Slashdot Mirror
802.11g Slows Down
Moosifer writes "Computerworld reports that in order to step on fewer 802.11b toes, the IEEE has reduced the actual throughput of 802.11g in its latest (and allegedly final) draft. I think I might keep old firmware on my linksys AP and card so that I can at least pretend I have faster gear." It's been moved from 54Mbps all the way down to 10-20Mbps, more than just a slight change.
This is the part where you begin to post the "802.11g is Dying" troll all across slashdot...
Hey, if it's going to be newer, more expensive, with very little increase in speed, what's the point?
Uhh, besides that, I'd be willing to bet most manufacturers will just say "screw it", and give their cards the full speed anyhow, standard be-dammed.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
US Robotics doubles up on 802.11g data rates
I don't think this has much to do with politics. Reading the article, it looks like a backwards-compatibility issue such that 802.11g devices don't interfere with 802.11b devices in a co-existing environment.
the blood has stopped pumping, and he's left to decay
the me that you know is now made up of wires
It almost seems that the new standard will more accurately reflect the real throughput for these devices, especially in mixed 802.11b/g environments. It's better to lower the expectations now before people purchase and are disappointed. I've read plenty of comments at amazon.com from purchasers of 802.11g access points where they were surprised that "backwards compatable" meant that mixing the b/g would make everything run slower.
"We make our world significant by the courage of our questions and by the depth of our answers." Carl Sagan
A lot of people have probably already invested a lot of money in 802.11g equipment because of the 54mbps rate, and now, if they have a mixed environment, will end up with a slower rate than they had with 802.11b (10mbps vs 11mbps). I guess this is the fault of the industry for making promises and shipping equipment before the standards are finalized, but this greatly shrinks the market for 802.11g upgrades.
[dons tin-foil hat] I wonder if the 802.11a proponents *ahem* persuaded the IEEE to do this because they might have lost a lot of invested time/money if 802.11g took over the world... [/tin-foil hat]
Don't upgrade your firmware. You will be fine as long as you're in a 802.11g only environment. The problem comes when 802.11g devices coexist with 802.11b devices. As it is now, your hardware should be fine.
the blood has stopped pumping, and he's left to decay
the me that you know is now made up of wires
Just plug in a cable. While I admit (by the time this posts, like seven other people before me) that knocking it down below one half of the original throughput is weird, 54Mbps is not neccessary. If I need 54Mbps I'll just grab an ethernet cable.
Normal/casual connections need no more than a megabit per second anyway. Browsing, SSH, IM etc does not require a enormous connection. Maybe if there were a "safe mode" there would be both safety for 11b and speed when only 11g is present in the area.
802.11g is not and never has been 54Mps.
The effective throughput of 802.11 is about 22Mps.
54Mps is the effective raw bandwidth.
I have no idea what the new changes will do the speeds of 802.11g, but no one is or has ever gotten 54Mps.
It appears from the comments here that quite a few people haven't bothered to read the article (slashdot users commenting before reading the article? there's a suprise!).
This is NOT a proposal that's going to slow down all the 54Mbps cards out there to 10-20Mbps, all it's saying is "Hey, we were a little optimistic, these g cards have never been 54Mbps, and it would be a little more honest at this point to tell people that they're only 10Mbps-20Mbps cards."
So hold off on your firmware upgrades if you wish, but you still won't have 54Mpbs wireless!
Jolyon
Please read my Canon EOS tech blog at http://www.everyothershot.com
the IEEE didn't make a mistake.. This spec had to be approved by them and the current 802g products on the market aren't IEEE certified. Now that they got arround to cert'ing it they brought down the speed, thats all.
- what is the definition of simultanagnosia?! I've been meaning to look it up!
The headline is more than a bit sensationalist; 802.11g still tops out at 54mbps. It's just that in a network with 802.11b equipment, it'll throttle back to 10-20Mbps. It's not quite as bad as you think it is, but you still may want to look at getting a 3rd party hardware solution. If you've got legacy equipment, you may want to consider picking up an 802.11a hub for your high speed equipment. I always thought that apple was silly by offering just 802.11g when all the chipset vendors have said that they're going to be offering combo solutions. Hell, a combo solution, used properly, can provide speeds of over 100Mbps. Someone's just got to create multilink support, much like the old trick of getting 2+ phone lines for dialup and using multilink PPP.
Marxism is the opiate of dumbasses
Guys,
The 54Mbps is the signal rate of the 802.11g modulation scheme. With the per-packet overhead, the effective data rate is around 20 Mbps, and they're trying to clarify that to consumers.
FastEthernet is 100 Mbps, right? Well, actually, the signal rate on 100Base-TX is 125 Mbps. It takes 5 bits on the cable to carry 4 bits of actual payload data.
802.11b is billed as 11, but it can only do 5. Make sure that you compare apples to apples. Even if the actual throughput is 10 Mbps, it's still double 802.11b's actual throughput.
The rate on the box != the actual throughput you get.
Due to protocol overhead, backwards compatability overhead, physical environment, yada yada yada, you'll see varying throughput.
With current implementations of the draft solution mixed mode performance is *terrible*. 10 Mb/s mixed mode is an improvement. Right now your draft
The standards body hasn't throttled down
Still, by the end of the summer you'll see throughput at 30 Mb/s in pure
In a pure
The compatability
Everybody's going on and on about how it's hardly faster than .11b. Read the freakin article:
.11b is about 5Mbps true throughput. .11g will be twice that in "safe mode" and four times that in pure .11g mode.
