[ 
home | about twcn | nodes | projects | docs | search | index ]
 |
|
|
|
|
As William Waites has put it, the flexibility/accessibility versus resistance to abuse debate will become prominent as the wireless network grows. Whether technology can completely satisfy all the concerns remains to be seen. One proposed solution is by a group called NoCatNet, sometimes referred to as NoCatAuth. For a more indepth analysis of the NoCatAuth project see their whitepaper (in text format) on it.
A general overview of how Jason Hammerschmidt sees the evolution and handling of the abuse vs. accessibility topic in TWCN follows this text below. As with the original posting, this overview may be better understood if you have read the Wireless Network Access Point Set Up Guide.
Jason Hammerschmidt
Wed, 23 Jan 2002 08:44:35 -0800 (PST)
A three tier approach to users of TWCN will probably work best. There's lots of (some unjust, some just) concern about hording bandwidth, leechers, transit of unwanted packets for various reasons (pr0n, DoS, crackers). This could be mitigated by Type III style nodes with rate limiters and proper firewalling and a three tier approach to user classes. So we should be aiming at supplying people with such nodes. I think the WRP project is a good idea for this. Companies like ReSource may be able to supply us with cheap hardware (PC's). We could build or buy at large discount cost electrical equipment such as amplifiers and antennas. But I'm getting too ahead of myself. A three tier approach to user classes will also help us mitigate such unwanted things.
Tier 1 users are the owners physically located at the given moment in time at thier premises using thier equipment. They have full unabridged access to all thier available bandwidth with the highest priority. I call tier 1 the owners.
Tier 2 users are owners in another part of the city. So if the owner of the TWCN Hammerschmidt node travels to the node vacinity of TWCN Emir, he's considered tier 2 class. I call tier 2 the co-ops.
Tier 3 users are the public. No owned infrastructure of thier own other than a computing device they use to connect. This class would be rate limited and/or bandwidth limited (depending on the node and the node owners paranoia quotient), have the lowest priority, be firewalled the most, etc.
You may be asking how the technology of such a system would work. Well authentication (think an Open Toronto Certificate Authority) along with tunnels (think IPSec) could be used together. Authentication to say "hey I'm a co-op in another place", no authentication means you're the public. I'm not sure if you could rate limit etc. on tunnels, but it is concievable. And we're talking about something that's so far away, by the time we get there, something will be in place (and if not, we hack some FreeNIX source :).
However a three tier system means jack unless you have a dense enough population of nodes. Furthermore many such concerns are right now handled on an individual basis. It doesn't make much sense to invest time, money, effort into such silly ideas at the moment. Instead we should be aiming towards node density or critical mass. I call this stage 1. There are three stages of evolution in my world Shirley.
Stage 1, evolution to the point of dense overlapping nodes, or critical node mass. We'll have to sort out things like boundaries, channel allocation, antenna's/amplification, hidden station problems, all of that. But that's what we're here to do. To organize ourselves. We could loosely all share, but that wouldn't take us that far. The goal here is to provide ubiquitos wireless networking everywhere, it may not be perfect, it may not route, it may not provide roaming across WNAP's, but at least it's ubiquitos. This will probably be the biggest hurdle, trying to show everyone.
Stage 2, routing via nodes to other nodes. Completely bypassing the public wired networks (ie. Ma'Bell) using either our own tethered means, or wireless means. In effect we create our own massive internetwork, operating independantly of the big carriers' support. Tieing into a few central peering and uplink points (say TORIX) for transnational transit. The goal here is to be independant of any commercial entity, the Internet in Toronto becomes decentralized again, we own it!
Stage 3, Roaming and three Tiers. It would be an awesome thing to implement once we finish stage 2, roaming features such as described in one of the many IETF RFC's on the subject. Combining this with the three tiers from above provides me the affordability to get instant messanging alerts while driving down Dundas in my car all while using the same single address.
The beauty of stage 3 is simple. But it's a deeply philosophical thing involving the enlightenment of the masses. To much to go into here but feel free to ask me during a monthly meeting.
On the topic of supplying people with nodes, we should be doing our best to supply them with Type III's. I doubt Type I's can evolve to the above dreams.
On the topic of evolution and dreams, it should be noted this is pretty much a big pipe dream, and the evolution from one stage to another will surely be organic. You'll see some stage 2 development while stage 1 is still being built in other areas. There will obviously be no overnight solutions. As well, while I don't see much use in implementing the three tier system until stage three, undoubtedly we may see some inroads during stages one or two.
On the topic of credit, I'm a firm believer that everything is just a rehash of older ideas, we're standing on the shoulders of giants so to speak. The three tier approach comes from the NoCAT project. The multiple stages came from various ideas from various people in various mediums. It's sort of a view of how I see the natural organic growth of TWCN.
--
Jason Hammerschmidt
If you're not paranoid, you just haven't been paying attention.
If you would like to stay informed on current events at Toronto Wireless Community Network, please join our mailing list.
|
© 2002, Toronto Wireless Community Network (TWCN). All Rights Reserved. |