The Dangers of Off-Grid Social Networking in Hong Kong and Beyond
BY Sascha Meinrath | Thursday, October 2 2014
By popular demand... A quick overview post on (just a few) of the reasons all mesh isn't created equal (a.k.a., why Wi-Fi Direct is not a replacement for ad-hoc mesh and why Hong Kong coverage of mesh "off-the-grid" communications is scary-wrong):
1. Let me just start by saying the folks at Open Garden are damn cool, and the FireChat application is super-neat in what it can do. I view the project as a fellow traveller and support both their work and the innovations that they're making to bring distributed communications to the masses. This post is meant to discuss some of the issues that have been left out of mainstream reporting, *not* to be a rationale for railing against FireChat.
2. The awesome folks at Citizen Lab did a great write-up of FireChat, here (just scroll down to the FireChat section). Most importantly, in the words of Open Garden's Vice President of Sales and Marketing, Christophe Daligault:
People need to understand that this [FireChat] is not a tool to communicate anything that would put them in a harmful situation if it were to be discovered by somebody who’s hostile. It was not meant for secure or private communications.
3. Wi-Fi Direct has an affiliation-security problem. As cryptologist Dan Harkins summarized:
It’s 34 frames of compromised and questionable security versus 8 of strong security. And both options can use the identical credential for authentication, it’s just that mesh does it right and WFD does it wrong.
In non-geek speak, it means that users of any system relying on Wi-Fi Direct may leak personal information whenever they join together (say, to create a FireChat group). This is a problem endemic to the underlying tech -- which means that anonymity is *way* more difficult to maintain.
4. Link-state changes: Wi-Fi Direct uses infrastructure mode (i.e., a hub-and-spoke model where clients can link to a "group owner" node); while groups can be linked together, this ownership has profound implications for propagating link-state changes (e.g., mobility, changing RF environments, etc.).
Furthermore, if the owner (who acts as an access point for everyone connected to her) leaves a group, the group is entirely disbanded. On the more scary side, owners could also silence all affiliated devices (via the "notice of absence" mechanism in Wi-Fi Direct) -- which means they can forcibly break links between themselves and their clients as well as any clients that those devices are connected via them. Again, this is endemic to the underlying tech and is one of the reasons why ad-hoc mesh is such a different beast than systems that interlink a bunch of hub-and-spoke architectures together.
5. Relatedly, there is an issue with scalability. The routing overhead necessary to create large-scale networks makes Wi-Fi Direct untenable. Judging from the specs, routing overhead is linear (and possibly supra-linear), making large-scale networks difficult (and possibly impossible). See Table 3 here.
If I had to bet, we've seen very few multi-hop Wi-Fi Direct-based networks (say with 4 or 5 or more hops), and probably nothing greater than a few hundred users on any one off-grid network (looking at all the stats in the media, they talk about simultaneous users, not size of any FireChat network -- and yes, if you're wondering, the only way they could know simultaneous users is if FireChat clients were phoning home with this data).
6. I'm more than a little scared about the overhyping of FireChat -- not because it's not a cool technology, but because it's being *way* oversold for a use case where people may be surveilled. For example, I'm not sure how it can be completely "off the grid" but still require registration of users with a central database -- that would seem to require being on the grid (as do the usage stats that are being collected).
FireChat is synergistic with, but completely different from Commotion Wireless and both have their own strengths and weaknesses (none are silver bullets for all situations). But, in this case, the over-hyping could be putting *a lot* of people at substantial risk (and could create substantial blow-back for many other important mesh and circumvention tech projects that are doing great work all around the globe).
Sascha Meinrath is the founder of X-Lab, an innovative new platform that will address the future of tech policy. He was formerly the Director of the New America Foundation's Open Technology Institute. This post, which originally appeared on Facebook, has been lightly edited and reposted with his permission.