Really be that Fast? The 5G cellular standard calls for eventual speeds of 100 Mbps. If 5G follows the development path of 3G and 4G, then those speeds probably won’t be fully met until near the end of the next decade. Even if 5G network can achieve 100 Mbps in ideal conditions there is still a huge challenge to meet those speeds in the wild. The 5G standard achieves 100 Mbps by bonding multiple wireless paths, using different frequencies and different towers to reach a customer. Most places are not receiving true 4G speeds today and there is no reason to think that using a more complicated delivery mechanism is going to make this easier.

Cellphone Coverage is Wonky.  What is never discussed when talking about 5G is how wonky all wireless technologies are in the real world. Distance from the cell site is a huge issue, particular with some of the higher frequencies that might be used with 5G. More important is local interference and propagation. As an example, I live in Asheville, NC. It’s a hilly and wooded town and at my house I have decent AT&T coverage, but Verizon sometimes has zero bars. I only have to go a few blocks to find the opposite situation where Verizon is strong and AT&T doesn’t work. 5G is not going to automatically overcome all of the topographical and interference issues that affect cellular coverage.

Would Require Significant Deployment of Small Cell Sites. To achieve the 100 Mbps in enough places to be a serious ISP is going to require a huge deployment of small cell sites, and that means the deployment of a lot of fiber. This is going to be a huge hurdle for any wireless company that doesn’t have a huge capital budget for fiber. Many analysts still believe that this might be a big enough hurdle to quash a lot of the grandiose 5G plans.

A Huge Increase in Wireless Data Usage. Using the cellular network to provide the equivalent of landline data means a magnitude increase in the bandwidth that will be carried by the cellular networks. FierceWireless along with Strategic Analytics recently did a study on how the customers of the major cellular companies use data. They reported that the average T-Mobile customer today uses 18.4 GB of data per month with 5.3 GB on the cellular network and the rest on WiFi. Sprint customers use 18.2 GB per month with 4.4 GB on the cellular networks. Last year Cisco reported that the average residential landline connection used over 120 GB per month – a number that is doubling every three or four years. Are cellular networks really going to be able to absorb a twenty or thirty times increase in bandwidth demand? That will require massive increases in backhaul bandwidth costs along with huge capital expenditures to avoid bottlenecks in the networks.

Data Caps are an Issue.  None of the cellular carriers offers truly unlimited data today. T-Mobile is the closest, but their plan begins throttling data speeds when a customer hits 50 GB in a month. Sprint is stingier and is closer to AT&T and Verizon and starts throttling data speeds when a customer hits 23 GB in a month. These caps are in place to restrict data usage on the network (as opposed to the ISP data caps that are meant to generate revenue). Changing to 5G is not going to eliminate network bottlenecks, particularly if we see millions of customers using cellular networks instead of landline networks. All of the carriers also have a cap on tethering data – making it even harder to use as a landline substitute – T-Mobile caps tethering at 10 GB per month.

Putting it all into Context. To put this into context, John Legere already claims today that people ought to be using T-Mobile as a landline substitute. He says people should buy a multi-cellphone plan and use one of the phones to tether to landline. 4G networks today have relatively high latency and 4G speeds today can reach 15 Mbps in ideal conditions but are usually slower. 4G also ‘bursts’ today and offers faster speeds for the first minute or two and then slows down to a crawl (you see this when you download phone apps). I think we have to take any claims made by T-Mobile with a grain of salt.

I’m pretty sure that concept of using the merger to create a new giant ISP is mostly a red herring. No doubt 5G will eventually offer an alternative to landline broadband for those homes that aren’t giant data users – but it’s also extremely unlikely that a combined T-Mobile / Sprint could somehow use 5G cellular to become the fourth biggest ISP starting ‘a few years from now’. I think this claim is being emphasized by the two companies to provide soundbites to regulators and politicians who want to support the merger.

Spectrum and 5G

All of the 5G press has been talking about how 5G is going to be bringing gigabit wireless speeds everywhere. But that is only going to be possible with millimeter wave spectrum, and even then it requires a reasonably short distance between sender and receiver as well as bonding together more than one signal using multiple MIMO antennae.

It’s a shame that we’ve let the wireless marketeers equate 5G with gigabit because that’s what the public is going to expect from every 5G deployment. As I look around the industry I see a lot of other uses for 5G that are going to produce speeds far slower than a gigabit. 5G is a standard that can be applied to any wireless spectrum and which brings some benefits over earlier standards. 5G makes it easier to bond multiple channels together for reaching one customer. It also can increase the number of connections that can be made from any given transmitter – with the biggest promise that the technology will eventually allow connections to large quantities of IOT devices.

Anybody who follows the industry knows about the 5G gigabit trials. Verizon has been loudly touting its gigabit 5G connections using the 28 GHz frequency and plans to launch the product in up to 28 markets this year. They will likely use this as a short-haul fiber replacement to allow them to more quickly add a new customer to a fiber network or to provide a redundant data path to a big data customer. AT&T has been a little less loud about their plans and is going to launch a similar gigabit product using 39 GHz spectrum in three test markets soon.

