• January 13, 2015

Internet for the Next 3 Billion

Last month, I traveled to Doha, Qatar to participate in the ITU’s Telecom World conference. While there I got to understand how a satellite provider brings Internet access to South Sudan using medium-earth orbit satellites and, amazingly, achieves terrestrial latencies to a region where reliable terrestrial connections simply don't exist!  The mission of this company is to help close the digital divide by extending Internet access to the estimated three billion people on the planet who are currently not served.  Our measurements show the that performance improvement over traditional satellite can be dramatic.

ITU Telecom World

First, let me say a few words about the conference itself and then I'll review this intriguing new satellite service.  In Doha, I was on a panel entitled Affordable International Backhaul and chaired by Abu Saaed Kahn of LIRNEAsia, a telecommunications policy institute primarily focused on the Asia-Pacific region.


Panel Session:Affordable International Backhaul
On the panel, Siddhartha Raja of the World Bank and Khaled Naguib Sedrak of NxtVn described creative approaches to the common problem of liberalizing telecom markets in developing countries. While it is an established fact that a liberalized telecom market yields better service for its customers and spurs greater economic growth, getting there from a closed market is not an easy transition. Closed markets typically involve large dominant incumbent operators, who, not surprisingly, happen to be major employers in their country. Changing an entrenched system involves difficult political decisions, as it puts those jobs at risk.


As evidence of the benefit of liberalized markets, I referred to an outstanding academic paper from 2005 called "New trends in the Latin American telecommunications market: Telefonica & Telmex" by Judith Mariscal, and Eugenio Rivera.  The authors provide an excellent analysis of comparative regulatory approaches in the markets of Latin America, which have been dominated during the past two decades by Spain's Telefonica and Mexico's Telmex (a.k.a. América Móvil).

In a nutshell, Mexico (as well as Spain for that matter) followed a regulatory approach of building a "national champion" in the telecom sector. The government allowed the champion to dominate its home market, so that it ultimately had the resources to take over other telecoms in other countries.  Brazil took an alternative approach of greater competition by dividing up its former state-owned telecommunications companies. The result was that as Brazil's markets privatized in the 1990's, its telecoms were weaker and were bought up by Telmex and Telefonica, the national champions of Mexico and Spain.  But ultimately, greater competition led to lower prices and better service in Brazil.

In my LACNOG presentation in Medellin, Colombia in 2013, I observed that the Internet of Brazil was growing by "two Mexicos" per year while the Mexican market was stagnant.  Professors Rivera and Mariscal's paper goes on to show that the domestic markets of Mexico and Spain weren't well served by their dominant incumbent, when measured by price or quality of service.  While Telmex's dominant position made its owner Carlos Slim the richest man in the world, the OECD stated in 2012 that the "welfare loss attributed to the dysfunctional Mexican telecommunication sector is estimated at USD 129.2 billion (2005-2009) or 1.8% GDP per annum." Hence the recent efforts by President Peña Nieto to reform Mexico's telecom market - not an easy task to say the least.

Connecting up the Other 3 Billion

Now let's talk about connecting the rest of the world. Steve Collar, the CEO of satellite Internet provider O3b Networks was also on my panel.  The name O3b refers to their target market: the Other 3 billion people in the world currently without access to high-speed Internet service.  It is a worthy mission, but it is their approach that caught my attention.

The typical storyline of bulk satellite Internet service around the world goes like this: as terrestrial/submarine cable arrives to serve a market, the local satellite business craters.  Take for example my post from 2011 about satellite service in Lebanon (and Uganda). Because traditional satellite is more expensive and suffers from high latency, it is hard for it to compete against terrestrial fiber optic lines.  O3b's approach is to bring the satellites closer to the earth, thus reducing the travel time to space and back. They are betting that the telecommunications market has changed enough from the 1990's to avoid the fates of previous low-earth orbit satellite providers like Iridium and Teledesic.

While walking the showroom floor, I spent time talking to the delegations from various developing countries (mainly African) about the challenges they are facing as they increase and expand service in their respective markets.  I spent a great deal of time with the telecom operators of the world's youngest nation: South Sudan.  In late 2013, we tweeted the appearance of the first two non-satellite connections for South Sudan here and here, however, unfortunately each of these links did not last long.

The business director of South Sudanese provider RCS was excited about his company's recent transition to service from O3b Networks and wanted to see how his network looked in our latency data. (For insight into latencies to major markets of the world, check out Dyn Internet Intelligence). A visualization of the transition (pictured below) is nothing short of remarkable.  Note the graph along the top that shows a precipitous drop in median latency from our over 150 measurement servers located around the world as RCS moved from traditional to O3b's MEO satellite service.


In fact, as it turns out, this type of reduction in latency is not uncommon when providers (often in underserved parts of the world) go from traditional satellite to O3b's service.  Similar transit shifts are depicted below for PNCC of Palau and Timor Telecom in Timor-Lest.

17893 38077-2

A Role for Low-latency Satellite

At LACNIC 22 in October, I participated on a panel covering the state of Internet infrastructure in Latin America.  At the end of our session every panelist agreed that further development of international terrestrial connections should be the top priority for the region in order to provide greater resiliency to submarine cable cuts, to support better performance for regional traffic, and to protect regional traffic from US surveillance.

At the submarine cable conferences (Suboptic and Submarine Networks World) where I have presented during the past couple of years, there has been a similar theme. Whether it is MainOne talking about Nigeria or WIOCC talking about east Africa, cable operators will tell you that getting affordable transit to the shore is one thing, but extending that affordable service deeper inland requires dependable terrestrial connections that often don't exist.

Finally, our analysis about Iraq's reliance on Kurdish providers from this summer showed that the country has yet to fully utilize its capacity from its submarine cables (one of which is presently down).  Like the previous African examples, terrestrial connectivity in Iraq is part of the problem.  However, for Iraq, Matt Smith of Reuters was able to identify other contributing factors in his piece about the Iraqi Internet.

These are all areas where a low-latency satellite provider could fill a role, at least as a stop-gap measure until the terrestrial infrastructure becomes more reliable.  Not to mention places like Myanmar, where no infrastructure exists in much of the country.  It is fascinating to think that bulk satellite service (a communications mode written off by many) could reinvent itself and again play a critical role in extending Internet connectivity in the developing world.

Last month, O3b launched another set of satellites (video below) in low-earth orbit to further expand their constellation.  This has the potential to be a game changing development for the "other 3 billion".



Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.Captcha