Slashdot Mirror
Ask Slashdot: Overcoming Convention Hall Wi-Fi Interference?
bbowman writes "One of my job responsibilities is to set up the small network for our company's exhibit at the trade shows we attend. The mobile demo devices we use depend upon a reliable Wi-Fi connection to a router I have in the exhibit. In the days leading up to the opening of the trade show, W-iFi connections are reliable and work as expected. However, as soon as the show opens none of our devices can reliable maintain a Wi-Fi connection to the router. The devices we use at the trade shows are Windows-based laptops, iPods/iPads, Android tablets, and a variety of Wi-Fi enabled cell phones. I have tried using channels 1, 6, and 11 (as well as the others) and used different routers (Linksys, D-Link, Netgear) without success. I'm sure it is likely that there are poorly insulated electrical cabling, fluorescent lighting, and other issues that would contribute to Wi-Fi interference in the convention hall. A quick scan shows dozens and dozens of discoverable Wi-Fi networks nearby. If I take the router back to my hotel room, I have zero connection problems. How can I overcome this so that Wi-Fi works reliably in the convention hall?"
I'm sure everyone will understand.
Alternatively, install a giant metal Faraday cage. (Good luck with that.)
The World Wide Web is dying. Soon, we shall have only the Internet.
Perhaps a Faraday cage?
http://en.wikipedia.org/wiki/Faraday_cage
Try an RS232c null modem cable.
If your devices are 802.11n compatible, you could put your router in n only mode... The 5.4ghz band may be less crowded.
- Holy crap, I've got MOD points! Who thought that was a good idea.
If you want your equipment to work with the current standards, you need to mod it to use another radio frequency. In their current as-written form, the 802.11* standards just don't have a way for equipment negotiate graceful degradation in this situation, and that's what it would take.
It's a Japanese channel.
For the devices that support it (decent laptops, iPad, and possibly other tablets), going to the 5 GHz band is a huge win. There are plenty of non-overlapping channels, and congestion is lower. The problem is that most WiFi enabled phones only support the 2.4 GHz band, so this will not cover all cases.
Illegally, you could pick up your wifi equipment from a country that allows channels higher than 11. I don't know what those frequencies are used for in the US, but there's a better chance you won't pick up interference (depending on what those frequencies are actually used for). Or, of course, use 5GHz (a or n). Less range, but also less interference.
Get a good enterprise class AP + controller. The controller will do most of the job avoiding interference, auto channel utilization, etc. Saves your time and headache.
It's technically illegal but if you set your router and network card location to japan (I think), you can use channel 14 which is quite far (on the spectrum that is) from the "standard" channels. Most wifi devices physically support this but they may be limited by the driver
At least for the laptops. There's a lot more spectrum there, and it's much less saturated. Probably not an option for the phones, though. Also, wired ethernet when possible.
"'Tis great confidence in a friend to tell him your faults, greater to tell him his." --Poor Richard's Almanac
It's why wired networks are better
This is a big problem for me. Every time I travel, I get interference when I sleep.
Here's what happened last time:
http://www.youtube.com/watch?v=guGchg4mbLs
Use a omni-directional antenna that has as much gain as you can afford, and avoid channel 6.
Alternately, crank up a wifi monitor and show that there are 200 separate networks in 15 channels, and then run everything via cat 5/usb.
Get a enterprise class AP + controller. That will help you to do auto channeling, interference avoidance, etc. Saves lots of your time and effort.
Use a dish off the access point pointed at the device under demonstration.
Overpower the interference. This would be especially true if it truly is background (lights and such) but
will work just fine overcoming 'The Masses"
Below channel 1 :).
The problem is intrinsically a hard one, 802.11* wasn't really designed for a zillion flacks in a large room, each toting personal cell routers and whatnot.
However, it is possible that the problem could be solved by money. Let's just say that "(Linksys, D-Link, Netgear)" isn't exactly an honorable lineup of the finest names in Serious Wifi. Cheap, yes, quite delightfully so. Built right down to price? Well, you could say that...
You might want to do some looking into the world of "industrial wifi" products. The environmental resistance of such will be total overkill for a tradeshow floor; but (successful) offerings in that sector are designed for people who need their network to work despite the fact that it is in the middle of a factory floor or next to the arc welder or what have you.