But there are also a number of announcements for using 5G with other spectrum. For example, T-Mobile has promised to launch 5G nationwide using its 600 MHz spectrum. This is a traditional cellular spectrum that is great for carrying signals for several miles and for going around and through obstacles. T-Mobile has not announced the speeds it hopes to achieve with this spectrum. But the data capacity for 600 MHz is limited and binding numerous signals together for one customer will create something faster then LTE, but not spectacularly so. It will be interesting to see what speeds they can achieve in a busy cellular environment.

Sprint is taking a different approach and is deploying 5G using the 2.5 GHz spectrum. They have been testing the use of massive MIMO antenna that contain 64 transmit and 64 receive channels. This spectrum doesn’t travel far when used for broadcast, so this technology is going to be used best with small cell deployments. The company claims to have achieved speeds as fast as 300 Mbps in trials in Seattle, but that would require binding together a lot of channels, so a commercial deployment is going to be a lot slower in a congested cellular environment.

Outside of the US there seems to be growing consensus to use 3.5 GHz – the Citizens Band radio frequency. That raises the interesting question of which frequencies will end up winning the 5G race. In every new wireless deployment the industry needs to reach an economy of scale in the manufacture of both the radio transmitters and the cellphones or other receivers. Only then can equipment prices drop to the point where a 5G capable phone will be similar in price to a 4GLTE phone. So the industry at some point soon will need to reach a consensus on the frequencies to be used.

In the past we rarely saw a consensus, but rather some manufacturer and wireless company won the race to get customers and dragged the rest of the industry along. This has practical implications for early adapters of 5G. For instance, somebody buying a 600 MHz phone from T-Mobile is only going to be able to use that data function when near to a T-Mobile tower or mini-cell. Until industry consensus is reached, phones that use a unique spectrum are not going to be able to roam on other networks like happens today with LTE.

Even phones that use the same spectrum might not be able to roam on other carriers if they are using the frequency differently. There are now 5G standards, but we know from practical experience with other wireless deployments in the past that true portability between networks often takes a few years as the industry works out bugs. This interoperability might be sped up a bit this time because it looks like Qualcomm has an early lead in the manufacture of 5G chip sets. But there are other chip manufacturers entering the game, so we’ll have to watch this race as well.

The word of warning to buyers of first generation 5G smartphones is that they are going to have issues. For now it’s likely that the MIMO antennae are going to use a lot of power and will drain cellphone batteries quickly. And the ability to reach a 5G data signal is going to be severely limited for a number of years as the cellular providers extend their 5G networks. Unless you live and work in the heart of one of the trial 5G markets it’s likely that these phones will be a bit of a novelty for a while – but will still give a user bragging rights for the ability to get a fast data connection on a cellphone.

Cellphone Data Usage

I’ve never seen any detailed information about the amount of data that customers use on cellphones. We have the global statistics from Akamai and others that look at the big picture, but I’ve always wondered how much data the average cell phone user really uses. This is something that is important to understand for ISPs because cellphone usage on home WiFi can be a big chunk of bandwidth these days.

FierceWireless has now partnered with Strategic Analytics to look in more detail at how people use their cellphone data and how they pay for it. The data used in the analysis comes from 4,000 android phone users who agreed to allow their usage to be studied.

Following is a comparison on an average month for the amount of Cellular and WiFi bandwidth used by customers with different kinds of data plans:

‘                                                               Cellular             WiFi               Total

No Data Plan (pay-as-you-go)              0.9 GB              8.8 GB            9.7 GB

Monthly Data Cap                                 2.8 GB            14.0 GB          16.8 GB

Unlimited Data Plan                             5.3 GB            12.3 GB          17.8 GB

Interestingly, there is not that much difference in the total bandwidth used by customers with unlimited data plans versus those with caps. But the unlimited customers obviously feel freer to use data on the cellular network, using twice as much cellular data per month as those with monthly caps.

What is surprising to me is the small amount of data used by unlimited plan customers. There are truly unlimited plans like T-Mobile, but even the quasi-unlimited plans from AT&T and Verizon allow for over 20 Gigabytes of download per month on cellular. But these statistics show that customers, on average, are not using much of that data capability. It looks like many people are buying the unlimited plans for the peace-of-mind of not exceeding their data caps. This reminds me a lot of the days when telcos talked people into buying unlimited long distance plans, knowing that most of them would never use the minutes.

These statistics also show that unlimited data customers are not putting a lot of pressure on cellular networks, as the carriers would have you believe. They have always used the excuse of network congestion as the excuse for charging a lot for cellular data and for having stingy data caps. These statistics show just the opposite and show that, in aggregate that customers are not using cellular data at even a tiny fraction of the bandwidth they use on their home broadband connections.

These statistics also indicate that there are not a lot of people using cellphones to watch video. T-Mobile may give access to Netflix, but it looks like people are either watching the video on WiFi or on a device other than their cellphone. It doesn’t take much video to get to 5 GB per month in download.

To put the total usage numbers in perspective, the average landline broadband connection uses around 120 GB per month according to several ISPs. I’ve seen numerous articles over the last year talking about how cellular data use is exploding, but these numbers don’t back that up. This shows that consumers still go to landline data connections when they want to do something that is data intensive.

