This thread is a couple weeks old, but I felt this is relevant. The discussion dealt with microsegmentation for wireless networks, and lo and behold this popped up on slashdot: http://frottle.sourceforge.net/.
Apparently, someone else decided that wireless traffic needed better management. From the site: Frottle (Freenet throttle) is an open source GNU GPL project to control traffic on wireless networks. Such control eliminates the common hidden-node effect even on large scale wireless networks. Frottle is currently only available for Linux wireless gateways using iptables firewalls, with plans to develop a windows client in the future. Frottle works by scheduling the traffic of each client, using a master node to co-ordinate actions. This eliminates collisions, and prevents clients with stronger signals from receiving bandwidth bias. Frottle has been developed and tested on the large community wireless network of WaFreeNet. We have found running frottle has given us a significant improvment in the network usability. Testing results will be documented here as time permits. RAther than find a way to segment the RF spectrum, it looks like they're running standard 802.11b wireless in conjunction with a deterministic token passing method. In addition to granting a host permission to send, the token tells a host how much data it can send. Good times. Enjoy, Hal > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Sent: Monday, July 21, 2003 6:06 PM > To: [EMAIL PROTECTED] > Subject: Re: Bandwidth, QoS, and Contention networks [7:72645] > > > Quick addendum: > > As Howard mentioned, the problem with a contention medium > isn't only how > much bandwidth is available but also how quickly a station > can access it. If > the stataion listens before it sends and can't send for long > periods of time > because the medium is not free, this is bad news for voice and video. > > Shared Ethernet has gone the way of LocalTalk, (almost?) but > shared wireless > is gaining popularity, as Chuck has been saying. I haven't > seen any studies > yet that address medium access delay on wireless networks > based on load and > access characteristics, but maybe they will get published at > some point, > (although it's pretty unpredicatable with bursy data sources, > of course). > > Suffice it to say, "microsegmentation" for wireless networks will be > necessary to some extent, just like it was with shared > Ethernet, depending > on delay and delay variance requirements of the applications. > > Priscilla > > "Chuck Whose Road is Ever Shorte wrote: > > > > ""Zsombor Papp"" wrote in message > > news:[EMAIL PROTECTED] > > > "Chuck Whose Road is Ever Shorte wrote: > > > > after viewing the presentation, you tell me - is > > > > this not saying > > > > that 5 megabits is more than adequate for voice, video, etc? > > > > > > I don't think that was the point of the presention. > > Regardless, I can tell > > > you that 5Mbps is enough for voice. :) For video, it depends > > on the > > quality. > > > > 30 frame per second video over ISDN requires 3 ISDN lines ( 6 > > B-channels ) > > for a total of 384K. I've spoken about video over WAN links on > > several > > occasions with the video guy in my group. He tells me he likes > > to reserve > > 500K over T1's typically. > > > > Howard, if your reading, I just looked over my unsent drafts of > > the question > > regarding bandwidth, and the point I failed to raise in this > > posted thread > > is that of global synchronization. One of the major benefits of > > such QoS > > mechanisms as RED and WRED is that the phenomenon of global > > synch can be > > controlled, meaning a more efficient use of bandwidth ( no > > periods of > > congestion followed by periods of low activity because of the > > TCP backoff > > mechanisms ) > > > > Now supposing, even in a contention medium, I could fine tune > > my queueing > > such that I no longer suffered from global synch. I set my > > voice queues and > > my delay sensative queues such that sufficient badwidth was > > available, and I > > used RED or WRED for the general queue. I'm wondering if there > > are studies > > done, papers in the IETF working groups demonstrating that > > given proper > > queueing mechanisms that less bandwidth is required or > > necessary? > > > > Cisco does offer downstream only QoS on their wireless product > > line. Not > > sure I understand the mechanism completely, but I have to > > believe it is > > based on enough solid study such that given a reasonable > > design, there would > > be less concern for voice in the mix. Recall that Cisco will > > soon be > > releasing their own wireless IP phone, and the whole point of > > it is to > > provide untethered mobility throughout an enterprise. > > > > > > > > > > > > > I'm just asking - if people are more productive, > > > > despite the obvious lack of bandwidth, and > > > > despite the step back to a contention medium, > > > > is there something to be said about the > > > > perceived need for 100 megabits to the desktop? > > > > > > I agree with Fred, such a perception is probably misguided in > > most cases. > > > Most people are very happy even with their 1.5Mbps DSL line. > > But all this > > > depends on what you want to do. Full-screen DVD quality video > > won't work > > > over DSL. Even online gaming could use more than 1.5Mbps. > > FWIW, I've heard > > > that in Korea, there is a serious market for dedicated > > 100Mbps connections > > > to the *home* due to wide-spread online gaming (I don't know > > if this is > > > true, I find it a bit hard to believe). > > > > > > can't comment much regarding the needs of on-line gaming, but > > the guy > > teaching the wireless class two weeks ago said he was an avvid > > on line gamer > > and that his DSL was plenty fine for what he did. He also said > > he had a > > couple friends around his neighborhood who did on-line gaming > > via a sireless > > AP that he set up for their use. Don't know the particular > > game, so I can > > say as to whether it is the same one you play. > > > > > > > > Also consider that pure 10Mbps Ethernet interfaces are > > getting pretty > > rare; > > > most of the Ethernet interfaces are 10/100. So in a campus > > network, in > > most > > > cases, there is no real reason to not have 100Mbps to the > > desktop. > > > > indeed. and with three teenage boys around the house, I am > > happy to sell > > lots of these things in order to keep the refridgerator full. > > ;-> yes, Mr > > Customer, you never can tell when your users will need this > > bandwidth, what > > with internet radio, lots of databases out there on the web, > > and all the > > e-mail attachments people need to read to get their work done. > > ;-> > > > > > > > > > > > > Thanks, > > > > > > Zsombor Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=73478&t=73478 -------------------------------------------------- **Please support GroupStudy by purchasing from the GroupStudy Store: http://shop.groupstudy.com FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
