Solar power may soon bring the Web to remote areas

"One laptop per child" was the ambitious promise a group of MIT faculty members made to the world in January last year. The idea to develop and produce millions of simple, Internet-capable $100 laptops and give them to the poorest kids on the planet sounded appealing. But just how children in remote villages in sub-Saharan Africa were supposed to hook up to the Internet was unclear.

The solution may come out of thin air. Literally. All you need is a rooftop and the sun, claim the inventors of a solar-powered wireless device. "Green" initiatives like theirs, together with a range of wireless technologies, may extend the Internet's reach to developing communities around the globe.

An increasing number of politicians, entrepreneurs, and organizations (including the United Nations) hail the Web as a tool with the potential to help transform developing societies. Access to it provides inexpensive communication as well as vast amounts of knowledge and real-time information. Eventually, the Web will aid education efforts in developing communities and help lift them out of poverty – or so the thinking goes. Government institutions in developing nations already present themselves on the Web, and their middle-class city-dwellers increasingly have access. Even in smaller towns, Internet cafes are springing up. Yet Web access stops short of reaching millions of people in villages and remote towns – arguably the places where it could have its biggest impact.

Experts call this failure the "last mile problem." Rough terrain such as mountains or rain forests makes construction of a wired network in many far-flung places a prohibitively expensive option.

As a result, schools and community centers in many remote villages cannot hook up to the Web, even if Internet connections are available in nearby towns. Given the lack of reliable electricity in many developing regions, the prospect of wireless networks also did not seem a viable alternative – until now.

"You can't have a network if you don't have electricity," says Marc Pomerleau of the Green Wi-Fi project. But together with cofounder Bruce Baikie and partially funded by the One Laptop Per Child (OLPC) initiative, he has developed the prototype of a solar-powered wireless router. The device, planned to be unveiled in late September, is essentially an off-the-shelf wireless router connected to a battery that is recharged by a solar panel. To this, Mr. Pomerleau and Mr. Baikie added their own invention – an "intelligent charge controller" that regulates the power consumption of the router. Test runs indicate that this addition allows the wireless node to run for up to four weeks, even if the sky remains overcast for prolonged periods of time.

A mesh of these Green WiFi nodes mounted on rooftops could keep a wireless network between villages running for weeks. In principle, only one of the nodes needs to be connected to the Internet to enable the rest of the wireless network to access the Web – potentially via the $100 laptops, which are each fitted with a crank or foot pedal to generate the electricity needed to run them.

Such a network could have aided communication with communities cut off from power by the 2004 tsunami in Asia as well as neighborhoods affected by hurricane Katrina last year, says Pomerleau.

Setting up a backbone network to connect the hundreds of major access points throughout a region can be achieved by using three existing wireless technologies, says Daniel Aghion, executive director of the Wireless Internet Institute. WiFi hotspots serve areas extending up to several miles. WiMax, a more advanced standard, covers distances of up to 30 miles depending on the terrain between the access points. Third generation high-speed mobile networks may also be used, especially in places where cellphone networks already exist. Which technology works best depends on a range of factors, including population density, terrain, deployment costs, and how easily future increases in network size can be achieved.

"If I had to design a backbone network from scratch, I would use all three," says Mr. Aghion.

In fact, developing countries have a unique opportunity to "leapfrog wired infrastructure," he says. The lack of competition from wired – or "legacy" – technology may allow them to deploy wireless networks much more quickly than could be done in developed countries, Aghion asserts.

Existing technologies like WiFi have been tried and tested only on a small scale. "Using [WiFi] for long range communications requires changes," says Mubaraq Mishra of the University of California's Berkeley Wireless Research Center. He has analyzed wireless Internet pilot projects in the southern Indian state of Kerala. What needs to be modified includes the way a WiFi network handles the information it transmits. Chip companies also have to adjust their hardware. "They need to support long distance WiFi," says Mr. Mishra.

But the major obstacles to spreading the Web may lie elsewhere. Global technology companies eager to tap into what is potentially a huge market often cosponsor or run their own wireless initiatives in the developing world.

The innovative and cheap equipment developed by other organizations may be viewed as competition, hampering cooperation. "The $100 laptop is disruptive technology; there are some corporate interests that find us threatening," says Prof. Mary Lou Jepsen of MIT and chief technology officer of OLPC.

Government red tape relating to radio spectrum licensing, rights of way, or import duties poses another major challenge. "Regulation related costs [can] account for over 45 percent of the total costs," says Mishra.

Wireless Internet in developing countries is for the most part uncharted territory. The potential benefits to developing communities are big, as are the market opportunities for global companies. But in view of the many uncertainties, the initial costs of deploying wireless networks have to be kept small if the Web is ever to get off the ground in these nations – not least because "we don't have a handle on the demand," says Mishra.

The Internet may bring cheap communications to remote communities, allowing them to access weather forecasts and health information, track market prices, or simplify dealings with authorities. But just how big the demand is for such information remains unknown, says Mishra.

Formerly a part of Yugoslavia, and on the brink of ethnic war only a few years ago, Macedonia has become what may be the first "wireless country" in the world. With $3.9 million from the US Agency for International Development, the Macedonia Connects program has brought wireless Internet service to the country's 460 primary and secondary schools, which had already been equipped with 6,000 computers donated by the Chinese government.

On.Net, a Macedonian company, was contracted to build the $7 million wireless network. Instead of a myriad of WiFi hotspots, the network consists of a mesh of hot-zones – each stretching some 10 miles over cities and villages – that covers 95 percent of the country. After the schools were linked up almost a year ago, On.Net began offering the service to businesses and individuals. Now anybody with a WiFi enabled computer can tap into the network – free of charge in some areas, including parts of the capital, Skopje. Elsewhere in the country, users pay $20 or more a month for access.

In practice, however, only 13 percent of Macedonian households can actually access the Internet. People lack the money, have no computers, or as yet simply have no interest in the Web.