Slashdot Mirror
Is There Too Much Enthusiasm Over Wireless?
lukOh asks: "in the US, 802.11b/g (2.4Ghz) devices use an
83Mhz-wide frequency range; in-use channels spanning 22Mhz and centered on one of 11 5Mhz steps (badly named as "channels"). This means there should be no more that 3 networks in close proximity, 5 'channels' far from each other, to avoid harmful interference. Now, in the middle of the mixed area where I am, the number of usable WLANs (SNR>20dBm) has gone from 10 to an unworkable number of 20, in just one month. Has the community/the market overestimated the practicality of wireless networks? Are we generally relying too much on such a young, IMHO immature technology made on 'startups hope' and broken firmwares? How can this mess possibly be handled in a working environment, especially the moment your boss asks you to give him access to 'the wireless'?"
"Access points can be easily detected, but the same isn't always true for every single client (or Bluetooth device) searching or using a network. Bluetooth itself employs the same 2.4Ghz range with 1Mhz-wide channels and much less power. To avoid interference a device jumps channel-to-channel, when the currently selected one is busy.
Most WLANs are managed by less-than-perfect SOHO access points. Connecting to an AP in such an environment is a gamble (even from 1ft away), especially when: WPA/WPA2 must be used; 802.11g stability is a dream; anywhere up to 7 networks are on the same 'channel' (1 and 11, being the most used, are standard on many devices); and now 'channel wars' are very common (i.e. 2 or more users concurrently hunting to set a free channel for their network, making the entire range unusable for hours)."
Most WLANs are managed by less-than-perfect SOHO access points. Connecting to an AP in such an environment is a gamble (even from 1ft away), especially when: WPA/WPA2 must be used; 802.11g stability is a dream; anywhere up to 7 networks are on the same 'channel' (1 and 11, being the most used, are standard on many devices); and now 'channel wars' are very common (i.e. 2 or more users concurrently hunting to set a free channel for their network, making the entire range unusable for hours)."
No one else is using it.
everything will be wireless one day
.kyle
"Are we generally relying too much on such a young, IMHO immature technology made on 'startups hope' and broken firmwares? "
We're relying too much on an unregulated spectrum.
The extention of the 802.11b standard into 802.11g is a pain the arse for exactly this reason. All access-points should be limited to work on only channels 1,6 or 11, and rate limited so that anything too far away simply drops off, rather than throttling.
802.11a has a much better frequency spacing (8 non-overlapping channels in most juridstiction, 4 in the others), but many countries won't let you use it outside. The penetrating power at 5GHz is also less than at 2.4GHz.
Has wireless been overhyped? Hell yeah, but all we are seeing is the same problem that we all had when everybody went out and bought a 900Mhz cordless phone.
We need to either compress the channel bandwidth (OFDM with few channels around the center frequency), which would give less bandwidth per channel, extend the number of non-overlapping channels available. Jacking up the frequency would give better overall throughput and less channel conflict at the cost of range.
Norman Cook's Ode to Sl
Wireless technology is great.... when you absolutely need it. Take the Conference Room scenario, whats wrong with a high port density switch under the table, accessible via a central panel? You end up with a 'spiderweb' of Cat5 cables, but with wireless, you still end up with all the power cables.
Yes, its useful to avoid snaking a cable from your desk to your bed in your dorm room, but is it a necessity?
Or have consumers bought into the "I need my data everywhere" ideal promised by the wireless people (Centrino! Get it, you'll be a hipster Blue Man Group Guy) and the constant bombardment of high speed wireless access ads from the phone company (Verizon)?
Back in my day, we had vt100 and 9600 baud, and we ran long serial cables or keyboard extension cables if you needed to be able to compute while wandering around your dorm room or a lab. How much real progress has been made with the WWW and 802.11 ?
I want to delete my account but Slashdot doesn't allow it.
I have to troubleshoot a network timeout problem that doesn't happen in wired locations I support, but the wireless one times out when a certain application isn't used for about 10 minutes.
And the wireless printer there suddenly decided forget how to get an IP address from the wireless router.
It's not a happy time in Wirelessville.
Saskboy's blog is good. 9 out of 10 dentists agree.
If your device uses batteries then wireless makes more sense. If it's something that you don't ever want to replace batteries for, then you have to plug it in. At that point, you may as well plug it into the non-wireless network.