The trouble with going upmarket, though, is that it can be somewhat hard to tell what is genuinely better at wireless networking vs. what is just the same old shit on the wireless side; but in a POE, ruggedized, -40/+135 thermal resistant, with baked-in proprietary management protocols in the firmware, container. You really want the former, not the latter...
Why is Internet access and Wi-Fi always so terrible at large tech conferences?
You may be able to politely suggest to your neighbors that everyone turn down their signal and agree upon a pattern of channels that do not overlap, thereby
1) reducing the possiblity that your closest neighbors are directly interfering and
2) limiting the amount of signal that interferes with your next-closest neighbors.
As an example, think of the coloring structure of a political map, where two entities of the same color cannot touch.
-sweBs
Same two options as always. Either overpower the interference or turn off interfering devices with a universal remote
Build your exhibit with a dungeon/prison them to hide the faraday cage that isolates you from the rest of the auditorium. Add lots of dry ice and flashing lights and not only will you have a working exhibit, it will look cool as well.
I am Slashdot. Are you Slashdot as well?
and directional antennas, if your literally 6 foot away from an AP with a +6 db ant pointing in your general direction your going to have a hell of a hard time not getting a connection, one in each corner of your exhibit area should do the trick.
There are plenty of tools to try to find free channels:
https://market.android.com/details?id=com.farproc.wifi.analyzer&feature=search_result
Also make sure you've got security enabled, otherwise everyone and their brother will try to connect and probably overwhelm the router.
FYI Steve Jobs routinely uses out of spec channels. For WWDC, this used to be channel 13, which is not licensed for use in the US, but is in Japan.
This got to be a problem (leading to the famous "you've got a choice..." speech) when enough Japanese Mac developers attended without changing their locale, and all the Japanese machines ended up on channel 13 because it was "less crowded" (for obvious reasons).
-- Terry
Look into WiMAX gear in the 3.5 to 3.6ghz range. You should be able to find a home router and usb or pcmcia gear for your laptops. Cell phones on the other hand are probably a no go.
In the days leading up to the opening of the trade show, W-iFi connections are reliable and work as expected. However, as soon as the show opens none of our devices can reliable maintain a Wi-Fi connection to the router.
I doubt it's this:
I'm sure it is likely that there are poorly insulated electrical cabling, fluorescent lighting, and other issues that would contribute to Wi-Fi interference in the convention hall.
...and more likely this:
A quick scan shows dozens and dozens of discoverable Wi-Fi networks nearby.
I would recommend trying a few things:
- Reduce your RTS threshold, if your AP supports it.
- Reduce the fragmentation threshold, if your AP supports it.
- Play with data rates, reducing them if your AP supports it.
If your AP does not support any of those options, go out and get a real AP.
Never start vast projects with half-vast ideas.
http://en.wikipedia.org/wiki/Faraday_cage
"Liechtenstein is the world's largest producer of sausage casings, potassium storage units, and false teeth."
One solution is to use 5GHz network (802.11a) which has lot more channels than 2.4 GHz band (802.11b/g). But some of the devices you mentioned do not support 802.11a (eg. iPods). Although you could get mobile phones and pads that have dual band support (2.4 & 5). You could also try running your network at a higher power (look in your router configuration) and maybe use better antennas than those that come with the router by default. Also you could squeeze some more performance by running your network in 802.11g only mode, hence will not work for 802.11b devices.
If your router supports it, switch it to channel 14. You can do this with most open source firmware (dd-wrt) - Beware, it is not legal - but I doubt the FCC is going to bust you for it. Channel 14 should have a lot less traffic, and might be usable.
Some Cisco and other high end access points have beamforming networks that can place antenna nulls in directions of interferes (other AP's, microwaves, etc) and point the peak of the beam directly at a user among all types of other fancy tricks.
They work wonderfully well in noisy, cluttered environments. Give them a shot.
http://www.cisco.com/en/US/prod/collateral/wireless/ps5678/ps10092/white_paper_c11-516389.html
http://en.wikipedia.org/wiki/Assault_rifle
Get a N class router with Linux firmware support (I recommend the RT-N16 + DD-WRT/Tomato). Set it up to run at 300Mbps on the dual-band and the transmit power upped to 100 or 125. It's practically a 2.4GHz G wireless jammer.
