Increasing Internet Access Availability Across South Asia: What To Do Next
by Gary Kim
Spectrum Futures
Fundamentally, there are a few ways mobile operators and other Internet service providers can add more bandwidth, and those choices are crucial in Asian Internet access markets, for the simple reason that half the world’s Internet users, smartphone users and bandwidth consumption will happen in Asia.
Simply, mobile operators can gain the use of more spectrum; create smaller cells or improve modulation and radio efficiency. But most platforms–fixed or untethered–rely fundamentally on the concept of cellularizing coverage.
Wi-Fi suppliers can deploy more cells and use both 2.4 GHz and 5 GHZ bands, or use multi-array antennas.
Fixed network suppliers also rely on use of smaller “cells.” In principle, a fiber “serving area” (fiber to neighborhood) is a form of network cellularization. Fiber to the premise essentially creates a very small cell that can use all the bandwidth of a lightguide, consistent with the chosen optoelectronics.
Some potential new transport or access platforms also use a form of “cellularization.” Drones can orbit over a particular area. Free-floating balloons or low earth orbit satellites create moving areas of coverage that are the equivalent of “cells.”
Likewise, there arguably are several ways for national regulators to stimulate the quality and quantity of communications services and bandwidth. Among those methods:
- release additional spectrum
- license new competitors
- remove barriers such as high taxes, fees
- expedite infrastructure approval processes
“Regulation is the main lever by which governments can influence competition in a given market,” the International Telecommunications Union says.
Stimulating Investment and Competition
The key regulatory tools are regulation to unleash competition and provide additional capacity.
Price regulation often is applied to wholesale services or interconnection. That tends to work well enough as a means of protecting consumer welfare (prices), but does not generally help spur additional investment (better services, more bandwidth) very much.
Specific policies such as mobile number portability and regulating wholesale prices for local loop access can spur competition, typically resulting in lower consumer retail prices. But almost nothing works as well as allowing new competitors to enter markets.
Consider Costa Rica, which liberalized its market and encouraged market entry by new mobile firms in 2011, among them Claro and Telefónica, as well as two mobile virtual network operators.
As you would guess, the entry of four new providers lead to big price discounts and a big increase in customers. The new providers gained some 69 percent of net new account additions from 2010 to 2014, for example. Mobile penetration more than doubled, while mobile Internet services penetration increased tenfold, the ITU notes. And strong increases in investment were also recorded in 2012 and 2013.
Competition in the mobile market also drove the significant increase in the percentage of households with Internet access, which grew from 24 percent in 2010 to 55 percent in 2014.
“Mobile broadband is not only bringing new people online, but also making the Internet available at home for people who had previously accessed it elsewhere,” the ITU says. Some 33 percent of households with Internet access in Costa Rica relied on mobile access in 2012, for example.
Others might note that any policies that increase provider costs will result in higher retail prices for consumers. That includes taxes, fees, import duties, universal service obligations, spectrum fees and charges or limited release of new spectrum.
Consumer Demand
Consumer demand plays a primary role, in addition to methods to increase supply.
By 2020, for example, 54 percent of South Asia mobile customers will be using smartphones, according to GSMA estimates. Those choices by consumers necessarily drive demand for more bandwidth.
Between 2015 and 2021, data consumption per smartphone will grow nine times in Western Europe, five times in Central and Eastern Europe, nearly six times in the Middle East and Africa, nearly seven times in Asia Pacific, nearly six times in North American and five times in South America.
In much of Asia, as in some other regions, the smartphone is the gateway to use of the Internet. In India, for example, about 57 percent of the time, the smartphone is the access device of choice, according to the Google Consumer Barometer.
In the Philippines, about 39 percent of the time, the smartphone is the preferred or more-used access device.
Smartphone adoption will be a key driver of bandwidth demand, since mobile Internet consumption on smartphones is growing at a compound annual growth rate of 50 percent.
It goes without saying that attractive applications and services are the reasons people want to use smartphones and the Internet.
Mobile Drives Internet Access
There will be 2.51 billion mobile phone users in the Asia-Pacific region in 2015, a figure equal to 62.5 percent of the population, rising to 69.4 percent by 2019, according to eMarketer.
Also, Asia will account for 39 percent of global data consumption by 2019, as a result, according to Cisco. And India will represent a huge part of the growth.
India is on track to surpass half a billion mobile subscribers by the end of the year, according to a new GSMA Intelligence study. By 2020, India will account for almost half of all the subscriber growth expected in the Asia Pacific region.
The Mobile Economy: India 2015 notes that 13 percent of the world’s mobile subscribers reside in India. At the end of 2014, India’s mobile subscriber penetration rate was about 36 percent of the population, compared to a 50 percent global average.
But that is going to change, fast.
