India Mobile Data Traffic Will Grow 1300% Between 2014 and 2019

With the caveat that traffic is not revenue, legitimate worries about voice markets in India, based on growing use of over the top apps, will eventually be resolved as they have been everywhere else: by a shift to Internet access as the revenue driver.

In India, Internet traffic will grow 4.2 fold from 2014 to 2019, a compound annual growth of 33 percent, according to the latest Cisco Mobile Visual Networking Index.

Mobile data traffic will grow 13-fold from 2014 to 2019, a compound annual growth of 67 percent.

India’s mobile was 12 percent of total Internet traffic in 2014 and will be 38 percent of total Internet traffic in 2019.

Wi-Fi was 53 percent of total Internet traffic in 2014 and will be 47 percent of total Internet traffic in 2019.

IP video traffic will grow six-fold from 2014 to 2019, a compound annual growth of 43 percent.

Consumer Internet video traffic will be 76 percent of consumer Internet traffic in 2019, up from 50 percent in 2014.

Smartphones will account for 40 percent (651.4 million) of all networked devices in 2019, compared to 13 percent (139.8 million) in 2014, representing a 36 percent compound annual growth rate.

So revenue growth will come, as it has elsewhere, from more connected devices, more Internet access subscriptions and ever-bigger access plans.

In India, the average fixed broadband speed will grow 2.9-fold from 2014 to 2019, from 4 Mbps to 12 Mbps.

The average Internet user will generate 14.6 gigabytes of Internet traffic per month in 2019, up 187 percent from 5.1 gigabytes per month in 2014.

The average Internet household will generate 40.9 gigabytes of Internet traffic per month in 2019, up 236 percent from 12.1 gigabytes per month in 2014.

FCC Leans Against 600-MHz Auction Set Asides

The U.S. Federal Communications Commission is  leaning against a smaller company spectrum reserve in the upcoming 600-MHz spectrum auctions, according to a Reuters report.

T-Mobile US and others have been arguing for such a spectrum reserve, of up to 40  percent of the total spectrum, for smaller carriers that do not have licenses to use much lower-band spectrum (800 MHz and lower frequencies).

Some argue the spectrum set asides are not needed. Others might argue that reserving spectrum for smaller providers does not work.

The FCC faces a problem, however. The two-stage auction process will work best if high prices are expected. That will cause more broadcast license holders to sell their spectrum.

A spectrum reserve will lead to lower prices, as it bars AT&T and Verizon, the firms with the deepest pockets, from bidding on a substantial portion of the spectrum. Less competition means lower prices.

For that reason, the FCC might be stuck. It might otherwise create the set asides. But doing so imperils the auctions. Licensees must first be persuaded to sell, and they will do so only if the prices are high.

Set asides pressure prices, causing less spectrum to be made available. So logic suggests promising license holders attractive prices, so they can sell. Only then can the buyers act.

Ofcom Proposes Two Auctions for 3.4 GHz and 2.3 GHz Spectrum

Flexibility is the apparent objective of an Ofcom proposal to separately auction 2.3 GHz and 3.4 GHz spectrum for commercial use in the United Kingdom. Ofcom is asking for potential bidder comment on an option where Ofcom would award most of the newly available spectrum later this year, or early in 2016.

The remaining frequencies would be held back for award at a later date. “Our initial thinking is that it may be appropriate to withhold around 60 MHz of the 190 MHz of spectrum available,” Ofcom says.

The 3.4 GHz spectrum will be licensed in 5 MHz lots, to retain maximum flexibility for applications other than Long Term Evolution mobile service. That band is expected to support license-exempt applications such as Wi-Fi, Bluetooth, ZigBee, some medical monitoring equipment and assistive listening devices. As much as 150 MHz will be made available in the 3.4 GHz band.

The 2.3 GHz spectrum is more likely to be used for LTE mobile service, which benefits from larger spectrum blocks. So the 2.3 GHz spectrum will be awarded in 10 MHz lots. A total of 40 MHz of spectrum is available.

Spectrum Futures

The two-award approach would avoid either an auction delay or full auction of all the spectrum.

As proposed, Ofcom would issue licenses for the 2.3 and 3.4 GHz bands for an indefinite period, but with an initial term of 20 years.