These numbers also counterbalance the predictions I keep reading that cellular data will eclipse landline data in a few years. That might true around the world since there are a number of places where almost all ISP connections are through cellphones. But in the US the landline data usage still dwarfs cellphone data usage and is itself still growing rapidly.

The usage by cellular carrier was also reported, as follows:

‘                                                          Cellular             WiFi                Total

AT&T                                                 2.4 GB            11.4 GB          13.6 GB

Sprint                                                4.4 GB            13.8 GB          18.2 GB

T-Mobile                                           5.3 GB            13.1 GB          18.4 GB

Verizon                                             3.6 GB            14.4 GB          18.0 GB

My one take-away from these numbers is that Sprint and T-Mobile customers feel freer to use their smartphone for video and data downloading – but even they mostly do this on WiFi. These numbers also show that the stingy monthly data caps from AT&T and Verizon have trained their customers to not use their cellphones – even after those companies have increased the monthly caps.

Regulation and Capital Spending

At the recent Mobile World Congress, FCC Chairman Ajit Pai said that one of his reasons he wants to reverse Title II regulation is that it has had a drastic impact on capital spending by ISPs. He says that the new regulations have been a disincentive for the ISPs to invest in broadband.

The Chairman bases that position on statistics provided by USTelecom which are based upon work done by Hal Singer, a Senior Fellow at GW Institute for Public Policy. Mr. Singer created the following table that shows the domestic capital spending for the big ISPs for 2014 through 2016. And indeed, this table shows a 5.6% drop, or $3.6 billion a year from 2014 to 2016 – which Mr. Singer attributes to Title II regulation.

2014

2015

2016

AT&T $21.1 $17.3 $17.8
Verizon $17.2 $17.8 $17.1
Comcast $6.4 $7.1 $7.7
Sprint $3.8 $3.9 $1.4
Time Warner Cable $4.1 $4.4 $3.8
T-Mobile $4.3 $4.7 $4.7
CenturyLink $3.0 $2.9 $3.0
Charter $2.2 $1.9 $3.1
Cablevision $0.9 $0.8 $0.6
Frontier $0.6 $0.7 $1.3
US Cellular $0.6 $0.5 $0.5
Suddenlink $0.3 $0.4 $0.3
   Total $64.6 $62.4 $61.0

But like with all statistics, it’s not hard to draw different conclusions from the same set of numbers. For example, all of the drop in capital spending can be attributed to AT&T and Sprint. Taking those companies out of the table shows that capital spending for the other big ISPs is up $2.1 billion or 5% from 2014 to 2016.

So what’s going on with AT&T? There are a number of reasons for their change in capital spending:

·         During these same years the company made massive capital investments in DirecTV ($3 billion over the last few years) and also on the company’s purchase and expansion of its cellular network into Mexico ($3 billion over 4 years). Those numbers are not included in the above table and it’s easy to argue that the company just set different priorities and diverted normal domestic capital to these two giant ventures. If you add those capital expenditures into the table then AT&T’s capital spending has grown – just not their ‘domestic’ spending on traditional broadband.

·         AT&T has been making a huge effort to update its cellular network using software defined networking (SDN) as described at this AT&T website. They have been decommissioning traditional hardware at cell sites and installing much less expensive, off-the-shelf routers that can now control the cell sites from centralized data centers. They have already converted over half of their cell sites and this upgrade means vastly reduced spending on traditional cell site electronics. The company has been bragging about this shift to investors for several years.

·         AT&T has also retracted from expanding traditional big tower cell sites. For a number of years AT&T has been spending money to get fiber to its more remote cell sites, and that upgrade is largely done.

Sprint can also be easily explained. This is a company in trouble and that has been well documented over the last few years. A number of attempts to find a buyer has fallen through. What’s not shown on this table is that in 2013 (the year before the table begins) Sprint spent $6.4 billion on capital in a massive system-wide upgrade to LTE. Since then the company has very publicly stated that they are cutting capital spending to conserve cash. The company is only expanding now with carefully selected small cell deployments. But the company is clearly in network maintenance mode and is spending only what is needed to keep the cell sites functioning. Also included in the drop in spending is a change in the way that Sprint treats leased cellphones – they used to capitalize the phones and they now expense them.

There are going to be further decreases in future telecom capital spending across the industry. I expect capital spending for all four big wireless companies to keep decreasing due to efficiencies from SDN. We are now seeing a burst of spending from cable companies due to upgrades to DOCSIS 3.1, but when that’s done I would expect a significant decline in their capital spending as well. We are entering a time when improvements in software will lower the need for new hardware – not just in telecom, but in many other sectors as well.

I have always been annoyed when statistics are used to falsely justify public policy. There is no evidence that the big ISPS have changed their spending habits in any drastic way due to Title II regulations. The argument that Title II has affected capital spending comes directly from constant press releases from USTelecom, and the FCC Chairman should be above repeating arguments from lobbyists. If the FCC wants to undo Title II then it should just do it – there are a number of valid reasons why this might be good policy. But it’s disingenuous to cook up false reasons for why the change is needed.