The self-described "breaker of things", Josh has created a kernel module for Android that implements a wifi mesh network with smartphones. I caught up to him at a mobile security conference recently, and we got to talking. Here's what he has to say about his work.
Gandi: The SPAN project links wifi capable smart phones in a mesh network. Why did you create it?
m0nk: Before the project, I always felt slightly skeptical about our societal reliance on the cellular infrastructure. The service providers do a wonderful job, but in the end they are for-profit corporations, focused on ROI and the like. I honestly commend them on the reliability they have attained, but it's far from 100%.
Then Fukushima happened. I was taking paternal leave from work as my son had just been born, and the Internet just exploded with news and heartbreak. The one story that really got me was a YouTube video of a family trying to call each other to say "Hey, I'm alive". The cell phones just said "no service". I remember thinking, why are we so reliant on a technology that doesn't support us when we need it most?
I started doing research. I learned that the worldwide cellular infrastructure is designed to handle only 80-90% capacity. In almost every major catastrophe, the network dies. I also learned that most emergency management personal rely on the cell infrastructure, so it isn't just families not being in communication, it's the people trying to help and manage the crisis as well. I decided it was time to fix this problem, or at least attempt to show the world it was fixable.
Gandi: So who does the SPAN project benefit?
m0nk: I'd like to think it has the potential to benefit everyone with a cell phone. It's my belief that meshed cellular communication should be ubiquitous across the globe, and anything I can do to promote that cause will get us one step closer to never losing communications again.
Gandi: What’s the biggest barrier to adoption of the SPAN project?
m0nk: In America, we like to think positively. We don't really ever consider disaster striking so we never really think about being prepared. Those that do think about it are sometimes labeled as "crazy". Though some are, the entire mentality is usually dismissed from the public consciousness. That's a barrier.
In the end, all people need to do to adopt SPAN is download and install an application on their phones that they will hopefully never use. To be realistic, I just don't see that happening en masse.
Therefore, I am hoping to one day catch the eyes of Google and Apple, and let them adopt the technology into the backbones of the actual production phones. I can show them it's simple software running on their phones (so the cost is low), but I cannot provide them with an impetus beyond "it's for the greater good".
Gandi: Does the proliferation of Internet-connected devices make mesh networks more or less important?
m0nk: Probably more. The more reliant we as a global society are on the Internet, the more important it becomes for daily life. If connectivity is removed, the effect increases with the number of devices that rely on it.
Gandi: What do the carriers (ATT, Sprint, etc) think of your project?
m0nk: I've heard the full gamut from "Hey, that's really cool" to "It does VoIP and free internet tethering, we DO NOT approve". They might fight the project, since it shows there is a potential problem, and the technology allows for bypassing the network for VoIP calls. As someone who pays the bills, I wouldn't really fault them for protecting their public image and revenue stream. I just think it's short sighted.
Gandi: Do you have any allies in the Internet industry?
m0nk: Not really, I'm just a guy with a loud mouth who got a little bit of FFRDC funding to explore this concept. I spend most of my mesh time these days evangelizing the technology and trying to raise awareness.
Gandi: Is security of mesh networks a problem?
m0nk: It depends on the deployment. For situations like the Arab spring, very much so. If the government shuts the Internet down to keep people from organizing, you have to expect they will also monitor and track the mesh network if they can. Oddly enough, emergency response personnel also need security, to avoid terrorist snooping, for example. For a general deployment across the landscape, I'd expect security would actually come second to bandwidth.
In the end, we have spent a ton of time working the security angle of the project. We have some code and a lot of ideas. What we lack is time and sponsors to solve it.
I am trying to prove a point with the SPAN project, namely "Mesh networking is a viable alternative / backup plan for emergencies and we as a global culture should adopt it." I'm not trying to say that SPAN itself is the answer, it is solely my implementation. Maybe someone will be as riled up as me, and solve the security problem before we do. That would be awesome. If not, we will solve it in due time.
We at Gandi love to hear what the leading edge people like Josh are up to, and to help out when we can. As the technology evolves to make innovations like mesh networks possible, Gandi will do it's part to promote them for the common good.
Here is some more information on Josh's work:
Presentations by Josh
SPAN Project Code
Amphion Forum Powerpoint