My consulting business specializes in providing show management services to associations putting on large meetings. I work with Facilities to provide wifi zones, as well as information kiosks, and internet access kiosks for attendees. I recently did a large meeting in Chicago (30,000+) attendees. This presented many challenges but the biggest issue was on the exhibit floor where a site survey revealed 160 rouge access points on 2.4ghz. Not only was the building Wifi having troubles it was like the wild west among the exhibitors, most had to be within a few feet of their access points while some others used amplifiers or enterprise class APs to get the extra signal strength, if all exhibitors upgraded their signal strength everyone would be back to square one. I tried to alleviate the problem by setting up a secondary SSID on the building infrastructure for exhibitors with decent bandwidth and tried to encourage some of the exhibitors to use this instead of their own. I got the rouge access points down to 90, but certain areas of the exhibit floor where still flooded. This is an inherent problem with wifi that the more APs competing over the same radio frequencies the more noise and less throughput you will manage. Without a doubt the best solution is to move to another frequency, as I found very few were operating on 5ghz (802.11a), The real challenge is to make exhibitors and attendees aware of the shortcomings of this technology in an environment with so many users and competing APs.
Channel 13 is allowed almost everywhere except the US. Channel 14 is the Japan-only one.
What is damned is alot of equipment is hard-wired to not allow channels over 11 worldwide, lest someone go use it in the US - lowest common denominator and all that.
If the problem were fluorescent lights or something, then it wouldn't work before the convention. I'm pretty confident from your description that the problem is that you're all trying to use the same radio spectrum at the same time and there's just too many of you competing instead of cooperating.
Wifi radios are just that; radios. if you have too many, they all interfere with one another. I suggest that this is a problem for all the other people at the convention, too. Therefore, the people who put on this convention should have already heard such complaints (unless people are too embarrassed to complain?) and they should be happy to deal with this as a problem of the commons (radio waves in air).
So what needs to happen is that every exhibitor who requires wifi needs to get together with every other such as organize shared resources so you can all actually share the airwaves, instead of unsuccessfully trying to compete with everyone.
Every 6 or 7 exhibitors who are adjacent should share one Wifi router. That'll probably fix the problem.
Most trade shows use enterprise-grade WiFi gear to provide WiFi access on the floor and also to block unauthorised access points, i.e. your wireless router. They charge good money for their WiFi, but that is your best bet. You may also be able to negotiate with them to unblock your router.
duh
With the spectrum pollution in a trade show and the multitude of reflective surfaces getting any wireless connection to work reliably would require, as others have mentioned, either that everyone else shut down their equipment or that you and all of yours is enclosed in a fine mesh faraday cage. Even with all N capable equipment in the environment you describe, stable connections would be nearly impossible.
Demo shutting down all the demos.
Wait till the last second to setup, leave your gear in hardened, conductive cases...EMP...clear networking.
I believe you are boned. Go out of spec. Your going to have to go further then japan out of spec. Too many people will be on 13 and 14. Getting the phones working will be joy.
Get a 3g enabled phone for all the networks on the off chance that the telco overbuilt the conference floor. Leave a demo server available on the net just on the chance that your phones will have 'good enough' 3g connections. It would suck to be boned on 802.11* and only running on cables for the laptops (you know it's your backup plan), and find one of your personal smartphones still had connection, but no server configured.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
Why not simply use a narrow-field directional antenna for your demo? If you're just feet away from it, it seems unlikely that other nearby networks would be strong enough to drown out the signal.
"Tell me doctor, with all of your defenses, are there any provisions for an attack by killer bees?"
What is damned is alot of equipment
I did not realize that there was an Alot that was made of equipment.
Sounds like either forcing 801.11n only or using 801.11a is the only inter-operable alternative unless you can modify the devices and play with other parameters.
What about getting the convention hall organizers (or you and your nearby booths) to try and build a mesh, so everybody is on the same network (and can somehow tweak parameters to reduce interference)? Maybe coordinate the channels between nearby booths so they don't overlap? Not that there are than many channels to distribute.
How do you know that he did this, out of curiosity?
-- Nathan
Wifi channel space has obviously been used up. Use either:
http://michaelsmith.id.au
Use the 5Ghz range (either 802.11a or 820.11n) to get more channels.
You need reliability but if you don't need throughput reduce the channel width to as small as possible (usually whats supported by the station) to reduce interference.
Use directional antenna on the APs to reduce interference for the APs (won't help the omni stations though).
Use more APs running at lower power and more localised to the devices.
More mW is never better. Ask your neighboring stands to reduce the juice.