You sir, are weak. Being able to untangle that mass of cableage behind one's desk is a staple of geekhood, gross wire length (like number of LEDs in one's living space) is a status symbol. You fail.
Wireless is a tool and can be great when applied appropriately. It is not the answer to everything as some would like to think.
I work for a college and once or twice a year someone brings forward the idea of a mobile cart of laptops for a roaming classroom. All laptops using wireless networking.
It sounds great until you find out they want to 30 students doing graphics or medical imaging at the same time. Of course we mention that it may not perform up to their expectations and that they should do some testing. They never follow through with the testing.
I'll say it again. Wireless is a tool and can be great when applied appropriately.
Keep the Classic Slashdot.
One Solution that /nerds seem to find suitable is to login to whatever access points are chocking your channels and change their settings.
I know, its not exactly ethical, it is legally dubious, etc.
But, since most people rarely change the default admin password, you can't argue with the results from switching people off your channel, or reducing the broadcast power of their WAP.
Mostly though, the issue is that WAPs aren't 'intelligent.' They aren't spread-spectrum, they can't automaticall channel hop because they can't predict how good/bad your reception will be... There's a whole host of technical challenges to making them play nicely together.
Here's a super nerdy pdf with equations, pics representing signal intensity/overlap. If it doesn't answer your questions in a highly technical matter, I don't know what will.
[Fuck Beta]
o0t!
I imagine we'll wind up with some sort of consolidation. We'll have something similar to phone companies with wireless. A regulated monopoly actually seems the most efficient here. I'd say the best example is radio stations. They are regulated so that they don't interfere with each other. We'll probably have wireless "markets" where different companies work. And at least here two or three could co-exist in some markets. And since competition among wireless providers means an area is more attractive to businesses, then gov't will maybe subsidize tower-building?
And for a good reason, it is so cool to be able to work from anywhere. When I was in college. We had alot of tables with network connections. But with so much student they were always taken. We could only work at places that had no plugs.
So the hotspot were my savior. Now when I work in my office, I don't need to mess with cable anymore(beside AC or mouse). I would never get back to the "old" way now.
One day the wireless turned out not to work very well. We didn't understand it at first. Cause it was working so well for so long. We discovered that 2 other hotspot has been install near our own, so they were 4 wifi network near each other.
They was somekind of interference. We changed channel to an unused one instead of using our default channel 6. It helped, but the article is right. They should have tough of the hype. They could prevented this. Now it will get worst. And the only solution is upgrade to a newer better standard...
But for now, we moved to a new business adress, and we are the only one using wifi. Joy to my world.
It's funny how I make sense to others and not myself...
Others have described this as "The Tragedy of the Commons." Have a stretch of pasture where anyone can graze their cattle for free, and it'll soon be overgrazed. Have a stretch of the spectrum that anyone can freely use, and it'll become overused, so much so that no one gets any benefit.
I saw that in a town I visited where the water was unmetered. A local told me that at first it seemed a good idea. Water was so cheap and abundant, why go to the cost of metering and billing by usage? But unmetered led to waste and waste led to a search for new sources that turned out to be expensive. The result was that everyone, whether they wasted or not, had to paid sky-high water bills.
I hate to sound like a scold, but we need to make like good little hobbits and not trash our Technological Shire. We are going to have to discipline ourselves not to waste what's free. If wired can do the job with a trifling more effort (and probably less cost), we need use wire. Reserve wireless where it's necessary or particularly handy.
--Mike Perry, Untangling Tolkien
2. 802.11a does not use the same frequencies as b and g, and has more spectrum. Unfortunately, the wireless manufacturers aren't promoting it, but even if the complaint above - that there are only three distinct bands and therefore you can't have more than three networks in one place - was actually true, it is an actual solution.
3. The complaint noted by the article is false. While it is true that there are only three distinct, non-overlapping, slices of spectrum allocated for 802.11b and g, you can have more than one network using the same slice of spectrum, at the cost of efficiency and speed. It is not the case that having a network on, say, bands 1-4 suddenly means that no other networks can use that spectrum, either theoretically or practically. The more networks run on those bands, the poorer performance will be, that's all. In reality, the chances of the performance decrease being so bad that it actually makes more of a difference than your DSL's bandwidth is relatively low, especially in the US where 1.5Mbps is considered a really good connection.
