Effective Range/Throughput - Comparisons and review - (see postings 31-01-06)
in [Wireless Internet]
|
Prev: WiFi setup assistance needed
Next: new to setup
From: Jeff Liebermann on 17 Feb 2006 20:56 On Sat, 18 Feb 2006 02:23:27 +0100, "navprojects" <nav0785(a)iperbole.bologna.it> wrote: >Confirms some of my ideas on this. >We could build a Spread sheet on this with calculations included. >Great Expectations vs. Unforgiving Reality :-() What I really would like is the PHP source for the various calculations on the page: http://www.terabeam.com/support/calculations/som.php There are others available but I like this one because it's simple. (Warning: I'm a truely horrible programmist and can't even make my own web pile look good and work well). I could probably scribble my own but it's so much easier to just steal one that works. Incidentally, see my assortment of RF path calculators and spreadsheets at: http://802.11junk.com/jeffl/rf-calc/ Please note that one can accurately predict the range, at any connection speed with only: 1. The transition point where the speed changes from one speed to another. 2. The RX sensitivity at various speeds and modulation modes at some fixed reference level (BER = 1*10^5). 3. The desired fade margin. 20dB minimum is good enough for most calcs. I'll post the material attenuation specs when www.thirdbreak.org comes back online. Some work has already been done on comparing hardware. http://www.seattlewireless.net/index.cgi/HardwareComparison The problem is that the Receive sensitivity page link seems to have been "reorganized": http://www.seattlewireless.net/ReceiveSensitivity Apparently there has been some number juggling designed to make someones products look good. No clue how any of the measurements were actually done, or whether they were just plagerized from the manufacturers or chipset data sheets. Got one of these handy? http://www.us.anritsu.com/products/ARO/North/Eng/showProd.aspx?ID=744&cat=1&cat2=3&cat3=38&cat4=0 Much as I like Wiki's, the common vandalism makes me wonder if it's really such a good idea. Back to meetings, work, interruptions, more meetings, etc... Sigh. -- # Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060 # 831-336-2558 jeffl(a)comix.santa-cruz.ca.us # http://802.11junk.com jeffl(a)cruzio.com # http://www.LearnByDestroying.com AE6KS
From: Jeff Liebermann on 17 Feb 2006 21:12 On Sat, 18 Feb 2006 01:56:33 GMT, Jeff Liebermann <jeffl(a)comix.santa-cruz.ca.us> wrote: >The problem is that the Receive sensitivity page link seems to have >been "reorganized": > http://www.seattlewireless.net/ReceiveSensitivity This one works: http://freenetworks.org/moin/index.cgi/ReceiveSensitivity However, it has the same problem. No clue where the numbers came from, whether they are test results or from data sheets, and if measured, what proceedure was used. How to measure receiver sensitivity (one of several ways): http://www.demarctech.com/techsupport/rw-wireless-cards-support/wlan-receiver-test.pdf http://www.demarctech.com/techsupport/rw-wireless-cards-support/100mw-rx-test.htm I doubt if any of those posting numbers to the Wiki bothered to do any of this. -- # Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060 # 831-336-2558 jeffl(a)comix.santa-cruz.ca.us # http://802.11junk.com jeffl(a)cruzio.com # http://www.LearnByDestroying.com AE6KS
From: navprojects on 18 Feb 2006 04:21 Hello Jeff, [Just before leaving for out of office business] Many thanks for this post "Jeff Liebermann" <jeffl(a)comix.santa-cruz.ca.us> ha scritto nel messaggio news:d8ucv1pmtdk3b9v3kurrhgepbmcderb20v(a)4ax.com... > On Sat, 18 Feb 2006 02:23:27 +0100, "navprojects" > <nav0785(a)iperbole.bologna.it> wrote: > > >Confirms some of my ideas on this. > >We could build a Spread sheet on this with calculations included. > >Great Expectations vs. Unforgiving Reality :-() > > What I really would like is the PHP source for the various > calculations on the page: > http://www.terabeam.com/support/calculations/som.php > There are others available but I like this one because it's simple. > (Warning: I'm a truely horrible programmist and can't even make my > own web pile look good and work well). I could probably scribble my > own but it's so much easier to just steal one that works. Which is why I commented on a Open Web Database. I have seen a couple in thepast but they were relevant to HW performance. The Terabeam work is in fact an excellent example. > Incidentally, see my assortment of RF path calculators and > spreadsheets at: http://802.11junk.com/jeffl/rf-calc/ Great, I also had seen a couple of these (Cisco etc), but the others are very appropriate. > Please note that one can accurately predict the range, at any > connection speed with only: > 1. The transition point where the speed changes from one speed to > another. > 2. The RX sensitivity at various speeds and modulation modes at some > fixed reference level (BER = 1*10^5). > 3. The desired fade margin. 20dB minimum is good enough for most > calcs. In fact it's interesting to note that 'expected range' calcs based on specsheet data are hard to come by. It might be nice to devise one. Based on your approach above it should not be difficult. > I'll post the material attenuation specs when www.thirdbreak.org comes > back online. OK > Some work has already been done on comparing hardware. > http://www.seattlewireless.net/index.cgi/HardwareComparison > The problem is that the Receive sensitivity page link seems to have > been "reorganized": > http://www.seattlewireless.net/ReceiveSensitivity > Apparently there has been some number juggling designed to make > someones products look good. No clue how any of the measurements were > actually done, or whether they were just plagerized from the > manufacturers or chipset data sheets. Got one of these handy? I had exemined these 2 well known sites several times but it seems that the data is not as updated as one might hope and the list of suppliers are rather limited. > http://www.us.anritsu.com/products/ARO/North/Eng/showProd.aspx?ID=744&cat=1& cat2=3&cat3=38&cat4=0 > Much as I like Wiki's, the common vandalism makes me wonder if it's > really such a good idea. Do you know how much does the Anritsu costs? A site that might interest you is the following which offers a reasonably sophisicated Survey Tool for an acceptable price. http://www.ekahau.com/?id=4600 > Back to meetings, work, interruptions, more meetings, etc... Sigh. > # Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060 > # 831-336-2558 jeffl(a)comix.santa-cruz.ca.us > # http://802.11junk.com jeffl(a)cruzio.com > # http://www.LearnByDestroying.com AE6KS I should be back in the evening, Thank you and best wishes for a nice day.