The subscriber penetration rate in India is forecast to reach 54 per cent by 2020 as many millions more are connected by mobile.
India had 453 million unique mobile subscribers at the end of 2014, but is forecast to surpass 500 million by the end of 2015 and add a further 250 million subscribers by 2020 to reach 734 million.
That matters for reasons beyond the ability to communicate using voice and text. Fixed broadband penetration in India is about 2.5 percent, and will not increase much more than that, for business model reasons.
In contrast, 60 percent to 90 percent of the population have access to at least a 2G mobile service.
That means the mobile network will become the dominant means of getting access to the Internet. The number of individuals accessing the internet over mobile devices had expanded from less than 100 million subscribers in 2010 to nearly 300 million at the end of 2014.
Although India only launched 3G services in 2009 and 4G deployments are at an early stage, the move to mobile broadband networks is set to gather pace over the coming years.
Mobile broadband networks (3G/4G) accounted for only 11 percent of Indian mobile connections in 2014, but are expected to make up 42 percent of the total by 2020.
One factor driving the migration to mobile broadband networks is the increasing adoption of smartphones, made possible by low-cost devices.
More than half a billion new smartphones connections are expected in India between 2015 and 2020, bringing the total to 690 million, up from 149 million in 2014.
The penetration of mobile Internet has reached 24 percent of the population by mid-2015.
This figure will almost double again in the next five years to reach 44 percent of the population by 2020, with around 600 million mobile internet subscribers.
Increasing use of the Internet in India will be driven by smartphone adoption, higher use of 3G and 4G networks, lower devices costs, more-affordable subscription prices and greater understanding of the value of Internet apps.
More Bandwidth from WRC-15
There are a number of ways of increasing the capacity of a mobile network: use more spectrum; create smaller cells; improve modulation and radio efficiency.
The easiest way is to gain use of additional spectrum. That also tends to be the lowest-cost method for increasing bandwidth, from the mobile operator perspective.
That matter was on the agenda for the International Telecommunications Union World Radiocommunications Conference 2015 (WRC-15), which approved a globally harmonized 200 MHz portion of the C-band (3.4 GHz to 3.6GHz) for mobile communications.
That band is expected to help provide capacity in urban areas, for reasons directly related to signal propagation and bandwidth-carrying capacity of frequencies in the gigaHertz range.
The WRC-15 meeting also created a globally harmonized portion of the L-band (1427 MHz to 1518 MHz), for mobile communications, as well.
WRC-15 also expanded the 700 MHz band (694 MHz to 790 MHz) from a regionally harmonized band in the Americas and Asia Pacific to a global band to support mobile communications.
GSMA officials also applauded the use of sub-700 MHz frequencies, especially 610 MHz to 698 MHz for mobile broadband, in markets covering more than half the population of the Americas.
India is expected to use that band as well.
GSMA expects additional work at WRC-19 to identify high-frequency bands above 24 GHz for 5G mobile services.
The GSMA had proposed that WRC-15 identify four frequency ranges to mobile broadband, including:
- Sub-700 MHz UHF (470–694 MHz): In countries with a large geographic area such as India, the spectrum below 700 MHz will be important for providing extensive and affordable mobile broadband coverage in rural areas beyond 2020. This band is already allocated to mobile services in Asia Pacific.
- But the WRC-15 also decided to keep the 470 MHz to 694 MHz a primary band for TV broadcasting, with secondary rights for mobile use, until 2023, in Europe, the Middle East and Africa.
- L-band (1350−1400 MHz and 1427–1518 MHz): This band is capable of delivering additional capacity and coverage over relatively large areas, including inside buildings. It is the most supported band across all continents, with some European countries already auctioning part of the band in 2015.
- 2.7–2.9 GHz: This band, which is underutilized, particularly in South Asia, would provide important extra mobile capacity, and its proximity to 2.6 GHz will facilitate the fast deployment of extra capacity to the existing cell sites.
- C-band (3.4–3.8 GHz and 3.8−4.2 GHz): best suited for dense urban areas. Satellite interests will vigorously oppose this allocation.
Except for the 2.7-GHz to 2.9-GHz S band, WRC-15, as many likely expected, approved the changes.
Delegates from over 150 countries reached a decision to maintain the lower UHF frequency band (470-694 MHz) primarily for broadcasting in ITU Region 1 (comprised of Europe, the Middle East and Africa), as opposed to allocating it for mobile broadband on a primary or co-primary basis.
This places ITU Region 1 in alignment with decisions already made for ITU Region 2 (the Americas) and Region 3 (Asia).
Tools to Increase Bandwidth
In addition to gaining access to more spectrum, mobile operators can use smaller cells to increase effective use of any available resources. All cellular networks work by re-using a fixed amount of spectrum.