So, to recap. Wireless works. It could be more efficient, but as sold, currently, it's more than up to the task. The proof of the pudding is that people are actually going out there and setting up their own wireless networks, and keeping them. We're not facing any real problems yet.
You are not alone. This is not normal. None of this is normal.
Just wait until wireless power is available...
Wireless power is available today. The problem is capturing it and converting it. This is something that plants have gotten very good at over the past several hundred million years.
All the appliances in your home are battery powered? I bet she puts up with cords for a variety of things, even the television. Of course, most of those are hidden down the edge of the walls and behind the furniture, etc so I see your (hers) point but only somewhat.
;)
It's not the wires that are a problem, it's the builders philosphy of where to put outlets and what to have "outletted".
The solution is just modern home design with some better plug-age in the normal areas. The problem is, it is only in the last few years that ethernet and coax cabling to every room has become sort of common in new residential construction. Most places it is somewhat doable as a retrofit, others it is not or exceedingly difficult. Another problem is the aversion people have to floor as opposed to wall outlets, because it forces a somewhat fixed furniture location. A dictatorship like an office can get away with it, but most homeowners wouldn't want it unless it was camoflauged somewhat or they could be content with their initial furniture layout and add in the ethernet/cable outlets where they fit the best. Yes, work, but maybe worthwhile work.
And besides, it's always nice to have an excuse to get out the sawzall!
Wireless will only dominate where data security is not a real issue. Wireless is still too insecure and will always be less secure than wired networks, regardless of the encryption protocols used.
This does not work.
WiFi works on unlicensed spectrum, companies cannot claim it as their own... that spectrum is for all to use however they please within the limits set by the relevant agencies. Trying to bury the new guy by boosting your transmitter's output would most likely violate the peak radiation limit and get your transmitter shut down if not brought back within compliance.
What would be really useful is moving WiFi towards true spread-spectrum modulation to reduce performance degradation from multiple full and partial overlaps. The main issues with this would be extra complexity, longer channel scanning/sync times and yet more bandwidth (but at a lower mW/MHz density) per useful channel.
After a year of wireless I have just finished moving back to a wired home network. In retrospect it seems like a loony idea: why replace a reliable wired network with a whole bunch of expensive equipment that provided less performance with far less reliability?
Using your laptop on the couch or on the deck has great novelty value, but is useless from a work or ergonomic perspective. Throw in interference, inevitable drop outs, and the fact that real world performance is no where near the '54Mbs' marked on the box and it all adds up to an unappealing package in the home.
if it weren't for a pair of legitimate concerns that all the home users in a neighbourhood have like:
- ISPs putting ridiculously low monthly bandwidth limits on users - like mine, 10Gb/month down... If I download one Mandriva install DVD, that just about uses it up! Add in that Finnish Star Trek spoof, and I'll probably be getting a phone call saying I'm downloading too much.
- ease of setting up a network where your LAN is private, but the wireless router will allow other LANs to be set up... then people could reasonably share a single access point.
When I set up my network 2 years ago, there were 2 other detectable wireless networks. Now there's 10; 1 is open.
If I could be made to feel that my home network is reasonably secure (with wireless), and I didn't have to worry about download limits, then mine would be a public access point.
When I park somewhere to borrow Internet access, more often than not, the majority of WAPs are Linksys on channel 6. If it's a spot I might use again, sometimes I'll log in (l:admin p:password) and sort out the mess, putting some of the APs on 1 and 11.
I believe WAP manufacturers (the big three especially) have a responsibility to at least default each unit to a random channel (1, 6, or 11). Even better, have the WAP scan for and use the least cluttered channel on power-up.
Thank you, Edward Snowden.
"Arguments from authority are worthless." —Carl Sagan
From what I hear, Intel is planning to introduce WiMAX any day now. One of the big changes is that it will use licensed spectrum in some modes. Hopefully mortal people will be able to purchase a license. If that's the case we'll be able to have some recourse if someone causes our carrier to noise ratio (C/N) to drop to an unacceptable level. 2.4GHz is mostly unlicensed space, so there's not much you can do about it.
"Well, good luck finding a judge that doesn't run a bestiality site."
Slashdot has been filled to the top with slobbering wireless fanboys for years and years. This is the very first article I've seen where the poster isn't gushing all over the ISM band and how they'll put a brazillian bits/second through it from over the horizon.
