Currently being trialled in New Zealand, each balloon delivers a coverage area of 1250 square kilometres as it floats overhead. New Zealanders who want to access the service must have a special antenna fitted to their house that connects to the closest balloon. The signal is then bounced from balloon to balloon, until it joins the internet back on the ground. Solar panels power the balloons' antennas and communications equipment, storing energy in batteries to keep them working through the night.
Mariya Zheleva, who works on wireless networks for remote developing regions, says the project is inspiring: "I get very excited about solutions like this one that try to escape the conventional understanding for communication infrastructure."
Google claims that its setup allows it to deliver "speeds comparable to 3G", between balloons and the ground. It is unclear how well applications which rely on short communications times, or pings, like VOIP, will work given that the signal must relay through multiple balloons before even reaching to the wider internet.
Stephane Boyera, an information and communication technology (ICT) development consultant formerly with the Web Foundation, says he is convinced that Loon is a technically promising solution, but worries that it ignores a huge number of other problems, which may limit its broader impact.
"Connectivity, while being a problem, is not the primary bottleneck in providing ICT services which can improve people's lives," says Boyera. The real problem is two-pronged – a lack of content that's useful to people in developing countries, and the lack of widespread, Wi-Fi compatible devices, he says.
"It is a total myth to imagine a farmer in Mali using Google to find solutions for a disease his tomatoes have. Barriers are just huge: illiteracy, language, ICT training," Boyera says. The existing web is not that useful to the underprivileged populations of developing countries, and no amount of new connectivity options can fix that, he says.
Boyera also warns that strategies which rely on top-down hardware distribution are historically prone to failure. "I believe that the only scalable option is to focus on what is already available," he says.
Google Loon is not the first project to attempt to deliver internet access over a wide area but, as the project name suggests, it is probably the kookiest.
Google will rely on weather prediction to keep its balloons in the right place, moving them up and down to take advantage of different air currents. "Project Loon uses software algorithms to determine where its balloons need to go, then moves each one into a layer of wind blowing in the right direction," Google announced. "By moving with the wind, the balloons can be arranged to form one large communications network."
Zheleva warns that a balloon-based system faces some fairly unique technical challenges. "Stabilisation of the balloons is of great importance, since a slight shift of the balloon due to wind can result in major shift of the coverage area on the ground," she says.
Flight time is another issue. The Raven Aerostar balloons Google is using typically have a maximum flight time of 55 days, meaning that the floating internet links will need to be either replaced on a regular basis, or replenished with helium while aloft. Google claims that its tweaked design can stay aloft for more than 100 days.
A similar project called Antarctic Broadband is planning to use small, cheap satellites to bring steady, reliable internet to the continent. Currently researchers there must rely on signals from geostationary satellites, which are above the equator, meaning their signals barely reach the edges of the Antarctic.
Unlike Project Loon, Antarctic Broadband will not connect users to a nearby internet relay through its satellites, but to dedicated ground stations in western Australia and South America.
See also: Project Loon on Google+
Syndicated content: Hal Hodson, New Scientist