From: navprojects on 18 Feb 2006 04:29 "Jeff Liebermann" <jeffl(a)comix.santa-cruz.ca.us> ha scritto nel messaggio news:v40dv1hrombh1bb0qq6afiqnfbc4u4i7k1(a)4ax.com... > On Sat, 18 Feb 2006 01:56:33 GMT, Jeff Liebermann > <jeffl(a)comix.santa-cruz.ca.us> wrote: > > >The problem is that the Receive sensitivity page link seems to have > >been "reorganized": > > http://www.seattlewireless.net/ReceiveSensitivity > > This one works: > http://freenetworks.org/moin/index.cgi/ReceiveSensitivity > However, it has the same problem. No clue where the numbers came > from, whether they are test results or from data sheets, and if > measured, what proceedure was used. I had already viewed this one and it seems quite like the seattle on and lacks updating and other manufacturers. > How to measure receiver sensitivity (one of several ways): > > http://www.demarctech.com/techsupport/rw-wireless-cards-support/wlan-receive r-test.pdf > http://www.demarctech.com/techsupport/rw-wireless-cards-support/100mw-rx-tes t.htm > I doubt if any of those posting numbers to the Wiki bothered to do any > of this. In fact the Denmarctech apparently offers cards etc with remarkable Rx sens and Tx power. Till later > # Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060 > # 831-336-2558 jeffl(a)comix.santa-cruz.ca.us > # http://802.11junk.com jeffl(a)cruzio.com > # http://www.LearnByDestroying.com AE6KS
From: Jeff Liebermann on 18 Feb 2006 14:16
"navprojects" <nav0785(a)iperbole.bologna.it> hath wroth: >> Incidentally, see my assortment of RF path calculators and >> spreadsheets at: http://802.11junk.com/jeffl/rf-calc/ The Sputnik spreadsheet claims to calculate range for their products. http://www.sputnik.com/support/docs/docs_deployment.html I don't know how accurate it is becauase I haven't played with their equipment. >Great, I also had seen a couple of these (Cisco etc), but the others are >very appropriate. I have some issues with the Proxim spreadsheet, which uses 12dB as the target fade margin, and the Cisco spreadsheet, which uses 10dB. 10-12dB is barely functional and not very reliable. I deal with one 5.6GHz link with a 13dB caculated fade margin. (The aesthetics committee would not allow larger dish antennas.) It's down about 10% of the time, which is acceptable because all it does is unload non-time sensitive collected data. >In fact it's interesting to note that 'expected range' calcs based on >specsheet >data are hard to come by. It might be nice to devise one. >Based on your approach above it should not be difficult. Yet another project. However, it's something I find interesting. However, I will NOT concoct a fixed table or chart showing "typical" speeds and ranges. There are far too many variations and I don't wanna spend my life defending the numbers. >> I'll post the material attenuation specs when www.thirdbreak.org comes >> back online. The site is back up. No clue what happened (or why it only happens when I post a link to it). Probabaly a conspiracy. http://www.thirdbreak.org/pipermail/wireless/2005-June/000804.html Note that the material attenuation numbers are all over the place and vary with the source. I once read a report where an instructor had students measure wall, door, and window attenuation. The result was huge variations in attenuations. I was also amused at reading the table of attenuations "accurate" to 3 decimal places. Sigh. I've also had some entertainment dealing with people that claim they have perfectly good coverage going through 4 or 5 walls. It's possible, but more often, the signal arrives via a reflection through open windows that simply bypass the walls. I got involved in such a mystery when a company was doing a site survey and noticed that the signal was useable no matter how many walls they went through. When I showed up, I blocked a few windows with aluminum foil and wall attentuation returned to normal. In another, it was leaking through the HVAC ducts. Still another leaked trough the plenum above a suspended acoustic ceiling. Translation: It's REALLY difficult to get accurate material attenuation numbers as much depends on the layout. >http://www.us.anritsu.com/products/ARO/North/Eng/showProd.aspx?ID=744&cat=1&cat2=3&cat3=38&cat4=0 >Do you know how much does the Anritsu costs? It was just released in Dec 2005 so it hasn't hit the used market yet. List price is about 3 million Yen, which is about $26,000 or 15,000 UK pounds at current exchange rates. Just for fun, look at the prices for Anritsu accessories and tools: http://www.us.anritsu.com/products/ARO/North/Eng/forsaleonline.aspx $25 for a wrench that costs $2 at any hardware store? Agilent 802.11 test equipment and software: | http://www.agilent.com/find/wlan/ N4010A Wireless test set. | www.agilent.com/find/N4017A $13,000 for the basic box. About $6,000 more for the software. Testing Wireless Products: http://cpliterature.product.agilent.com/litweb/pdf/5988-3762EN.pdf >A site that might interest you is the following which offers a reasonably >sophisicated Survey Tool for an acceptable price. >http://www.ekahau.com/?id=4600 Ekahau is $6,000. Also see: http://www.airespace.com/products/appnote_wlan_planning_design.php Airespace is now owned by Cisco. -- Jeff Liebermann jeffl(a)comix.santa-cruz.ca.us 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558 |