I did see quite a few theoretical posts - ie there are three channels, a good engineer will use three 120 degree sectors. That is better, but they go on to say the next ISP that comes along is SOL. Not the case - they just elbow there way in, and people keep loving up the ISM band until it turns into packet bukkake - 100% utilization, 0% throughput.
Anyone who seriously wants to deploy that stuff should go google for "n9zia wireless" and read the Green Bay packet crazies ideas, which is where I learned half of what I know. The other half came from hard experience.
There will, of course, be two dozen fanboys all set to reply to this. You need to ask yourself the following questions:
Ever climb a tower?
Ever made a 21.7 mile shot using 802.11b?
Ever operated a wireless ISP in a metro area?
Ever been invited to speak at WispCon?
If you're not qualified, please shoot your mouth off on some other topic. Really. This article is a step in the right direction for Slashdot - away from wireless delusions of grandeur and towards a bit of realism.
I am very easy to get along with, but I don't have time to waste being nice to people who are being stupid. -Theo
I would sugguest that you read the FCC Part-15 rules that all WiFi equipment operate under.
To Quote:
15.5 General conditions of operation.
(a) Persons operating intentional or unintentional radiators shall not be deemed to have any vested or recognizable right to continued use of any given frequency by virtue of prior registration or certification of equipment, or, for power line carrier systems, on the basis of prior notification of use pursuant to 90.63(g) of this chapter.
(b) Operation of an intentional, unintentional, or incidental radiator is subject to the conditions that no harmful interference is caused and that interference must be accepted that may be caused by the operation of an authorized radio station, by another intentional or unintentional radiator, by industrial, scientific and medical (ISM) equipment, or by an incidental radiator.
(c) The operator of a radio frequency device shall be required to cease operating the device upon notification by a Commission representative that the device is causing harmful interference. Operation shall not resume until the condition causing the harmful interference has been corrected.
Most users believe that (c) is the "first-come-first-served" clause, but it actually refers ONLY to the primary user of the frequency, the ham radio operators. All secondary users are subject to (b) which is that you must accept any interference from a type-approved radio, including that which may cause unintended operation.
If you are ramp up your power, you are no longer type-approved, and can be subject to (c). As long as your equipment is type-approved, or under the EIRP power limits, you have as much right to the frequencies as anyone else.
Note: This does not allow you to specifically construct a "jammer" or simply change frequencies specifically to interfere with someone else. That would be willful interference, and WILL land you in civil court. If you ever end up there, be prepared to show that you made a good faith effort to mitigate the interference issue, and that there were specific technical reasons why you needed to be at the frequency you were at...
But his Etch-A-Sketch is already wireless?
If they actually have a computer, simply connect his computer to the wireless router via CAT-5.
Okay, it's sorta' like that. I used to have an economics teacher that always used to say, "If socialism worked, I'd be a socialist, but it doesn't, so I'm a Capitolist." He said that if communism and socialism worked (which would only be possible with leaders who truely did everything in the best interest to the people), then people would have pretty good lives, but people with power get corrupt, and so those societies don't work. So we turn to the more Laze Fare (I know, I misspelled it) ideas because they at least work better.
Anyway, if we didn't have the constant pessimism that regulators (of any technology) would abuse and misuse their powers, then we could get all of the frequencies regulated and probably be just fine. But, since various regulation groups tend to be bent one way or another due to various large interested companies, the unregulated seems like the best choice to us...
For certain applications, regulated sounds fine. In others, like in private homes, it'd be horrible. My solution: offer companies that are willing to pay extra the ability to use regulated frequencies to do their networking, and let the rest of us have at it in the unreg spectrum.
GentryDigital.com - A Digital Photography Education
I've been using wireless in the form of packet radio for almost 15 years and using aironet gear (the guys cisco bought) in production since 1998. While it's true that there are a lot of networks in the air most of them continue to function despite the interference. Isn't that strange??? no it's not ... all you need is a good enough piece of hardware to pick the sound from the noise. Perhaps more importantly I am writing this over a 5 meg full duplex wireless connection that gets fast internet to me from 13.2 miles away and hasn't dropped a packet in three years!!!
If there are 20 networks in range of you, why aren't they all doing the mesh thing to maximize bandwidth and get some redundancy too? This is just like with OSS - everybody wants to "homestead the frontier" instead of realizing that it's not a frontier anymore, and cooperate with what already exists. And those who are smart enough to do so want to secure the hell out of their networks too, not share with the neighbors at all. Just human nature, I guess.
