Slashdot Mirror
When 54 Mbps isn't 54 Mbps: 802.11g's Real Speed
eggboard writes "Matthew Gast, author of 802.11 Wireless Networks, filed this article for O'Reilly Networks explaining exactly how fast 802.11g really is: that is, what's the actual data payload and real throughput, not the rated maximum speed. His conclusion? In mixed 802.11b/g networks, which will be common for years to come, g is only 1.6 to 2.4 times faster than b, not 5 times faster as it is in its g-only mode. This article has real math based on the specs, rather than armchair speculation."
When you connect a 10bT NIC to a 100bT switch you get reduced throughput.
EVERY medium that I've seen specs for published the actual bit rate of the wire/cable/fiber, not the end user throughput. They can't know that because they don't know what protocols you will be running over the network.
Article X: The powers not delegated... by the Constitution...are reserved...to the people
This doesn't sound much better than armchair speculation either... Where are the real-world throughput benchmarks performed with actual equipment?
Gigabit ethernet is supposed to be 100 times faster than good ol' 10BaseT. It is, at the root layer. Most devices can't push that much data through the pipe, and with wireless, there is MUCH more error correction that needs to be done in communicating back and forth. Wired networks (normally) don't have the kind of interference that 2.4 GHz-band devices now suffer from.
100 Megabit Network does not actually deliver 100 Megabit transfer speeds. Film at 11.
Okay, I read the article, and here's a basic rundown (I think :):
:)
* 802.11g in a homogenous network (ie: only 802.11g access points) is faster than 802.11b (by a factor of five or so) *and* 802.11a (just a bit faster)
* 802.11g in a heterogenous network (ie: some 802.11g access points, and some 802.11a access points _which have been "assosiated" with the 802.11g_) is rougly 1.5 to 2.5 times faster than 802.11b, depending on the type of collision-detection algorithm used.
So, to sum up the summary: If you start replacing your 802.11b access points with 802.11g access points, you'll see some performance gain with 802.11g client devices right away. When all your 802.11b client devices are gone (and thus all the 802.11b access points), it'll be *way* faster. Faster even than 802.11a.
Why is this billed as a bad thing? You get compatibility with your existing infrastructure, a little bonus performance now, and when the time comes, *bang* you get a big boost.
This is the kind of thing that sysadmins such as myself LOVE
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Even the manufacturers make this point. From apple's site:
If a user with an AirPort-enabled computer or a Wi-Fi certified 802.11b product joins an AirPort Extreme wireless network, that user will get up to 11 Mbps and the AirPort Extreme users on the same wireless network will get less than 54 Mbps. To achieve maximum speed of 54 Mbps the wireless network may only have AirPort Extreme-enabled computers on it.
Its not like this was quite the surprise its being made out to be...
You're special forces then? That's great! I just love your olympics!
(Sorry for the parent post, I made a typo. Just s/802.11a/802.11b/ in the second bullet point. "oops" :)
Okay, I read the article, and here's a basic rundown (I think :):
So, to sum up the summary: If you start replacing your 802.11b access points with 802.11g access points, you'll see some performance gain with 802.11g client devices right away. When all your 802.11b client devices are gone (and thus all the 802.11b access points), it'll be way faster. Faster even than 802.11a.
Why is this billed as a bad thing? You get compatibility with your existing infrastructure, a little bonus performance now, and when the time comes, bang you get a big boost.
This is the kind of thing that sysadmins such as myself LOVE :)
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Wireless networks have greater latencies than wired networks. Its just a fact. Windows NT (and various linux/bsd/other systems) is usually nice enough to automatically adjust the TCP recieve window size to your network latency. Sometimes it gets it right. Other times it gets it wrong.
For this to be a usefull test, you will need to at least publish what the window size was on each end. Also, making sure the immediate area was free of microwaves and blenders helps a bit.
Now, I fully believe that the test was accurate and that the speeds listed are a accurate representation of what an average 802.11g network will experience. But there are many things you can do to tweak your throughput much closer to the theoretical speed.
Karma: SELECT `karma` FROM `users` WHERE `userid`=138474;
Oh, so I only get a 60% faster connection? Given that soon enough the price differential between B & G will be gone, I still think G is the superior choice. When the wireless cards are only $15 to $20, I think that pure G networks will be much more common. And then you will get much higher throughputs.