There will be no coverage obligations placed on this spectrum, because the frequencies being auctioned are better suited for high capacity and faster speeds, rather than achieving wide geographical coverage, Ofcom says.

The decisions are at least partly based on consolidation in the U.K. mobile market.

BT has announced plans to buy EE, while Hutchison Whampoa has reached agreement to acquire O2 from its current owner Telefonica. If the latter merger goes ahead it would reduce the UK wholesale mobile market from four major operators to three.

When Does "Innovation" Become "Restraint of Trade?" When Does Regulation Prohibit Innovation?

One of the great tensions in communications or Internet regulation is the balance between ordinary competition and unfair monopolistic practices that harm competition.

And that is where the network neutrality debate lies. Some might argue that “ordinary” practices in a competitive market are confused with unfair monopolistic practices.

The effort to ban the business practice known as zero rating provides an example. In any competitive market, the better products gain more share. That is “unfair” only in the sense that consumers will buy what they perceive to be the better products.

So bigger firms, with more market share, inevitably arise. That is evidence that consumers are choosing some products over others. It is “unfair” to less-valued products.

That is not the case where dominant firms restrict competition. Debates over network neutrality principles now are shifting into areas where there is less clarity and less agreement.

The difference is between permissible innovation and retail packaging, including the use of promotions, loss leaders, discounts or special offers by any firm in a market, and “unfair” practices by dominant providers that restrict competition.

Much of the debate implicitly contains assumptions about what is happening.

There seems to be a growing sense on the part of India regulatory officials that current application “zero rating” plans of two types violate the “spirit” of network neutrality principles, though some might suggest a final ruling is unlikely in the near term.

And the issues has become a major political issue, apparently. In fact, some might legitimately argue that India has become the current center of struggle over net neutrality policy.

Ironically, one unintended consequence of strong net neutrality rules that prohibit zero rating is that Internet adoption could be slowed and innovation thwarted.

The Progressive Policy Institute argues that bans on quality of service mechanisms, including zero rating, undermine innovation for real-time applications like telemedicine and high definition voice.

Many would argue that offering subsidized access to key apps is a proven and effective way to acquaint non-users with the value of Internet access and apps. That has proven to be the case in the Philippines and elsewhere where Internet.org has tested the concept.

Potential low-income users also suffer when subsidies such as zero rating are outlawed, the Progressive Policy Institute argues.

Still, the India Department of Telecommunications is said to have a "strong view" that 'Airtel Zero' and 'Internet.org' plans--which operate differently--do not follow net neutrality principles.

It is a horrendously complicated matter, one might well argue, and might not be resolved soon. The emerging “strong view” also is said to be “preliminary.”

“Any final decision regarding the two services will be taken only once the expert committee of the ministry submits its recommendations on the entire gamut of issues related to net neutrality,” one source says.

Further key input on network neutrality rules also are necessary and likely to take some time, as there is a change of Telecom Regulatory Authority of India leadership underway.

TRAI Chairman Rahul Khullar is leaving office and a new chair will likely need some time to develop a “detailed understanding” of an issue that is highly contentious and complicated.

So a delay of TRAI recommendations is expected.

The government is said to be investigating how other countries, including the United States and European region are dealing with the issue.

"It is a complex issue and all efforts are being undertaken to ensure that any best practices, if they are in place somewhere, are also considered by us," ETTelecom.com reports an official’s comments on the subject.

Airtel Zero might be considered the more controversial of the programs, though it is a well established concept in retailing.

Under the Airtel Zero program, application providers can subsidize data charges when Airtel customers use their apps.

The principle is similar to the notion of “toll free calling,” where a business pays for long distance calls made by its prospects, customers and partners. The “800” number is a good example.

Internet.org is different. Under that program, any app provider, without paying any money to anybody--not the mobile service provider and not Internet.org--is offered free of charge to users, who do not have to buy a data plan.

The intent is to allow consumers who do not use the Internet to sample key apps and learn the value of the Internet. Any app provider can apply to be part of Internet.org.

But some object to Internet.org principles that apps need to be optimized for low-bandwidth connections.