Perhaps the router could be connected to a dish above the booth pointed straight down like a street light. This could overpower other signals in your booth and reduce your interference to other users. Dishes at wifi frequencies aren't large.
use a cable!
I used to set up networks for training and conferences my group did. We picked up an Extricom http://www.extricom.com/. I never had problems in the dozen times I used it. Its an interesting system, it has a central Wifi unit, and you run cat-5 out to 4 remote transmitters. You can place them spread out over the area. Admittedly the model I used, the Extricom ESX400 and 4 radios no longer seems to be available but check it out. -Joe
In my experience, the only way to get reliable service at a trade show is to pay fifty bucks a day to the company running the show. Not sure how they do it, but my cynical guess is that they deliberately pipe in interference.
http://en.wikipedia.org/wiki/Nicolas_cage
Obi-Wan: "I felt a great disturbance in the Force, as if millions of voices suddenly cried out in terror and were sudden
On the first point I would recommend you stick with a simple ecosystem of same vendor / same model, or better then, with a spare one of another model of another vendor, just in case ;) (as far as monoculture is a concern, that would solve it). But the point is that routers of the same kind tend to have more reasons to interoperate better.
Another interesting choice : Make a map of the estimated coverage interlap of the routers in the field. Then colorize it with a color for each choosen channel, so that never two neighbors (closest overlapping person in that direction) share the same color.
With channel 1, 6 and 11, you would get 3 colors, but nstead of using 3 colors, use 4 colors by letting frequencies overlap better if 3 don't suffice. That would make channels 1, 5, 8, 11 (I prefer to give the extra space to the lowest frequency, on the logic that it spreads a bit better, so would be the least at ease).
Then, of course, if you could use some power control. Listen to neighbors, estimate their activity relatively to yours, and scale your power according to that difference. The less you're active, the less you merit to dispensate your imprint on the local spectrum.
My 2 cents, of course. But I admit I did a Ph.D in a related field.
[edit : sorry, that's the repost under my account. My bad.]
On the first point I would recommend you stick with a simple ecosystem of same vendor / same model, or better then, with a spare one of another model of another vendor, just in case ;) (as far as monoculture is a concern, that would solve it). But the point is that routers of the same kind tend to have more reasons to interoperate better.
Another interesting choice : Make a map of the estimated coverage interlap of the routers in the field. Then colorize it with a color for each choosen channel, so that never two neighbors (closest overlapping person in that direction) share the same color.
With channel 1, 6 and 11, you would get 3 colors, but nstead of using 3 colors, use 4 colors by letting frequencies overlap better if 3 don't suffice. That would make channels 1, 5, 8, 11 (I prefer to give the extra space to the lowest frequency, on the logic that it spreads a bit better, so would be the least at ease).
Then, of course, if you could use some power control. Listen to neighbors, estimate their activity relatively to yours, and scale your power according to that difference. The less you're active, the less you merit to dispensate your imprint on the local spectrum.
My 2 cents, of course. But I admit I did a Ph.D in a related field.
glop
Startup aircrack. Deauthenticate all mac addresses but your own.
Bump up to 5GHz There's a lot less people with the good stuff.
Tom's Hardware did an excellent and extensive test on WiFi networks not long ago. It is well worth the read, as is the first part of the series. http://www.tomshardware.com/reviews/wi-fi-performance,2985.html
Depends on the physical layout of what you are up to, but directional antennas on the mobile equipment and a 2.4ghz absorbing backdrop behind your stuff might work. Think old Pringles can style directional and a grounded copper mesh covered in cloth for the backdrop.
Comment removed based on user account deletion
Most hotels and convention centers locked in long-term Internet access contracts long ago, before Wi-Fi became the norm, and before free access became popular. The site's Internet provider might be intentionally crowding the spectrum so conference organizers are forced to pay for access, and attendees are expected to go through the conference organizers for a password.
The problem is that before the conference, no one is setup (or set up at weird times). Also, potentially depending on your conference, the conference organizers are running their own Wifi. This means that before the conference, your access point works great, and once the show starts, your access point is being drowned out by everything else.
If you're at a conference with central Wifi, you're probably screwed, because their APs are licensed and VERY LOUD (like Interop). If your conference doesn't have central wifi then using something louder will be helpful, but also switching to 802.11a will clear you more spectrum (possibly even if there is conference-wide wifi). Your other option is wimax, but it's hard to get a license these days because Sprint and Clear are using it for 4G cellular.