Maybe they should go after Dannon yogurt for decreasing the size of their container to 6oz from 8oz, but keeping the price constant. Then at least they would be reporting on something I could care about.
No, I don't trust in god. He'll have to pay up front, like everybody else.
It should be a good thing for SoIP, and for pissing off the RIAA
MoFscker
However, when you connect a 10baseT NIC to a 100baseT switch, you don't slow down the rest of the connections to the switch, which can still operate at 100baseT. The situation with wireless, a shared medium, is more analogous to connecting a 10 baseT NIC to a 100/10 baseT auto sensing hub--when you hook up that 10baseT card, it slows down the rest of the hub to 10 baseT.
5.) It's still too slow to download Celeste-Virtual_BJ.avi in a reasonable time .GIF icons.
4.) You're not a cafe communist with a computer and a four dollar cup of coffee.
3.) The low-bandwidth version of Slashdot doesn't have those cool 1997
2.) The babes dig retro shit these days, like 14.4bps dial-up.
1.) Your life revolves around physical things, not six-hundred dollar mp3 players (iPaqs, etc.)
Reply or e-mail; don't vaguely moderate. Ex-O'Reilly/MIT employee, now a full-time Google employee.
Thanks to MADWIFI and this postI was able to get my Netgear WAG 511 working in a laptop in under five minutes. A walk in the park compared to the last time I configured wireless on my laptop.
I have not had a chance to thoroughly test it in a multi-signal environment, but the throughput is solid on B. There have been some drop-outs but I blame the D-Link access point to which I am connecting. (DWL-1000AP=junk, but at least it was inexpensive).
The WAG511 was on sale at Fry's for $80; I haven't seen it significantly cheaper on line, so I grabbed two.
This afternoon I am working on getting another card to work in a desktop with a pcmcia adapter to act as a host so I can unload the D-Link; then the higher-speed testing can begin. I have nothing but good things to say about the Netgear card so far. Thanks to all those who are doing the heavy lifting to make A/G support possible.
The actual improvement in g-only mode is better than what the specs say.
Anyways, I don't know why anyone would have a mixed b/g network, unless they are offering it as public service. Its easy enough to upgrade everything to g-mode only. 802.11g sounds like a big win to me.
"The lesson to be learned is not to take the comments on slashdot too literally." --Vinnie Falco, BearShare
Definitely not 54mbps, my pr0n goes more than twice as fast on my 100mpbs connection at school.
ok, so on a straight g system, you get 5 times the rate of b wireless... b gets ~11Mper second times 5 = 55... a nice approximate number to 54... where is the problem? Why is this a controversy worth discussing?
Laws are for people with no friends.
There's two problems I see with the authors math. First it doesn't measure real real-world conditions. Don't get me wrong, office connectivity is valid. But where's the growth? It's in ISP implementations. Big cities are specifically making use of it. Needless to say there are a plethora of different sized buildings in those areas.
Second, algorithms are an important part of CS, but geez, I have yet to see where fluid conditions have been calculated with necessary precision with just a monolithic algorithm. These measures should have been "componetized" and solved with more refined equations.
what in the fud does that mean? /. might go off?
there's no way the planet/population is in jeopardy, as just last weak on tv, (between the boobies & the allergy medicine commershills) we were tolled 'things' were looking up/would be fine.
we'll never be billyonerrors if we don't pay the badtoll to the georgewellian fuddites, corewrecked, again? what a relief.
I'm just waiting for market to see this and realise that you can, totally theoertically on a full duplex connection double throughput. So 100Mb/s give 200MB/s so I would guess we will be seening "108Mb"
very soon
Rus
Cheap UK and US VPS
When you mix networks, you always get lower speed. What's new here? That it's wireless?
Next thing you'll tell me that each new version of MS Word doesn't increase my productivity. Yeah, just blow me away...
The only thing we learn from history is that we learn nothing from history.
"Furthermore, the model ignores the sophistication in the TCP acknowledgement model. To avoid constraining throughput, TCP uses "sliding windows" and allows multiple outstanding frames to be transmitted before acknowledgement. In practice, TCP acknowledgements can apply to multiple segments, so this model overstates the impact of higher-layer protocol acknowledgements."
