Are Community Wi-Fi Networks Doomed to Repeat the Security Mistakes of the Early Internet?

Posted At : January 19, 2005 6:38 AM | Posted By : Cameron
Related Categories: Wireless

Last night I attended a meeting of the Socal Free Net (formerly the San Diego Wireless Users Group) and heard about some very interesting and exciting projects they are doing in the communities of San Diego. They've been installing free wifi hotspots in (primarily low income) communities around San Diego.

Someone donates the equipment and the bandwidth and the folks of Socal Free Net provide the labor to setup the hotspots. When they get an opportunity to put up a hotspot - these guys get on rooftops, mount antennas, form rooftop-to-rooftop links between nodes, and stretch the signal to as much of the community as they can. It's really great stuff, free and open internet for anyone within range. It's a worthwhile project and I may join them at their next install this weekend.

These projects are very ambitious and promise to offer free internet to alot of people who may not otherwise be able to afford it. However, all this talk of free and open internet has me thinking - are these projects doomed to repeat the same mistakes that plagued the early internet?

In the early days of what is now the internet, everyone on the network was trusted to a certain extent. In the beginning, protocols were open, alot of stuff was sent in cleartext, and there were few failsafes preventing malicious activity such as Denial of Service attacks. As more users came onto the internet, worms started appearing, DoS attacks started taking advantage of weaknesses in underlying protocols, and the internet because a nastier place. Things have changes since those early days, and today the internet is a far different place.

So what does this have to do with the Socal Free Net projects? The spirit of these projects is to allow free and open access to everyone in the community. No-one is restricted. If you can physically get within range of the network, you can use the network. Also, due to the inherent difficulties in teaching an entire community how to enable WEP/WPA security, it's not enabled on these access points. Cleartext is visible in the air, and it's up to the individual user of the network to take whatever measured they see fit to secure themselves.

I see many similarities between these wifi hotspots and the mistakes (if you want to call them that) made in the early days of the internet. Eventually someone's going to figure out that they can sniff their neighbor's network traffic. Eventually someone's going to figure out that hardly anyone encrypts their POP account password when they check their email. And eventually someone is going to realize that a person's POP login is often the same as the one they use for online banking, forums, and the recipe of the week club website.

So, you say, what's the difference between this and any other wired network today? Alot of wired community networks have this same problem! Here's the difference: You don't have to live in the community to participate on this network. If you want to listen to traffic on a wired network, you generally have to either physically plug into it or compromise a box on the local subnet or at a router. With these networks, all you have to do is drive over within range of the network.

So let's skip forward a few years. June 2008 (why not?). There are now open networks in communities all over town. All over lots of towns. Some software programmer somewhere has just released net wireless application, and she built it to take advantage of the abundance of open wifi hotspots found in her local city. The software runs on a wifi enabled laptop or Pocket PC and constantly looks for open networks. When it finds them it starts gathering packets sent across the network. The software picks out POP passwords and logs them to a file, it picks out HTTP form posts and logs them to a file, it might even pick out data patterns, whole email messages or HTTP request/replies and logs them to a file.

Thousands of script kiddies get ahold of this software. People with bad intentions get ahold of this software. They load it up on their Pocket PC and just drop it into their pocket/purse/backpack, going about their day like they normally would. At the end of the day/week/month they take a look at their bounty. Files and files chuck full of personal data, full of passwords, full of account information.

Maybe this already exists.

Sure, if you're reading this blog you probably already have your own access point with VPN over AES encrypted WPA and work/live inside a Faraday cage. but what about these community networks? For that matter what about your neighbor 2 houses over who thought their access point was "plug and play"?

Are we headed down the same road all over again? It's easy to say the responsibility of security resides with the end user, but with these community networks, can we really expect the average end user to know how to configure themselves to communicate over wifi securely? Is it even offered to them?

I'm not really sure what the answers to these problems are, but those are my thoughts of the day about wifi, and it's about time I posted *something* to my blog.

Comments
Adam Howitt's Gravatar It's a very interesting thread but there can be a distinction between the community networks and the commercially provided ones. Take SBC Yahooo! DSL. For a small fee you can upgrade to a wireless router. They provide you with a 2Wire wireless router with WEP configured by default. The default complex password to each is located on the base of the router itself and to connect you must follow the installation instructions. I'm not implying that WEP is a safe protocol by any means (on the contrary I've heard techsnobs rant on THAT subject on many an occasion).

Are other commercially available wireless packages doing this type of thing?
# Posted By Adam Howitt | 1/19/05 8:16 AM
Adam Howitt's Gravatar Oh! and welcome back to blogging Cam ;-)
# Posted By Adam Howitt | 1/19/05 8:16 AM
Cameron Childress's Gravatar Thanks!

Funny you would mention that. At the meeting last night someone mentioned that those 2Wire routers ship with the wireless turned on by default. Apparently, alot of people get them and don't use the wireless, not knowing that it's turned on even if they never used it. In densely populated areas this causes problems for people who ARE trying to use a wireless network because the noise coming off the 2Wire router interferes with other networks in the area.

Worse than that, because the people who own them never used them, they have no idea they are causing this problem for their neighbors. Before Christmas I could see 4 wireless networks from inside my apartment. After Christmas I could see 6. The problem of crowding in the 802.11b/g space is only going to get worse, and those 2Wire devices aren't helping.
# Posted By Cameron Childress | 1/19/05 10:31 AM
Jonathan Bigelow's Gravatar Good post Cameron, and, yes, welcome back to blogging. Good to hear from you.

Definitely a good point and something to consider. I did a little Googling and research on the IEEE website and was surprised to not find anything about future wireless protocols implementing PKI.

If encryption can be handled automatically between me and my bank, or Win2K+ servers for that matter, I'm curious why nobody is planning on implementing PKI (regardless of what encryption is used) for wireless. It would eliminate the need to share and distribute WEP keys and the fact that if one had the key string, one could (granted with a little work) decrypt the data traversing a WEP-enabled network anyway. Thoughts?
# Posted By Jonathan Bigelow | 1/19/05 3:55 PM

Recent Entries

Archives By Subject

Tech Blogs

(Mostly) Not Tech Blogs