ICTworks by Inveneo

The WiFi hackers at Inveneo, famous for their long-distance network in Haiti, are piloting a new Internet connectivity solution that they hope will bring Internet access to currently unreachable corners of rural Africa. Deploying collar-mounted wireless transceivers on a herd's worth of goats, they've created a mobile, self-healing, self-configuring network they've dubbed the Goat*Net.

Rural WiFi challenge

Rural Africa suffers from two distinct disadvantages that hinder the easy and affordable provision of Internet access - long distances between users and no electrical infrastructure. One possible solution to these challenges often employed by Inveneo is long-distance WiFi.

Even WiFi has its limits, though. On a recent trip to Ethiopa, Inveneo found rugged terrain and highly mobile populations would make a standard WiFi deployment difficult or impossible. Transient herdsmen rarely stop at one location for an extended period of time and often wander in and out of coverage areas of traditional networking solutions.

The hairiest ISP you've ever met

The challenge became how to give these herdsmen access to the same benefits of information and communication technologies (ICT) as Africa's more stationary populations.

Goat*Net: Bleating Edge Rural WiFi

While conducting a site survey of the environment and ICT needs of Afar goat herders in the Ethiopian Danakil Desert, Inveneo Project Engineer Andris Bjornson was hard pressed to come up with a traditional WiFi solution that worked:

"The rolling hills and many tall trees were making it difficult to find lines of sight that would make the network feasible"

One hot afternoon while pausing for a mishkaki snack, Andris had an intriguing thought - why not build a mobile WiFi mesh network using goats? This was the start of Goat*Net.

A mesh network is different from more traditional networks because each mesh node acts as an independent router. A mesh network continuously reconfigures or "heals" itself - finding ways around broken or blocked paths.

Inveneo paired Ubiquiti PicoStations mounted on the collars of goats with stationary Ubiquiti Airmax sector antennas to form a grazing, wandering, self-healing mobile network. As goats wander from ridgetops to valleys, the network continually adapts routing to find the best paths back to the stationary antennas and eventually to the Internet.

Goat*Net Mesh Topology

Currently, the PicoStations are powered by long-life rechargeable batteries, but Inveneo is investigating a biogas power solution, enabling the devices to be powered by waste methane produced by the goats.

Goat*Net has had a huge impact on the lives of Afar goat herders, bringing the community into the modern world. As one was quoted as saying:

Goat*Net has brought real change to our community. We can now get CBOT spot prices for nyama choma

Inveneo is excited to push this new technology even further as part of its Goat Ahead Program^TM to bring technology to ewe. According to Inveneo CIO Mark Summer:

"We've already been able to do so much with just 50 goats in a network, but that's just a start. We have plans to scale up to a full 1 kilogoat next year. As wireless radios shrink further and prices come down, the dream of a full megagoat network could be realized within 5 to 10 years."

It's the classic dilemma: Google Earth has the fantastic interface you need to keep track of your GPS data, and high resolution imagery that lets you find the exact spot on the planet you're looking for. Radio Mobile puts powerful tools of radio line of sight calculation at your fingertips, but its interface isn't nearly as polished as Google Earth's. Where do you put your data first?

Do both! Radio Mobile units can be imported and exported seamlessly as KML files. In Radio Mobile, look under "File -> Units Properties" and click the Export or Import buttons on the right side of the window to get started. (screenshot)

Saving Google Earth places to KML

Importing a KML file to Google Earth is as simple as choosing "File -> Open". Exporting from Google Earth is almost as easy...just right click on the waypoint or folder full of waypoints and choose "Save Place As..."

When exporting, make sure you save your points as KML files (rather than the default compressed KMZ) otherwise Radio Mobile won't know what to do with them.

Thanks to import/export, you can use Google Earth as a tool to download your waypoints directly from your GPS, organize them into folders, and then export them to Radio Mobile. This saves time and helps you avoid typing mistakes that could cause you real problems later.

After you've modeled your wireless network, take the time to export any new radio sites you've created back to Google Earth and into your GPS. This will make it easy to find your project sites and turn them from points on a map into antennas on the ground.

Want more on Radio Mobile? Subscribe to ICTworks via RSS, Email, Twitter, or Facebook

.

In 1999, the US and Japan jointly launched the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) into orbit. One of the goals of the ASTER instrument mission was to conduct detailed terrain mapping of the surface of the earth.

Colorized ASTER GDEM Map of Africa

The ASTER global digital elevation model (GDEM) covers the planet from 83 degrees north to 83 degrees south at 1 arcsecond resolution and is the "most detailed 3d map of the Earth ever made."

In June of 2009, the ASTER GDEM was released to the public.

For us as builders of long distance wireless networks, this is mostly of interest because it allows us to do more accurate computer modeling of radio lines of sight to answer the basic question "If I build antennas at point A and point B...can they talk to each other?"

If you want to try it out for youself, see our detailed technical guide on how to download and convert this data to a format compatible with RadioMobile, the widely used free software for modeling radio propagation.

• Obtaining ASTER GDEM Data and Converting to SRTM1 for Use With RadioMobile

We recommend ASTER GDEM in addition to SRTM3 (previously the highest global terrain model available) as some of our partners in Nepal have had significant difficulties with "voids" or blank spots in the SRTM3 data right in the middle of their project sites.

Using SRTM3, RadioMobile will tell you a line of sight through a void will work, when in the real world it might not. We have seen significantly fewer voids in the ASTER GDEM data. If SRTM3 has voids where you need to plan a link, try the ASTER data.

As we mentioned in our recent post about testing WiFi antennas across the San Francisco Bay, we at Inveneo rely heavily on Google Earth to better understand the complex wireless network projects we implement in rural and developing regions.

Google Earth is fantastic tool for understanding any project involving geospatial data on a macro as well as a micro level.  Google Earth can also help you better communicate the where component of your project to your coworkers, your partners, and your funders.

If you're working in a remote rural region and have tried to put Google Earth to work on your project, though, chances are you've run into it's one key limitation:  it relies heavily on the Internet.

I recently discovered a free tool I'm excited to try out that looks like it will play a big part in my workflow as I analyze wireless networks for Inveneo.  Google Earth Voyager allows you to automatically fill your Google Earth cache with data for a project region you define. 

You can run Voyager to fill your cache when you have a fast connection to the Internet, and have all the data you need available offline when you're operating on your laptop out in the field.

Google Earth Voyager was developed by an Egyptian software engineer named Tamer Louis.  You'll find the direct link to the program zip file here.  The file is currently called gev6.zip, though it looks like the filename changes with each new version Tamer rolls out.

The guys over at Google Earth Blog are also thinking about this problem, and have provided a great writeup on optimizing your cache settings for offline use.  They also link to an interesting forum discussion on saving multiple versions of your GE cache for different regions.  You'll need to quit GE to swap the caches, basically tricking the program into thinking there's only one cache.

We'd love to hear about any other techniques you have found useful for dealing with Google Earth in offline locations out in the field.