This reduces the "TCP" he uses to a stop-and-wait protocol.
You didn't even look at the article, did you? There was no testing. The author didn't model TCP windowing at all, and he even failed to take delayed ACKs into account.
You'll save yourself some grief if you get yourself a wireless card. ;)
I got myself one too. No regrets.
It's easy enough to upgrade everything to g-mode only.
Like iBooks? Like PDAs? Like wireless security cameras? There's more than laptops with PCMCIA wireless cards in the world.
I have no complaints about the speed of my neighbor's wifi access point.
Strange women lying in ponds distributing swords is no basis for a system of government.
Now we have the a/b/g and maybe the x standard, and possibly more to come, them all being close in speed and performance to each other. Why arent they all rolled into one single spec which accomodates 5.4GHz, 2.6Ghz 11Ghz and more rather than making seperate specs and causing trouble to the manufacturers, users and buyers?
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
b, g, ??? Why? I've been using 802.11a in my school wireless laptops.... Granted I don't need distance associated with b or g, because the laptops are in the same room as the AP. Speedwise, it is much better, and I don't have to worry about blenders, microwaves, cordless phones, and other stuff in the 2.4ghz range.
Distance is the only thing that b has to offer, but if your within your building / room you need it, then it is fine... Unless you want to have to PAY to access the wireless at some street corner, be even a Verizon link that I've seen pictures of was in a big Verizon "booth" type device that you would know when your around it... I'd pay for bandwidth, not coverage... that is the providors problem.
but, just my $.02 worth....
--- Relax, that mass muderer is just trying to reduce our carbon footprint, one fetus at a time...
54 Mbps has never been the advertised real bandwidth for g. 54 Mbps is the speed at which data goes between your card and your router. Guess what? There's a lot of correction code, synchronisation, etc.
Maybe the author should read the docs(RFCs aren't that ahrd to find, are they?) before jumping on a juicy story?
Oh, and... DUPE! "lie" was already covered a few months ago. Heck, there even was the same conclusion: g gives you around 20 Mbps, VS what, 11 Mbps max on b?
Try Corewar @ www.koth.org - rec.games.corewar
Holy ballz son. You've discovered a new formula!
1) Post Insightfully with format errors
2) Admit to mistake and repost with corrections
3) Go from Insightful to Informative and reap in double the karma
Niiice.
Here are some real numbers.
Best Performance among various hardware
802.11g
wep off: 15.5Mbps
b card on network/wep off: 9.4Mbps
wep on: 10.3Mbps
802.11b
wep off: 4.8Mbps
Whenever the offence inspires less horror than the punishment, the rigour of penal law is obliged to give way...
"his article has real math based on the specs"
Kinda like judging a car's performance based on "real math based on the specs" when you can actually test the real thing in the Real World.
Whenever the offence inspires less horror than the punishment, the rigour of penal law is obliged to give way...
lol, you know I realised that after I'd posted the second one only to see that the first one had been modded way up in the meantime? :)
But in all seriousness, it doesn't matter - I had 50 the first (and incidentally, only) time I checked my karma, and that was after the cap was put in place (very shortly after, in fact). God knows what it was before.
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
... nobody has managed to blame this on Microsoft.
Why do people keep insinuating that comfortable furniture is somehow incompatible with brilliant thought?
Seriously, I've come up with many a clever solution upon taking pencil and paper to bed with me.
Gentlemen! You can't fight in here, this is the War Room!
Unfortunately, many retailers no longer stock any 802.11a equipment, other than a couple of "universal" a/b/g cards.
I was in Best Buy and CompUSA and it is wall-2-wall 801.11g -- all "54 MBps!" in big, bold print.
It is a shame, since the 5 GHz band is so less crowded. I think "A" equipment is going to fade into a niche and be harder and harder to find.
Learning HOW to think is more important than learning WHAT to think.
And I thought this 56K modem in my stack of cards in my closet really put out 56K! ;)
30% off web hosting. Coupon code "SLASHDOT".
when you hook up that 10baseT card, it slows down the rest of the hub to 10 baseT.