Use dd-wrt and crank up the signal strength...
I seem to remember that Japanese wifi devices had channels 12&13. I don't think they are FCC approved though
This doesn't meet the specs of the question, due to the particular devices you wish to support, and the fact that many of them are deliberately incompatible with it.... but there is a technology that I've used successfully many times in the past to overcome the problems inherent and unavoidable in any electromagnetic wireless communication system. This technology dramatically reduces signal attenuation due to distance, it reduces interference from external devices that use the same frequencies, it allows for dramatically higher data rates, and as a bonus it even adds a level of security/privacy: requiring extremely close physical proximity to sniff the signals. These features come standard with the tech itself, or can be enhanced with a special "shielded" version. There are even varieties capable of actually powering the devices. This technology is, of course: Wire.
One of these days I'm going to patent it, clean up on license fees for a few months before the patent gets invalidated, and retire to Mackinac Island.
http://alternatives.rzero.com/
Hardwire everything like in the old days
And then read up on microwave amplifiers and horn antennas. You can then run a lot more power on the 2390 - 2450 MHz band. Your router becomes the exciter, feeding the amp, feeding a horn.
1. Use a mist machine to both knock out the competing electronics and attenuate the other APs. Add a light show to hide the sabotage.
2. Make a video overlay for the demonstration if it could fail live.
3. See if your devices that can do networking over bluetooth or USB.
How do you know that he did this, out of curiosity?
-- Nathan
I am unsure if the channel 13 statement is true. I was in the room at WWDC when that happened, and I'd run up some tools on my laptop as the Wi-fi behaviour was awful.
There were over 1200 Mi-Fi type devices , a couple of hundred Android phones acting as wireless hotspots, and about 400 laptops sharing their interconnections.
I'd have thought the military term for nearly 2,000 Wi-fi networks in the same room was broad spectrum jamming ;)
make a cantenna
i work at a convention hall...are you sure they're not containing your AP? we don't allow outside managed gear although most people don't realize it until they can't connect to their router five feet away. we have our APs contain any rogue APs to avoid losing $$$ to folks showing up and trying to provide free wifi. most of the convention and exhibit centers we deal with do the same thing. the last thing we want is someone providing unsecure free wifi in the building and then we get blamed for 1) shitty bandwidth 2) mitm attacks 3) bad customer service because there's nothing wrong with our gear when you have an issue with dude in the next booth's cell phone tethered AP.
Yes, switching to 5 GHz should help, but not all your equipment may support it.
The other solution is to overpower them. Something like this can be cranked up to 19dBm assuming you only need internet-level speeds. You'll screw everyone else, but it should work.
Adding a higher gain antenna will only solve the downlink interference issue. The uplink towards the AP will still be a mess with 150 rogue APs floating around. This would be like trying to hear someone talking in a room full of people. Yes focusing where your listening helps but ultimately if there are still a ton of 'voices' in your path, you will still not be able to hear the intended person.
With an 802.11 AP you should also keep in mind that it uses a CSMA protocol which basically means everyone accessing the AP will transmit at random intervals. If by coincidence, two users transmit randomly but at the exact same time, the data will collide and be lost. The long and short of it is, as the number of users trying to connect to an AP increase, the quicker the overall system throughput decreases and the more terrible the performance becomes. Your only way around this is to either a) add more APs or b) change technologies (WiMAX/LTE). Of course, adding more APs will only add more interference thereby compounding your issues.
All that being said, my suggestion would be to go with a neutral-host cellular in-building solution, retransmitting an operator's LTE or WiMAX throughout the building and give up on unlicensed spectrum WiFi. The licensed frequency of this system and the higher 'legal' transmit power will increase the performance 10fold. Unfortunately, this will also increase the cost and you will give up control of the system but at least you could guarantee the system will work without a pissing contest!
You could ask the cellular operator to put WiFi on their DAS which will put the responsibility of the performance on them (getting rid of your headache).
One last thing, adding amplifiers and using higher gain antenna may put you over the power limitations of the 2.4GHz/5.8GHz band imposed by FCC. Depending on the ethical stance of your business, you may want to avoid going this route.
I have been setting up wired and wifi networks in trade shows for 10+ years now. :)
Welcome to wifi Hell
I do not think achieving 100% reliability is a sane goal in that context, but I found that there are simple ways to greatly improve the odds.
> Consumer-grade routers / AP are no good. They often do a fine job, and they always give up quickly.