Many who have experienced similar low-bandwidth situations can understand the performance challenges and the value of optimizing apps for better performance under challenging conditions.

Indeed, “better performance” is a recurring concern. It is why Google Fiber was launched, for example.

So one might hope the examination of network neutrality both thoughtful and detailed. The issue is horrendously complex, has many subtleties and might have direct bearing on how fast demand for Internet access is stimulated, in India and elsewhere.

LeoSat Picks Up Where the Others Leave Off

LeoSat, which plans to launch a new constellation of 80 or more low earth orbit satellites to provide high-throughput Internet access covering every square inch of the earth, thinks its wholesale business model and high bandwidth makes it a potential partner for virtually every other satellite capacity supplier or retailer, aside from the core markets it has identified.

For starters, LeoSat is focusing exclusively on wholesale capacity for business customers, not the consumer business and not business segment retail.

“We wouldn’t compete with anybody in the current milieu,” says Fotheringham. “Our lowest service tier begins where traditional satellite ends.”

The lowest tier of service offers 50 Mbps to 100 Mbps of Internet connectivity. The middle range offers 100 Mbps to 500 Mbps while the top tier supports 500 Mbps up to 1.2 Gbps.

“We do what they cannot,” Fotheringham says of the comparison with legacy satellite services. So he believes LeoSat will have “many chances to align with incumbents who are delivery partners.”

Strictly focused on business-to-business customers, LeoSat’s primary focus will be delivering “ industrial-grade communications to major organizations,” both commercial and government, says Fotheringham.

In fact, even early on, when its first satellites are launched, LeoSat sees “some very interesting opportunities for early-entry strategies, where with a single plane you can get 24/7 coverage in the high and low latitudes, where we find big gaps in infrastructure with not a whole lot of people but very high-value people with high-value requirements.”

One shorthand description might be that LeoSat sees its target market as the biggest 3,000 companies in the world. The segments of those customer requirements that make the most sense are mission-critical and high bandwidth Internet communications in locations where such bandwidth is hard to secure.

That logically could include exploration, scientific or extraction operations. LeoSat would offer lots of bandwidth to users from big firms that have high requirements, in places where capacity is sorely lacking, and high ability to pay for quality connections.

Spectrum Efficiency, Not Just "Spectrum," is Required

Spectrum efficiency is one reason spectrum sharing seems to make so much sense. And allocating spectrum for use does not always mean spectrum is deployed commercially.

“It could be argued that close to 19 percent of the RF spectrum allocated to mobile service in Chile is being underutilized,” says Jose Felipe Otero Muñoz is the director of Latin America and the Caribbean for 4G Americas.

In Ecuador as much as 38 percent of allocated spectrum is less than fully used, while in Mexico 30 megahertz of spectrum in the 1900 MHz band was allocated at the end of the last century, but never has been used for commercial mobile service, Muñoz says.

Some might say spectrum is being “warehoused.” Others might simply say there is a mismatch of supply and demand, caused, in part, by less than fully competitive market dynamics.

The other issue is that additional spectrum that might be used to stimulate competition has not been brought to market.

The International Telecommunications Union calls for making available 1,300 Mhz at frequencies lower than 3 GHz by 2015.

While in developed countries the amount of spectrum allocated to mobile services is more than 50 percent of that figure, by the first quarter of 2015, only one Latin American market has allocated more than 40 percent of the spectrum recommended by the ITU, according to Signals Telecom Consulting,

Mobile Operators in India Can Share Spectrum, TRAI Says

Spectrum sharing can take a number of forms. In India, it primarily is an immediate issue with respect to mobile operators pooling spectrum in the 2G and 3G bands.

In essence, the ability to share spectrum is similar to earlier forms of asset sharing where mobile operators agreed to share the cost of mobile tower sites.

In other words, two or more operators with spectrum in the 800 MHz band, for example, could agree to pool their respective spectrum holdings to create larger or more-contiguous blocks of spectrum.

The Telecom Regulatory Authority of India (TRAI) recommended allowing mobile operators to share spectrum, including capacity at 2100 MHz, for the purpose of  improving spectral efficiency.