Not at all. An auto-sensing hub (does anyone still make these?) is actually a 10mbps segment bridged to a 100mbps segment. Each port connects to whichever segment it can talk to, and they're switched together internally. The whole thing does *not* drop to 10mbps when any 10mbps devices are present.
It would be nice if B and G played that nicely in the same spectrum, but they don't.
I think that not taking them alive was an unfortunate, but reasonable, alternative.
I used to be a cop, and did SWAT for about seven years... an assault on a fortified target like that is difficult in the best of circumstances, let alone in the midst of a hostile city, where you may or may not be able to guard your flanks. If that situation had turned into a prolonged siege, the brothers might have had the opportunity to contact local resistance elements, get some media attention, and shift the balance of that situation in their own favor, costing american lives. We already had multiple americans wounded in several attempts to make entry... how many more would have satisfied the critics?
An assault on a fortified target like that is very, very difficult, particularly when the occupants are armed to the teeth and unwilling to be taken alive. With a single stairway as a choke point, you'd have to attempt to breach elsewhere to gain a tactical advantage... nobody is going to want to advance into that stairway's "fatal funnel."
You would have to try to breach the ceiling (very hollywood, and not very practical in the real world), or breach one of the windows (thick, fortified glass). You could try to make entry from multiple points, or simply gunport those additional entry points... but you are talking about a prolonged, complex SWAT operation, something the military may not necessarily be set up to do, especially in that environment. You could try gas, but that doesn't always work. I've been on ops where we gassed the hell out of people, and they shook it off, even without gas masks. You could try the Russian "fentanyl" gas... might kill them anyway, and they'd be just as dead as if they were shot...
This operation was 101st, as I recall, and they are reasonably high speed. Don't get me wrong... the Delta/SEAL operators are the best, and their CQB skills are top-notch; I've trained with some of them, and I was impressed... but their mission and mindset are a bit different from a civilian SWAT team. They are soldiers in a war, not police officers, and their response in a hostile environment may not be optimal in a perfect world, but is certainly objectively reasonable considering the circumstances.
It would have cost lives, and valuable time to attempt to inject civilian SWAT tactics into that environment... I can certainly understand why they chose to do it the way they did.
They at least made an effort to take them alive... if I were that local commander, I wouldn't have squandered the lives of my men on two scumbags like Uday and Qusay either.
A trial and some iraqi justice would have been nice, but even so, they got what they deserved.
Even if a man chops off your hand with a sword, you still have two nice, sharp bones to stick in his eyes.
I stand corrected. But were there not more simplistic hubs that would drop to 10 mbps for all segments?
isn't this already known. I knew if you put a b device on a g network it slows down other g devices since g came out. I thought this was well know. Is this article important because no one has actually done the numbers before?
A properly-equipped laptop with PCMCIA can do anything an "iBook" or "PDA" or whatever new-fangled device can do.
"The lesson to be learned is not to take the comments on slashdot too literally." --Vinnie Falco, BearShare
I have an auto-sensing 10/100 hub (not switch) and it does indeed maintain 100 Mbps among the 100 Mbps nodes, in spite of a 10 Mbps node also being plugged in. What I haven't tested is how much actual bandwidth is available to the 100 Mbps nodes while the 10 Mbps node is going full blast. My SWAG is that there's still about 90 Mbps left.
Yes, there were. However, they simply ceased selling when smart sensing became available due to their catastrophic effect on bandwidth and the relative inexpense of adding smart sensing.
I think they sold maybe a hundred of those worldwide, total.
Smart sensing came fast when customers realized their investments in 100mbit became worthless as soon as they plugged in one legacy device.
I've heard that CompactFlash interface can't support those speeds anyway. Is that true?
When it's hi-speed 802.11g!
"This article has real math based on the specs, rather than armchair speculation."
How, exactly, is sitting around doing math not "armchair speculation?"
Property is theft.
So you mean my g network card would be highspeed but not fullspeed??!?!!?
... that the 54mbits number measured how many bits fly through the air, not how many bits of the data you want carried from one end of the other. If it takes half the bits to guarantee delivery, then you still have a 54mbit connection, but only 27 of that is the data that you actually see.