"Li estimated that that in mixed 802.11b and 802.11g networks running standard TCP/IP Internet protocols, this will reduce actual throughput to 10Mbit/sec. -- while pure 802.11g networks will have actual data rates of around 20Mbit/sec. Li pointed out that even at these data rates the 802.11g devices still outperform 802.11b devices, which have a raw data rate of 11Mbit/sec. but an actual throughput of about half that speed. "
See that? He's saying
The article would have been much clearer if he had said ".11g is being reduced from 54Mbps raw data rate to X Mbps raw data rate, and from Y Mbps true throughput to 10 or 20Mbps true throughput." Instead he says it's getting reduced from 54Mbps raw data rate to 10 or 20Mbps true throughput. Way to mismatch your units to get the biggest reduction possible.
The post and article compare incompatible metrics, 54Mbps is the theoretical bandwith, vs. 20Mbps measured throughput. The maximum throughput of the draft devices is between 22-24Mbps. The new 10Mbps mode is only when an 802.11b network is detected in the same channel, which is better than the nasty and unpredictable timeslicing that happens with most draft equipment. So... real speed loss = 22-24 to 20. Bad, but not that bad.
-Ryan C.
-Ryan C.
> I work in wireless at a bug company
How are things at Microsoft?
This phenomenon is one of the signs of the apocalypse.
You were 80% angel, 10% demon. The rest was hard to explain. - Over The Rhine
"Math in a song is good."-Linford
TS. It wasn't a ratified standard. Too bad. I mean seriously, if you implement non-standard systems, this is the price you pay. If you didn't point out to your customers that what you were selling them wasn't a ratified standard, then it's your butt in the sling when they complain.
This sig has been temporarily disconnected or is no longer in service
The problem here is just that the reporter seems to be twisting the numbers to try to make it sound worse than it is. His very first sentence compares "true throughput for Internet-type connections of between 10M and 20Mbit/sec" with "54Mbit/sec. raw data rate", which is misleading. Raw data rate and actual throughput are (unfortunately) only vaguely related. If you want accurate numbers for g and b, compare apples to apples. According to the article, if you pay close enough attention, the real numbers are:
Now, maybe in earlier drafts the actual throughput numbers for 802.11g were supposed to be higher, but you wouldn't know it from reading the article. Looking at his past articles it seems like the reporter might just not know the difference, he uses 'throughput', 'data', 'data rate', 'raw data rate', 'data speeds', 'raw data speeds', and 'bandwidth' all interchangeably. The differences between some of those are subtle (or non-existent), but if he's confused enough then comparing 'raw data rate' to 'actual throughput' could conceivably have been an honest mistake...
"So I was downloading this file, and my WiFi card was all like "beep beep beep"... and I was all, like, huh? It devoured my bandwidth. Then I had to finish downloading the file only it was really slow and it wasn't as good. It was, like, a bummer."
In Soviet Rush, today's Tom Sawyer gets high on you.
So if they made some change to the final 802.11g standard such that the througput is only 20 Mbps, that's not much of a change from the draft.
And it has always been the case that in a mixed enviornment (802.11b coexisting with 802.11g), you can't get maximum 802.11g throughput. The exact amount of slowdown will vary.
So in summary, I'm not convinced that anything this Computerworld article is reporting about the 802.11g standard is actually a significant change from the draft. They've just compared some numbers in a meaningless way to sensationalize the story.
Disclaimer: At work I'm involved in the development of 802.11g products.
I can't believe it took this long for someone to pick up on the fact that they were comparing to completely different numbers.
The drop in effective data throughput in pure 802.11g environments is only about 2Mbps (from 22Mbps to 20). It's nothing to sneeze at, but it's hardly the 24Mbps drop that the headline would imply.
Is this even really news?
At work we've been using 802.11a and 802.11g devices (not to mention 802.11b) since the absolute first days they were each available. All the testing I've ever done was far from impressive and probably close to what they are saying in this article:
802.11b
Advertised Speed: 11 megabit or 1.38 megabytes/sec
Advertised Range: 150 feet
Real-world Speed: 4.5 megabit or 0.55 megabytes/sec
Real-world Range: 100-250 feet depending on interference
802.11a
Advertised Speed: 54 megabit or 6.75 megabytes/sec
Advertised Range: 150 feet
Real-world Speed: 21.5 megabit or 2.7 megabytes/sec
Real-world Range: 50-100 feet (outside of that and the link is so weak the real throughput is worse than 802.11b)
802.11g
Advertised Speed: 54 megabit or 6.75 megabytes/sec
Advertised Range: 150 feet
Real-world Speed: 19.5 megabit or 2.45 megabytes/sec
Real-world Range: 100-200 feet (at 200 feet you can still get better than 802.11b throughput, while 802.11a usually is completely gone at 100 feet unless you are in an open field)
The reality is that they had better start advertising the true speeds and problems of 802.11a/g because a lot of people get disappointed when they compare them to standard 100Base-T wired connections -- to me it's flat-out false advertising. The real-world range of 802.11g is similar to 802.11b and its real-world throughput is consistently 3-5 times faster than 802.11b.
But to say that 802.11a/g are "54 megabit" so people compare them to a 100 megabit ethernet connection is REALLY wrong. It reminds me of the "56k" modems we have in our computers that never connect faster than 40k-45k for most people.
(for the record, our wired 100Base-T network that all these devices are plugged into is very fast -- we have no problem getting 8 to 11.5 megabytes-per-second of throughput)