Sharing is allowed in 800 MHz, 900 MHz, 1,800 MHz, 2,300 MHz and 2,500 MHz bands, so long as all the carriers wishing to share have licenses in the shared bands.

Google Wants WRC Support for Internet Access Using Unmanned Aerial Vehicles

Google has asked the Federal Communications Commission to push the World Radiocommunication Conference to expand spectrum devoted to High Altitude Platform Stations (“HAPS”) as a method of supplying Internet access, across a wide range of geographies.  

Google has been testing the use of unmanned aerial vehicles for delivering Internet access services at altitudes between 20 kilometers to 50 kilometers above the surface of the earth.  

Scale, Google argues, is essential if the new method is to become a sustainable way of providing Internet access. Right now, spectrum is too limited, as well as lawfulness across most countries, to support a robust business case, Google argues.

Connected Cars Could Add 97% More Data Traffic at Some Cells in 10 Years

If nothing else is done (and obviously, something will be done), connected cars at rush hour could double data traffic double in certain cells, researchers at Machina Research say.

The study, commissioned by analytics company TEOCO, predicts a 97 percent increase in data traffic over the next decade, caused primarily by connected cars.

Some of you would not be surprised by a prediction of 100 percent increase in traffic, on any network, at any site, over a decade.

What the study intends to highlight is the specific new demand created solely by the connected cars.

By 2024, Machina Research predicts “machine-to-machine” mobile network connections increase from 250 million in 2014 to more than 2.3 billion worldwide.

Obviously, all those new devices, an order of magnitude more, would produce, all other things being equal, an order of magnitude additional demand. But the usage profile might be notable for its difference from human-used smartphones, tablets and PCs.

M2M applications and services will account for just four percent of overall network traffic in 2024, Machina Research predicts.

So bandwidth consumption is not the problem so much as network resource management, which always is more complicated than comparable fixed network planning, since the nodes are stationary.

The study suggests a number of ways the management  problems can be addressed, including the integration of capacity supplied by unlicensed networks and offload.

When Spectrum Prices Get Too High, Bidders Respond

High spectrum prices might encourage Verizon Wireless and possibly other mobile service providers to use the other tools at their disposal to increase effective bandwidth.

The hope of getting such high prices also might encourage TV broadcasters to withhold their spectrum from relicensing unless high prices are offered.

At the same time, possible Federal Communications Commission spectrum set-aside policies could contribute to lower prices paid by smaller firms, and higher pirces potentially paid by Verizon and AT&T.

But high spectrum prices represent danger, and will encourage AT&T and Verizon to use the other tools at their disposal to increase capacity.

As always, high prices encourage users to become more creative about using alternative and substitute products.

Verizon isn't as hungry for spectrum as it once was, in part because prices have gotten too high, Verizon CFO Fran Shammo suggested at a J.P. Morgan investor conference.

Shammo said that given the premium at which spectrum went in the most recent AWS-3 auction increases the value of alternative approaches, such as adding more small cells.

"I think that the AWS-3 auction has given the FCC a little bit of a challenge around unintended consequences," Shammo said.

Basically, the high AWS-3 prices likely encourage TV broadcasters to hold out selling their licenses unless they get similar, or even higher prices. That means less spectrum will be made available.

At the same time, with the Federal Communications Commission also considering bidding rules that would further reduce the amount of spectrum available to major carriers, the gains Verizon could make are further limited, and likely also work to push prices higher.

"When you bid up spectrum so high and now you have broadcasters sitting on the sidelines thinking their spectrum is worth a certain amount of money, and then you have where the FCC is trying to draft the rules around favoring some carriers over others, I think you have a problem,"  Shammo said.

Unlike the recent spectrum auctions in India, where service providers virtually had to acquire new licenses to support their present operations, and therefore had to spend whatever it took to do so, Verizon and some other U.S. mobile service providers will not “need” to rely on spectrum purchases to expand bandwidth.

As always, assuming a reasonable amount of physical spectrum is available, service providers can use network architecture to effectively reuse existing spectrum, and take advantage of offloading to Wi-Fi more extensively.

Curation Does Not Violate Net Neutrality

Many would agree, in principle, that walled content gardens are "bad." Such "curated" experiences are not the "open Internet." The issue is that every streaming video supplier uses a curated approach. No exceptions.