Maybe I'm just used to marketing-ese. I remember when video game cartridges were measured in bits and not bytes. I remember being stunned that the Sega CD could store 4.7 gigs of data. Too bad I had to divide that number by 8.
Come to think of it, floppies were like that. "2 megs unformatted!"
Marketing really sucks for computer geeks. We want hard data, they want to give us the highest (or lowest) numbers. Go fig. This particular industry would do much better to appeal to practical #'s and develop trust based on that.
"Derp de derp."
yes, if you equip it with a hack saw and a roll of duct tape so I can put it in my pocket and reassemble it afterwards.
I've had this sig for three days.
The "airport" slot in my iBook is really an internal PCMICA slot. G is definitely possible on this sucka.
Good luck; you'll need it.
I don't go around telling people I'm 8'11" just because it's my theoretical maximum!
And you figured out a new formula!
A) Wait for someone to repost.
B) Correct and poke fun at him.
C) Reap karma!
A NYC lawyer blogs. http://www.chuangblog.com/
This allows segmenting the slower clients out of the higher speed network, and increases your overall bandwidth. Older clients use the (somewhat crowded) 2.5 GHz spectrum, high demand clients use the 5GHz bandwidth. Neither interfere with each other, and in fact both can transmit at the same time. You also gain more available channels on the 5GHz spectrum, which makes deployment easier.
I've set up a lot of these in people's homes and I'm at the point where I'm practically begging them to get an electrician to run Cat5 behind the walls. Why? Because 2.4 ghz phones interfere badly with them, and the ranges are nowhere nearly as good as what the manufacturers claim to be, and they just keep calling me back whenever their connectivity cuts out.
The problem with a is the high frequency limits its range. 2.4GHz is already fairly short range, 5GHz is even shorter. Can't you just use different channels on the 802.11b and 802.11g subnets?
"The lesson to be learned is not to take the comments on slashdot too literally." --Vinnie Falco, BearShare
Certainly makes an arguement for forgoing a cutover, but rather keeping two seperate networks. I will keep my current 802.11b network seperate then, to increase speeds for my actual PCs on my 802.11g segment. Why maintain a 802.11b network? Because I'm not going to replace 802.11b wireless usb adapters in machines that are only connecting to the internet and nothing else, such as my TiVo and Playstation 2. But at the same time I want as close to 54mbps as possible when I'm grabbing files off the laptop in the living room.
OpenBSD leader Theo states that there are 7000 users of OpenBSD. How many users of BSD are there? Let's see. The number of OpenBSD versus NetBSD posts on Usenet is roughly in ratio of 5 to 1. Therefore there are about 7000/5 = 1400 NetBSD users. BSD/OS posts on Usenet are about half of the volume of NetBSD posts. Therefore there are about 700 users of BSD/OS. A recent article put FreeBSD at about 80 percent of the *BSD market. Therefore there are (7000+1400+700)*4 = 36400 FreeBSD users. This is consistent with the number of FreeBSD Usenet posts.
Due to the troubles of Walnut Creek, abysmal sales and so on, FreeBSD went out of business and was taken over by BSDI who sell another troubled OS. Now BSDI is also dead, its corpse turned over to yet another charnel house.
All major marketing surveys show that *BSD has steadily declined in market share. *BSD is very sick and its long term survival prospects are very dim. If *BSD is to survive at all it will be among OS hobbyist dilettante dabblers. In truth, for all practical purposes *BSD is already dead. It is a dead man walking.
Fact: *BSD is dying
Not at all. An auto-sensing hub (does anyone still make these?) is actually a 10mbps segment bridged to a 100mbps segment.
No. Some hubs, long time ago, had the feature of dropping speed whenever there was a 10 Mb connecting. This is long ago, but there was....
Assembling etherkillers for fun an profit
So if you have a homogeneous 802.11g network, then some war-driving schmuck pulls up to the curb in front of your house with an 802.11a card, does your network suddenly get slower?
I'm no signaling expert, but couldn't the loss in range be compensated for by a better antenna?
test
<a href="http://www.joblessjimmy.com">Work is dumb and so is Jobless Jimmy.</a>
The airport slot on my G4/533 Digital Audio is definitely the same connector type as PCMCIA, but it's definitely not PCMCIA compatible.