Use of smaller cells allows more-intensive use of any available amount of spectrum, balanced against the cost of many more towers and a more-difficult session hand-off problem.
In addition, mobile operators can sub-divide their macrocell serving areas, creating smaller cells.
In fact, by some estimates, smaller cells account for as much as one or two orders of magnitude more effective bandwidth than “new spectrum” or “better technology.”
In other words, some would argue, spectrum re-use has been the most-important way of increasing effective mobile network bandwidth. That is one reason many are optimistic about new ways to share spectrum, using a variety of techniques.
Mobile operators also can use a number of other technology approaches, including use of frequency hopping, power control, overlaying microcells, using half-rate codecs and adaptive antennas at the base station, according to Ericsson.
New air interfaces can help as well. Long Term Evolution fourth generation networks, for example, are more bandwidth efficient than third generation networks, and can support an order of magnitude more bandwidth, all other things being equal.
Nokia, among others, has argued that mobile networks can grow capacity in three ways: adding more new spectrum; using more efficient modulation and antenna technologies; and using microcells (small cells).
Those approaches can be summarized as efficiency, spectrum and density.
Regulatory, Market and Supplier Actions
Technological advances, prepaid payment methods and a marked fall in prices driven largely by a high degree of market liberalization and economies of scale, plus efforts by regulators to stimulate competition and release new spectrum, have helped increase supply of Internet access products for consumers in all markets, including consumers in South Asia.
Between 2013 and 2014, prices continued to fall across both developed and developing regions, in relative as well as in absolute terms, albeit at lower rates than in previous years, according to the International Telecommunications Union.
Even in the developed countries, mobile retail prices have been falling, from an average of 1.5 percent of gross national income per capita to 1.4 percent, the ITU says.
In 2014, the mobile services typically represented 5.6 percent of per-capita GNI in developing countries, a vast improvement from 11.6 percent in 2008.
In the lesser developed countries, mobile prices in 2014 cost 14 percent of GNI per person. That arguably still is too high for mass adoption, but prices are significantly reduced from the 29 percent levels of 2008.
Fixed network broadband services, though, have ceased to become cheaper or more affordable since 2014, on a global level.
That might suggest we are reaching a point where the business model is driven and limited by the cost of physical infrastructure and operating costs, while all other cost contributors are less an issue.
In fact, prices are rising in a number of developing countries. That might be driven in part by fixed network business models (less revenue from voice means all fixed costs must be recovered from other sources), and in part by changes in the products consumers are buying (faster speed services generally cost more, per month).
In 2014, fixed network broadband cost an average of 29 percent of GNI per capita, up from 25 percent a year earlier.
Globally, fixed-network broadband costs grew from 17.9 percent of GNI per person to 20.8 percent of GNI per person.
That might suggest policy needs to be constructed to remove barriers to fixed network investment.
Some steps can be taken by commercial providers. Compared with the benchmark of $10/Mbps in North America and Europe, Asia’s transit pricing is approximately seven times as expensive ($70 per Mbps, based on the benchmark).
When peering is taken into account, however, the effective price of bandwidth in the region is $32 per Mbps. Lower transit and backhaul prices would help, even if an unwelcome development from a transport provider perspective.
Regulators have key roles to play, as well. They control the release of new spectrum, and terms and conditions for use of existing spectrum.
Spectrum sharing, for the moment, is under active consideration in the United States and European Union, largely centered on the 3.5 GHz band in the United States and the 2.3 GHz band in the EU.
As always is the case, large addressable markets are required to create incentives for volume production of equipment.
For the moment, the hope is that the United States and European Union, separately, will provide enough of an opportunity to create reasonable sales volumes that also can be reaped by other nations and suppliers.
Assuming the trend does not materialize in the near term in other regions, mobile service providers will require additional spectrum to keep deploying next generation networks. The expectation is that those additional assignments will be on a licensed and exclusive basis.
But TV white spaces could provide a chance to innovate in the Internet access space, in many emerging markets.
At the same time, impediments to lower mobile retail costs–and agnostic or supportive policies for new platforms–also are worth examining.
In the Internet access ecosystem, consumers ultimately pay all value chain costs. It therefore behooves us to contain or lower all other inputs, whenever and wherever possible.
Regulators can help by removing barriers to investment, releasing spectrum and allowing innovators to experiment with new access or transport platforms. Suppliers can help by creating and sustaining business models that allow retail pricing to hit targets no higher than a couple of percent of per-person income.
The good news is that history seems to be on the side of those who want, and need, lower retail prices to connect everyone.
Ability to reduce costs is going to be essential, if one believes, as does the International Telecommunications Union, that retail consumer communications service prices are going to drop by 40 percent by 2020.
Prices for mobile services, including mobile broadband, have been falling falling for some time, according to Ofcom, for example.
