05 July 2017

There are plenty of technological solutions to cost effectively connect the unconnected, including Google’s ambitious Project Loon that has attracted much interest in Asia. 

Five billion people across the world are now mobile subscribers.

That milestone was reached earlier this year and, according to the GSM Association (GSMA), a further 620 million users will be added by 2020 to reach almost three quarters of the global population.

Asia will drive the growth and account for 60 per cent of new subscribers globally. 

Sub-Saharan Africa will represent 16 per cent (99m additional users) while MENA is forecast to account for seven per cent (41m additional users).

But in its Global Mobile Trends 2017 report published in September, the GSMA warned that the rate of growth is slowing.

It said that while it took four years to move from four billion global subscribers to five billion, reaching the next billion will take longer and will be the “toughest challenge” yet. 

And with 50 per cent of the world’s population still not online, the report said: “The digital divide is greatest in India and sub-Saharan Africa which account for 42 per cent of the world’s unconnected, with more than 60 per cent of their respective populations not yet on the internet.”

At current rates of progression, Ericsson predicts that mobile broadband will provide network coverage to around 95 per cent of the world’s population by 2022.

So will that be sufficient for wireless service providers? With dwindling ARPUs continuing to afflict mobile operators everywhere, what incentives do they have to invest in rolling out their networks to remote and rural areas? 

Sort your costs out

Canada-based NuRAN Wireless has developed cost-effective mobile network infrastructure to enable rural connectivity in emerging markets.

Bradley Shaw, the company’s MEA regional manager, believes that there are profits to be made in such low ARPU environments, as long as operators make the right equipment choices.

“You just have to be efficient with the capex and the opex. Operators had no interest in expanding service into remote rural areas with traditional infrastructure because it meant operating at a loss.

“We now see operators rolling out sites based on NuRAN, and soon-to-come OpenCellular, for less than USD30,000, and paying back the investment in less than 18 months. Some low-traffic sites even get built for USD10,000, all inclusive. The efficiency in terms of spend is critical when you’re working in low ARPU environments.”

OpenCellular is one of the initiatives being developed by the Telecom Infra Project (TIP).

Established in 2016, TIP describes itself as an “engineering-focused” collaboration between operators, suppliers, developers, integrators and startups.

Their aim is to come up with fresh technologies, examine new business models, and drive investments into telecoms.

Its OpenCellular project group focuses on the development of wireless access platforms and is co-lead by experts from NuRAN Wireless, Facebook, Keysight, amongst others. 

Earlier this year in June, NuRAN presented details of its new OpenCellular product, the OC-2G to TIP members.

The company said that the base station will be integrated with its proprietary software stack and base station controller in order to form a complete RAN solution for carriers looking to expand their footprint to communities of 400 to 1,500 inhabitants.

But, as has been well documented in the industry, it’s not just a question of deploying mobile infrastructure in remote and rural areas.

For example, Shaw says: “Voice traffic, as we all know, is declining and data services are increasing. The use of data and the increase of ARPU through data services is largely device-driven, but in rural areas you have very low penetration of smartphones.”

Of course that is likely to change moving forwards with the GSMA pointing out that, like subscriber growth, smartphone uptake is also being driven by developing markets.

In its Global Mobile Trends 2017 report, the association said that Nigeria is one of five markets forecasted to account for more than 40 per cent of the 1.6 billion new smartphone connections by 2020 (the others are India, China, Indonesia and Pakistan). 

The power to succeed

However, the challenges of remote and rural connectivity cannot simply be solved by building low capex and low-cost networks and making affordable handsets available. From Africa to Asia, another basic problem in many emerging markets is a lack of grid power.

“We are seeing sites that are being closed down because they are not profitable,” says Shaw. “Why is that the case? The operator is running a diesel generator which is, say, five hours from the closest urban environment.

“So the cost of purchasing the diesel on top of the cost of shipping it makes that site unsustainable.

“Whereas if the operator had put in a solar, low-powered base station, the returns from that site might be marginal but at least it would still be breaking even.”

Specifically designed to reach the next billion subscribers, it is said to only consumer 65W, thereby minimising the capex associated with solar panels and batteries, or opex in the case of diesel-powered sites. 

The firm adds that the hand-carried, tower-mounted LiteCell does not require any machinery to install, nor any kind of shelter to protect it.

Antennas connect directly to the unit, while an all-IP interface makes it easy to connect to any IP-based terrestrial or satellite backhaul. 

Earlier this year in May, wholesale operator Raeanna Group announced that it would use NuRAN’s system for more than 1,000 sites in Nigeria over the next five years.

This followed a separate deal with Global Communications Extension Services which will also use the vendor’s platform as part of an initial deployment for an unnamed Tier 1 MNO in Nigeria.

Of course, NuRAN is not the only company to make specialised infrastructure for remote and rural mobile sites.

Since 2004, India’s Vihaan Networks Limited (VNL) has  been developing and offering low-powered base stations that can be run using solar energy as part of its WorldGSM system.

The company, which is part of the Shyam Group, says its systems has since been deployed to rural areas in Kenya, Uganda and Ghana, as well as many other Asian countries.

