I talk about max achievable throughput of 802.11n ,that may be got in ideal conditions. It is obvious that noask gives more throughput then with ask in ideal ( for example in lab via coax cable ptp connection without collisions and interference ) conditions. I agree that in noisy environment using of proprietary protocol like Nstreme may give more stable link and higher throughput then standard protocol 802.11a/b/g/n due to it's some useful features, like packet aggregation and link adaptation. Does Airmax have simular features? Is it only noask ? Noask is acceptable when packet losses rate due to interference are very small. Throughput is not only one link parameter. Also packets losses rate(BER) , delays, jitter, MOS ( VoIP applications) and others are also important. Using of proprieatry protocol in 802.11 a/g/n units very often help to improve these link parameters. We carried out field test of standard and proprietary 802.11n systems in various LOS and nearLOS conditions. One of the conclusion that Nstreme ON in 802.11n is not able to improve link in case multipath fading interference, but Nstreme On in 802.11a in the same link really makes connection more stable then in standard 802.11a/n mode.
Vyacheslav Vasilyev Unidata 2010/9/6 Scott Carullo <sc...@brevardwireless.com> > I'm not sure your assessment of UBNT not recommending to use airmax on PTP > as a general statement hold true. It is unlikely that they would have built > in a specific PTP noack mode into airmax configuration if it was their > suggestion not to use it. I use it on lots of links and it works very well. > > You are incorrect in saying airmax and nstream cannot increase throughput > in comparison with 802.11n. In the real world in the wild a lot of times it > is only airmax or nstream that will even let a link perform reliably, > regardless of what the textbook says. Not to mention it allows our > throughput to increase in contrast to your statement. > > > Scott Carullo > Technical Operations > 877-804-3001 x102 > > > > ------------------------------ > *From*: "Vyacheslav Vasilyev" <s...@unidata.com.ua> > *Sent*: Sunday, September 05, 2010 3:50 PM > *To*: "WISPA General List" <wireless@wispa.org> > > *Subject*: Re: [WISPA] mikrotik vs ubiquiti > > > > > 2010/9/5 Jeromie Reeves <jree...@18-30chat.net> > >> On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev <s...@unidata.com.ua> >> wrote: >> > >> What is a Rockets PPS with airmax on? >> > Ubnt does not recomend to use Airmax On in ptp due to lower performance. > We did not test it. > Ubnt , MT and any other atheros 802.11n based products have aprox > equal max throughput in standard 802.11n mode. > But when airmax and nstreme are ON they have different performance. > Tecnically Airmax is polling ( round robin algorithm ) like nstreme or > turbocell/ outdoor router Proxim . Ubnt polling uses latest Atheros > chipset clock timing ( ubnt calls it "tdma" ) , that may be usefull only > in ptmp . > Nstreme 2 is Nstreme 1, that also uses clock of atheros chipset > So both airmax and Nstreme 1,2 can not increase max throughput in ptp in > comparison with standard 802.11n (hardware atheros aggregation On, 2 chains) > in ptp in ideal conditions( no interference). > Nstreme 1,2 is able to improve 802.11n link in comparison with standard > 802.11n ( Nstreme Off) mode in presence of interference or/and multipath > fading due to it's feature of link parameter adaptation according packets > losses rate. > I do not know is Airmax support link adaptation ( modulation ) or > not . I suppose -not yet. > > > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the > simular > > platform ( Alix, CM9) . It also has poor throughput at small packet > size > > ( but much better then standard 802,11a) and it is may be improved by > > using more powerfull h/w. > >> Ive read up on Sams work and have been very impressed. I looked at the BSD >> TDMA >> driver back in late 08 or early 09 (been a while), when I did it was >> missing glue and needed >> a bit of polish. It looked like adding GPS sync to it would have only >> been a matter of getting >> 2 or more AP's to hold then start on the same signal and keep the >> timing windows synced. >> > > Software TDMA Linux/freebsd implementation based on 802.11 chipset > hardware is separate issue . I think it may be useful in ptp and our test > showed promising results. With regards to ptmp IMHO it is not viable. There > is standard fixed TDMA BWA techhology called fixed wimax 802.16-2004/2009. > There is 802.16-2004 miniPCI cards - ASIC hardware TDMA implementation . > There is TDMA 802.16-2004 BS/CPE Linux based software. For what a lot of > people want full software TDMA implemenation? > Vyacheslav Vasilyev > UNIDATA > Fixed BWA solution > > > > -------------------------------------------------------------------------------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > > -------------------------------------------------------------------------------- > > WISPA Wireless List: wireless@wispa.org > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ > > > > > > -------------------------------------------------------------------------------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > > -------------------------------------------------------------------------------- > > WISPA Wireless List: wireless@wispa.org > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ >
-------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