The issue is that all media experiences are curated; distinct; original. Curation itself is part of the issue for zero rating. "No cost" access to curated apps strikes many thoughtful observers as a violation of Internet openness principles.

But openness has been narrowing, steadily, as the Internet has become important. Government app blocking is an extreme example. But language communities, interest communities, social networks and other "natural" human expressions of choice create a functionally curated Internet experience, even when the Internet itself remains "open."

There are important and fundamental principles at stake, to be sure. But the Internet has been growing all sorts of non-traditional walls, ranging from government content restrictions to curation itself.

The debate over zero rating is complicated because it spans domains; tactics for increasing Internet access and use, and also highly valued principles.

Some of us would argue that Internet.org not only works, but leads precisely to faster use of the open Internet. There is no fundamental danger to Internet openness, over time, even if the near term sampling program is curated.

On one hand, some think Internet.org is a proven way to dramatically increase sampling of the Internet by people who have not used it before, as well as a bundle of value that has proven to create significant sustainable demand for Internet access services.

Some matters are incontestable: if one’s objective is to quickly encourage people to try the Internet and its apps, Internet.org has proven it works.

It also is true that Internet.org raises some thorny issues about “permissionless innovation” and the notion of curated apps.

No doubt, Internet.net curates apps. It perhaps is not, as the Electronic Frontier Foundation argues,

And “neutrality” actually is not the central issue. Curation is the issue.

As EFF notes, the  “guidelines are neutral as to the subject matter of the site.” In other words, no category of apps, or type of content, are excluded.

On the other hand, Internet.org knows full well there are severe bandwidth limitations in its target markets. To improve end user experience, Internet.org deliberately wants apps curated so that they work reasonably well even in a very challenging bandwidth environment.

In other words, Internet.org wants the apps to work even on inexpensive feature phones, on slow data networks. So yes, apps have to “ask permission” in terms of being architected to work on bandwidth-limited networks.

Specifically, that means stripping out images greater than 1Mb in size, videos, VoIP calls, Flash and Java applets and JavaScript.

EFF would prefer a uniformly rate-limited or data-capped free service. “We have confidence that it would be possible to provide a limited free Internet access service that is secure, and that doesn't rely on Facebook and its partners to maintain a central list of approved sites,” EFF argues.

The matter is complicated. Keep in mind that nobody gets paid anything as part of the Internet.org program. All apps and app providers can participate. The program works.

Against that, EFF suggests a program that simply offers free access.

Ignoring the philosophical objections to curation, it is not so clear that simple “no charge” access is sustainable. That is not to disparage the effort; simply to acknowledge that large numbers of mobile service providers have not yet done so.

We do know, on the other hand, that time-limited programs such as Internet.org has organized, with “free access” ending at some point, have worked to significantly expand the base of regular Internet users.

Globe Telecom in the Philippines found the number of data users on Globe’s network doubled, after offering Facebook on a zero rated basis for about a year.

The portion of Globe’s prepaid subscriber base who were active on mobile data expanded from 14 percent in September 2013 to 25 percent in November 2014, Facebook and Globe say.

In other words, the mobile Internet customer base nearly doubled.

Globe’s Free Facebook campaign (and similar internet outreach efforts by other players in the market), led to a six million increase in the number of active mobile internet users in the Philippines as a whole.

Both EFF and Internet.org support rapidly getting people to use the Internet. Globe Telecom found “free Facebook” did so.

There are philosophical issues and principles at stake, to be sure. But Globe Telecom’s experience might suggest the disagreement is more about tactics. Rather quickly, people migrate off the promotions and to the open Internet.

Almost nobody (except for some government authorities) would agree that curated Internet access is he desired end result. But, sometimes, a bit of near term curation can lead to unfettered use of the full Internet. 

It's an emotional issue. But curated access is a temporary sampling mechanism. It helps get lots of people exposed to a new product they haven't realized they needed. 

Is an Era of Communications Spectrum Abundance Coming?

With some important qualifications, spectrum abundance is coming. Consider only millimeter wave frequences.