Another innovative infrastructure specialist that made its debut in the telecoms market a few years ago is Range Networks, the US company that claims it developed the industry’s first commercial open source cellular system.

Range says its software runs on off-the-shelf hardware that is typically less than 20 per cent of the cost of custom hardware to deliver full-featured mobile services. It reckons this allows the operator to make a profit while charging a price that “almost any” subscriber can afford.

Range has designed its system to support different radio interface protocols.

It says the system can run as virtual machines on the same standard Linux-based server hardware while sharing the same ‘IP core’.

The same software is utilised for microcell to macrocell coverage, with the operator or systems integrator mixing-and-matching COTS hardware for the most appropriate, complete coverage solution. The solution provider can either virtualise network functions or implement a self-contained Linux OS base station.

“This means that a greenfield carrier can start with a simple 2G network and, over time, develop a mixed 2G-3G-4G system, using whatever technology is best adapted to particular sites,” states the firm.

“Core network upgrades are just capacity upgrades, replacing existing servers with more cores or faster processors as the traffic volume increases, or by adding incremental software upgrades to provide new features, such as MMS, as they become available.”

In 2013, working in collaboration with the University of California Santa Barbara, Range installed a two-tower cellular extension in rural Zambia to expand coverage of an existing cellular network.

The village is home to more than 130,000 people but is spread out over an extremely large area. Many large sections of it are without cellular service.

A water tower and school building were used to mount two new systems running Range Networks’ software. Working in tandem they added approximately 35km² of new coverage for the residents. It’s claimed that it took just two days to establish this new infrastructure.

One project that has been garnering headlines over the last few years is Project Loon.

The initiative is being developed by Google as a way of putting broadband within reach of millions of currently unconnected people.

It involves 12-metre tall balloons that act like floating mobile towers.

They fly on stratospheric winds at altitudes twice as high as commercial planes, and are fitted with low-powered electronics to beam an internet connection down to the ground. As one Loon drifts out of range, another moves in to take its place.

The project has undergone trials in many countries, notably in South Asia.

For instance in Indonesia, it has been tested by the country’s three biggest cellcos (see News, p5, South Asian Wireless Communications, Q4 2015 issue), while in Sri Lanka the government has bought a 25 per cent stake in a joint-venture setup with Google in return for the spectrum that will be allocated for the project.

But in February 2017, it was widely reported that the ITU blocked Google from using the same frequency as Sri Lanka’s public broadcasters over fears of interference.

Faster than fibre – the new space race

When it comes to connecting remote and rural users, satellite technology comes into its own in terms of its speed of deployment and ubiquitous coverage. 

But at the same time, critics often point out the high price of satellite capacity which, it would seem, is at odds with the idea of MNOs drawing a profit by investing in building networks to low ARPU outposts. 

The satellite industry is countering by talking about decreasing satellite prices, particularly in terms of the cost per megabit rather than cost per megahertz.

And if the analysts are to be believed (see Wireless Business, SAWC, pp12-13, Jul-Aug 2017 issue), the market looks set for further price falls as the latest generation of smart and efficient high throughput satellites from the likes of ABS, Intelsat, Yahsat, et al, find their way into space.

But the real game-changer is likely to come with the launch of the low-Earth orbit satellites (LEO) that have been much talked about over the last few years.

Even the GSMA in its Global Mobile Trends 2017 report believes that satellite “has re-emerged from the ashes of failed attempts in the early 2000s” as an alternative connectivity option.

It said the technology could provide an alternative backhaul option in reaching rural unconnected areas in emerging markets and serve as a complement to mobile networks, offering capacity wholesale to operators. 

One of the companies that has attracted some big name backers for its LEO mission is OneWeb. With directors from major players such as Airbus, Bharti, Coca-Cola, Intelsat, Virgin and others on its board, the company’s aim is to fully bridge the digital divide by 2027. 

OneWeb says its small satellites will feature fewer components and weigh less than 150kg, thus making them easier to produce at scale and cheaper to launch. Once in space, they will create a ‘mesh’ style network by intelligently interlocking with each other to create a planet-wide footprint.

Working with manufacturer Airbus and its launch partner Virgin Galactic, OneWeb plans to send its first 10 satellites into space early next year.

Assuming these successfully pass all in-orbit tests, the full launch campaign will begin six months later with services going live in 2019. 

OneWeb will ultimately use a constellation of 648 satellites orbiting the Earth at an altitude of around 1,100km.

The company reckons this closer position will result in much better web performance, and is targeting latency of around 30 milliseconds – that’s much lower than the 240ms delay geostationary satellites suffer from as they circle the planet at an altitude of approximately 35,786km above the equator. 

US-based LeoSat is aiming to go even better with its constellation of around 78 to 108 high-power Ka-band satellites that are planned for launch in 2019.

They will use polar orbits to provide full global coverage, and each one will be interconnected using unique laser links.

Once uplinked to the constellation, LeoSat says data will travel from satellite to satellite until it reaches its downlink destination – there is no need to interconnect with any third-party network or any satellite gateway infrastructure to carry data.