The lack of organization is really inefficient. I'm surprised there aren't more organized free community networks nowadays; I really thought that was going to happen more, and that the big corporate empires wouldn't be as efficient about covering large areas with hotspots and then charging big fees to use them. A lot of hams have some sense of duty to use their skills for community service, but a lot of wifi hackers don't, apparently.
Wireless in my apartment is hopeless. There are only a couple of networks but even before there were any others my wireless wouldn't even work in some rooms. And it drops so many packets that it's unusable for anything but light webpage browsing. Doing anything that requires a constant connection fails horribly. I had to go back to a wired which obviously works fine.
-- You're too stupid to be an atheist.
ah if I only had a laptop... But once a Real Programmer(TM) has started coding, nothing will get in his way (particularly not pesky little things like ergonomics) until the job is completed, or he passes out from lack of nutrition/sleep/water.
"I think it would be a good idea" Gandhi, on Western Civilisation
-"A lot of FHSS devices are used in warehouses and the like, as FHSS is fairly immune to noise, but had a much lower limit on it's speed (2 Mbps or slightly more if you go vendor specific)." Uhmm, no. Frequency hopping (FH) is no more immune to noise than any other technique. This is because noise power tends to be spread equally over all frequencies (white noise) so there is no benefit in hopping from one frequency to another. It is however more immune to interference since this (sometimes) tends to be unequally distributed over frequency. -"That's not to say it's not possible though, and many FHSS networks have very poor security (as they haven't gone further than WEP with 128 bit keys)." This is misleading. If there's security issues then they may be due to the choice of protocol, encryption, hop codes, etc, but not to the use of FH itself. In fact, with proper choice of hop codes, security is considered to be one the strengths of FH. This is one reason for its common usage in military communication systems (difficult to intercept a signal transmitted at a frequency which is varied over time in an 'apparently' unpredictable manner).
Unfortunately you mistake what you're paying for. That $45/month isn't for the local link, it's mostly for the transport between your ISP and the rest of the Internet. Your scenario B has you paying $45/year for access to... well, nothing really. Only those computers associated with the same wireless access point as you. Basically exactly what you'd have if you set up an 802.11b access point of your own that wasn't plugged into a LAN. Now you have to go about getting your access points connected up to the rest of the Internet, and that costs money. Just like your home AP requires a router and connection to your ISP to let you talk to the Internet, a wireless service provider's access points require peering and transit service from a backbone provider or peering point to let the people attached to them talk to the Internet or they're just isolated LANs.
Obviously you use your computer differently than I do.
I have a laptop. 75% of the time, it sits on the desk, right next to the router. But it's that other 25% of the time that made my $30 wireless router worth it. Being able to yank the power plug and usb hub, and pull the laptop onto the couch while watching tv. Or reading slashdot, deciding I want to take a dump, and hauling the computer onto the pot. Or when the wife has company over, so I want to go take the computer and hide in the other room.
Why mess with all the hassle of setting up HPNA, or running wires, when you could just buy a $30 router and it works ANYWHERE in the apartment, and you are free to move around?
Sure, I'll be the first to admit that there are problems with wireless, just like everyone else is saying. Speed is slower, and overcrowding, yeah yeah. But I don't think it's overhyped -- I use it because it's cheap and convenient. That's not hype, it's quality. And if the high quality means that the rest of the non-geek world catches on and wants to use it, well, so be it. Sure, it could have the side-effect of lowering the quality slightly for me, but that doesn't mean that there's too much enthusiasm. It means the world has caught on to something useful.
Dude, if you're gonna shoot, shoot. Don't talk.
Besides, nobody may understand what the hell you're talking about, but some will pretend they do, just to score some karma. Hell, you might even get a response that reads like it was lifted directly from a rejected fanfic script for ST:TNG!
Geordi: True, a 1-6-11 spread will maximize the spectrum with minimum nodular-crossinterference, but a triple-stepped 4 channel spread will remodularize the wifi matrix-
Riker (adjusting uniform): which would maximize headerless throughput and give us another 10 Teracycles on the core processes!
Picard: Make it so.
Troi: I'm feeling a sense of accomplishment here.
Chewbacca: Arrrr!