Why Google Should Bring Loon Networking to Android


People love large and shiny objects. So we can be forgiven for being absolutely blown away by Google’s idea of relaying IP across the skies via giant balloons to remote areas where Internet connectivity would otherwise not exist.

The most jaw-dropping aspect of the Loon project is the fact that the system uses algorithms to convert published windspeed and direction data into navigation using algorithms. (Balloons are moved by finding an altitude at which the air is moving in the right direction.)

So much about this project is dazzling — the scope and audacity of it; the solar-powered servers-in-the-sky; and the fact that balloons will deliver the Internet to remote areas — that the core aspect of Loon is easy to overlook.

The key thing about Loon is mesh networking.

The idea behind mesh networking is that instead of user terminals (phones, tablets and PCs) getting their connectivity via physically wired notes, such as cell towers and WiFi routers, each component in a mesh network is a sender, a receiver and a relayer of data.

Mesh technology extends networks far beyond the reach of the wired connection to the Internet. As long as there’s at least one relay chain of nodes, each within range of another, the network extends across long distances.

To set up a chain of notes in the sky is disruptive, but it doesn’t disrupt existing technology or monopoly or anything else except the absence of technology. Loon disrupts the limitations of poverty and geography. And that’s awesome.

But if Google wanted to be truly disruptive — really the most disruptive thing I can think of — they would bake Loon-like mesh networking right into Android.

Anyone finding themselves outside the range of both WiFi and mobile broadband data connections — or anyone who simply can’t afford a data connection, could simply turn on Mesh Networking in the Android phone’s settings, and not only get connectivity but share it with others as well by becoming a node in the network.

Unlike balloons, which could be shot down by repressive governments and which would not be deployed in areas where wireless carriers have political power, smartphone-based mesh networking could blend into the environment. Authorities and carriers could not tell who is using a licensed, paid account and who is using a mesh networking account.

Technically, phone-based mesh networking is feasible and has been demonstrated.

A project called the Smart Phone Ad-Hoc Networks (SPAN) used existing WiFi chips in Android phones to create mesh networking. Networks created with SPAN enabled not only text based communication after the 2010 earthquake in Haiti, but also VoIP — phone calls over the data network.

The trouble with SPAN was that the system had difficulty functioning with constantly shifting locations of phones. So another project to develop a more flexible routing protocol, called the Better-Approach-To-Mobile-Adhoc-Network (BATMAN), was launched and continues.

Using the BATMAN protocol, smartphones can connect to each other from up to 100 feet away, and VoIP works up to five hops from the physical connection.

Another group of researchers from Australia’s Flinders University found a way to extend Android phone calls into the Outback and beyond the reach of cell towers.

The geniuses behind this project still appear to be working on this. They call it the Serval project.

These projects are poorly funded and limited in scope. As a result, they have not lead to anything sustainably usable and they’re not getting the traction they should get.

They also don’t have access to the more broadly deployed Android code or have any influence over how Android handsets are designed and built.

What Android mesh networking needs is a big-budget, Google-style “Loon shot” project that results in mesh networking being built into every “Google Experience” or “Nexus Experience” phone.

And, come to think of it, I’m wondering if this capability might be built into the upcoming Moto X phone everybody except Google is talking about. According to rumors, Google is doing something very “different” with the Moto X, something to do with brand new technology.

Motorola CEO Dennis Woodside even hinted that the new phone would “change people’s lives.”

And finally, we hear that the Moto X phone might be priced very low — as in, targeted at emerging markets.

A Moto X phone with built-in mesh networking in which mobile broadband and Wi-Fi connectivity are optional would fit all the rumors.

Either way, Google is the right company to mainstream mesh networking. And not just any mesh networking, but really good mesh networking that’s fast, flexible and long-range. Google’s skill with algorithms is probably the key to making mesh networking function well at scale.

Such a feature would be the most disruptive deployment of technology in history, putting Internet connectivity into the hands of the poor, the remote, the oppressed — and bringing life-saving communication to people in disaster areas and during power failures.

While Loon balloons can bring Internet connectivity to remote places, Android mesh networking could spread it around once it gets there.


(Pictured: Paul Gardner-Stephen, technical architect, co-founder of the Serval Project)