According to the company, all this effectively creates an optical backbone in space which is about 1.5 times faster than terrestrial fibre backbones.

It promises an average latency of below 120ms, which would make it better than terrestrial fibre.LeoSat’s website explains that it is all down to physics:

“Light travels faster in free space than it does in a fibre optic cable once that cable reaches a certain length. Our services will start making up the extra distance [light] has to travel back and forth to the spacecraft (at 1,400km), and then get ahead of fibre. That critical cable length is about 5,000-5,500km, subject to the type and age of cable, the amount of switching panels on the route, the latitude of the begin and endpoints of the connection, to name a few variables.”

2G, 3G, 4G or ‘white elephant’?

When building networks in greenfield sites today, MNOs may face a dilemma: should they invest in basic but higher margin 2G networks, or enable first-time users in remote and rural areas to ‘leapfrog’ technologies and benefit from faster but pricier next-generation infrastructure?

“You have got to look at the device penetration in these areas,” advises Shaw. “The way that operators can do that with 100 per cent certainty is to put up a 2G network, cover everywhere, and then see what devices are registering on their network.

“Where they find there are pockets of high-penetration 3G devices, build 3G networks. And if by some chance they find pockets of very high 4G-enabled devices, they should put up a full LTE base station.”

Ericsson agrees here. In its Mobility Report published in June, the vendor offers detailed advice about how operators should go about choosing the right generation mobile technology.

For instance, in areas already covered by 2G, it says factors such as demand for connectivity, availability of device types, cost sensitivity among mobile subscribers and operator business case will influence whether upgrading to 3G or 4G coverage will be preferred as an initial solution.

One of the ways operators can decide which sites to upgrade from 2G to 3G and/or 4G is by using CDRs associated with the existing network. Ericsson says the data here can determine which 2G sites have the highest number of expected mobile broadband-capable users. 

Another useful exercise for MNOs is to see how their spectrum assets match the capabilities of their subscribers’ device capabilities. “Existing spectrum assets, spectrum re-farming opportunities and device penetration (supported technology and bands) influence the revenue potential of 3G and 4G deployments,” states the report.

Example of solutions for providing mobile broadband coverage in remote rural areas. Source: Ericsson Mobility Report, Jun 2017

But echoing the point about deploying mobile infrastructure in remote and rural areas and hoping the divide will be bridged, the GSMA is keen to point out that operators should avoid a ‘if you build it, they will come’ type mentality.

Despite the fact that most advanced countries now have national 4G networks, its says take-up patterns are mixed.

It even describes India as an anomaly: “Coverage is out of sync with consumer demand. With operators only able to reduce pricing so much in an already competitive market, the risk is that 4G becomes a ‘white elephant’.”

So what about delivering mobile broadband to areas where there is no coverage, 2G or otherwise?

Here, Ericsson says that any villages that are within 2G coverage zones, can be upgraded with 3G or 4G. Villages outside these zones can then install an outdoor high-gain antenna that can be used to provide fixed wireless broadband access to important hotspot sites within the community. 

“This solution requires low investment and the 4G site can serve a hotspot that is located 20-80km outside the 2G coverage range,” says the report. “In this scenario, the school or hospital is equipped with a roof-top antenna which, as an example, would get 3Mbps downlink speed (wireless indoor coverage and LTE modem connected to, for example, a Wi-Fi router) at a distance of 100km away from the 4G-upgraded base station site using 2x10MHz of spectrum.”

Making the connection

Clearly, connecting the next billion people requires a monumental effort, and the responsibility does not lie solely with the operator. An entire ecosystem of vendors, developers, regulators, governments, etc., has to be mobilised in order to make it happen.

The GSMA has developed a Mobile Connectivity Index that measures and quantifies the barriers to mobile internet access across four key enablers: infrastructure; affordability; consumer readiness; and content.

The index is built up through 39 specific indicators, such as mobile tariffs, handset prices, spectrum, local incomes, etc., to ultimately give each country an aggregated score from 0-100 for each of the four enablers. 

According to the index for 2016, Australia topped the rankings of 150 member countries with an overall score of 87.3. It was followed by Norway (85.5), New Zealand (85.2), Finland (83.9) and Singapore (83.4). 

The first African country to appear is Mauritius which ranks 75th with an overall score of 62.71. South Africa comes in next in 84th place with 59.97, followed by Tunisia at 89 with a score of 57.42 (see GSMA Mobile Connectivity Index table, below).

Extrapolated data for African countries from the GSMA’s Mobile Connectivity Index. Source: GSMA Intelligence

African countries go on to dominate the lower end of the table with Niger is at the bottom with an overall score of 17.22. 

Thus, mobile coverage is not the only barrier according to the association’s Global Mobile Trends 2017 report.

It states: “The largely rural populations and lack of fixed line infrastructure make extending coverage a longstanding challenge for many developing countries.

“Of the 3.7 billion not yet on the internet, around a third (1.2 billion) live outside a 3G or 4G signal and so could be considered excluded because they don’t have fast enough coverage. 

“The corollary is equally important: for two thirds of the unconnected, coverage is not the problem. Affordability, content relevance, literacy skills and gender factors are all part of the discussion.”