It sometimes is hard to comprehend just how much useful communications spectrum exists in the millimeter wave bands (3 GHz to 300 GHz), even if some blocks already are allocated and even if  high attenuation has key network architecture implications.

“You can take all of the useful spectrum we now use from DC to 30 GHz and drop it into the lower end of the millimeter-wave region and still have 240 GHz left over,” said Lou Frenzel of Electronic Design.

“The unlicensed band at 60 GHz contains more spectrum than has been used by every satellite, cellular, WiFi, AM Radio, FM Radio, and television station in the world,” some would say.

In addition, there are new plans to allow spectrum sharing in the 3.5 GHz and TV white spaces bands in the U.S. market, with investigations proceeding of whether sharing is possible in hundreds of megaHertz of additional frequencies, in addition to TV white spaces and 3.5 GHz.

Spectrum at 28 GHz, 38 GHz, and 70-80 GHz looks especially promising for next-generation cellular systems, many would argue.

The bandwidth available in the 70 GHz to 80 GHz frequency band, a total of 10 GHz, is more than the sum total of all other licensed spectrum available for wireless communication in the United States, for example.

Some four bands in the upper millimeter wave region have been opened for commercial applications.

Of the four bands, the 59-64 GHz band (commonly referred to as V-band or the 60GHz band) is governed by FCC Part 15 for unlicensed operations. The band us suited to very short range point-to-point and point-to-multipoint applications.

The 92-95 GHz band (commonly referred to as W-band or the 94 GHz band) is also governed by the FCC Part 15 regulations for unlicensed operation, though for indoor applications only.

The 94 GHz band may also be used for licensed outdoor applications for point-to-point wireless communication per FCC Part 101 regulations.

However the band is less spectrally efficient than the other three bands due to an excluded band at 94 to 94.1 GHz.

The 71 GHz to 76 GHz and 81 GHz to 86GHz regions (commonly referred to as E-band or the 70 GHz and 80 GHz bands), are reserved by the U.S. Federal Communications Commission for licensed operation, as the most ideally suited millimeter wave band for point-to-point wireless communication applications, some would traditionally say.

With new spectrum sharing techniques developing, it is not so clear how access might eventually evolve across the millimeter wave region.

What is clear is that a prodigious amount of new communications spectrum is available, albeit with signal attenuation issues.

source: University of Notre Dame

Spectrum Sharing Might Free Up Nearly 1,000 MHz of U.S. Spectrum for New Commercial Users

Spectrum sharing will underpin the use of huge new blocks of U.S. spectrum over the next several years.

The U.S. Federal Communications Commission already is making 150 megahertz of spectrum available for shared small cell use in the 3.5 GHz band, and 100 MHz of that bandwidth necessarily will be made available only because spectrum sharing is possible, according to Lawrence E. Strickling, National Telecommunications and Information Administration assistant secretary.

“Next in the pipeline, we are evaluating the feasibility of increased sharing by unlicensed devices in up to 195 megahertz of the 5 GHz band,” Strickling said. “We are also working with federal agencies to assess their spectrum use in another five bands accounting for 960 megahertz of spectrum.”

“We will then be in a position to prioritize some of those bands for detailed sharing feasibility studies,” Strickling said.

Collectively, that means hundreds of megahertz of spectrum will be made available to new commercial users, without the requirement to clear existing licensed users.

The end result will be more efficient use of communications spectrum, at far less cost and time than would traditionally be necessary to free up the spectrum.

70% to 80% Smartphone Use in Thailand by 2020

About 70 percent to 80 percent of all Thai mobile phone users will use smartphones by 2019, in line with other emerging Asia–Pacific countries, according to Analysys Mason.  

Thailand could increase mobile and fixed broadband adoption to 133 percent by 2020, up from 52 per cent in 2013, the GSMA also estimates.

The GSMA study, prepared by Analysys Mason, suggests that goal will be aided if Thailand can increase mobile spectrum in the 900 MHz and 1800 MHz bands.

Thailand also should make available the globally-harmonized 700MHz “Digital Dividend” band.

Other helpful policies would establish a business-friendly environment and a level playing field with state-owned companies through transparent processes.

In part, that requires independence of the National Broadcasting and Telecommunications Commission.