In my view, using small-business equipment is a better way to go : still affordable, and a lot more resilient.
For about 400 $ you should be able to find a basic gateway and access point (new - I'm thinking sonicwall / netgear prosafe / hp pro curve...)
And if it matters, you still have web-based interfaces available to configure them.
> Use fixed ip when possible. Avoid encryption if you can, avoid wep if you can't.
> Reserve the network for your demo gear. If you also have to provide internet access to people working on the booth, use a separate network/gateway (your current linksys/d-link/... router might do the trick if less critical).
> Also : wired.
Not an option for your whole setup as I understand, but maybe part of it.
I am no sysadmin/network admin/whatever so this is basic stuff, but works for me on a regular basis in your exact predicament.
Hope this helps, and good luck to you.
I was thinking that and I'm amazed that nobody had mentioned that earlier. Even relatively big parabolic antennas for WiFi (~1.5metre dish for 2.4GHz or 5GHz) are cheap and there are things like ubiquiti access points that even have much smaller directional antennas as part of the sealed unit.
There's a lot of different directional antennas out there.
Look at their "Airtime fairness" capability. Works amazing in places just like this.
CSMA (Carrier Sense Multiple Access) makes this a problem. The way this works is the AP and clients listen before transmitting. If they hear another signal above a certain threshold they will try to cooperate by not transmitting until the channel is clear. It's like trying to hold a conversation with a room full of people talking.
5 GHz will probably not work as your client equipment is most likely not equipped with 5GHz capable radios. If they have mini PCI cards you could upgrade them but otherwise your stuck with 2.4GHz.
Some high end WIFI APs have a signal threshold setting. This tells the AP to ignore signals below the set level. This could help as it could be set to ignore the radios further away, allowing the radio to transmit even when other weaker signals are present. This would severely limit your APs range but could help.
Terry works at Apple.
Congestion is what happens when there are too many clients busy on the channel that you are on. This is workable. Interference is what happens when there are clients/AP's on adjacent, overlapping channels (802.11b/g 2.4Ghz). This is unworkable. Just because you have your AP's on the proper channels, doesn't mean that those around you do. The best solution is to move to 802.11a (5Ghz) where there are a log more channels, and they don't overlap.
Get a network extender from sprint or verizon. connect it to the access point. connect everything via 3g to the extender. limit access to the phone numbers assigned to your devices. The extender ignores everyone else's device unless they're calling 911. Verizon - http://www.verizonwireless.com/verizon-network-extender.shtml Sprint - http://support.sprint.com/support/device/Sprint/AIRAVE_by_Sprint-dvc1230001prd
Look it up. These things will give you a more focused connection, but you will need setup the antenna correctly for the space you are using. I am assuming you have your wireless router/access point hardwired in your kiosk space. You need to setup 1 or 2 directional antenna and point them (preferably from above) at the area that you have your equipment/devices/demo space. The antenna you use will be dependent on the wireless connection frequency you use (802.11 a/b/g/n 2.4GHz, or 802.11n 5GHz). It will also depend on what kind of access point/router you have. Obviously this only works on routers which support different antenna, like the infamous Linksys WRT54GL.
:D
Personally I have a 19dbi panel antenna which I use on emergency, which I carry in my laptop backpack along with the said WRT54GL. I have been able to pickup and connect to networks 2 miles away (with line of sight) in bridged mode to allow my laptop to get on the net in a pinch (this was before phones had data/internet access). Worked out great for a school project once where the class was held at a remote location and we then had to write a team report/paper on about it. The professor gave bonus points to the first team to complete the assignment and sent it to him via email. We were already working on laptops, and I simply broke out the router, aimed the antenna back at our campus, spoofed my laptop's wireless mac address for the router's (since my laptop was on the approved wireless access list), and we sent the report. We also let our professor connect his laptop so he could receive it
We were all warned a long time ago that MS products sucked, remember the Magic 8 Ball said, "Outlook not so good"
You're running into the big problem with wifi: Everything on the same channel has to take turns. If there's 40 APs all in the same vicinity, they're going to start a round-robin game of who-goes-first, and if there's enough other interference, they're all going to keep yielding, and nobody gets heard. See "myth" #1 here: http://www.cisco.com/en/US/prod/collateral/wireless/ps9391/ps9393/ps9394/prod_white_paper0900aecd807395a9_ns736_Networking_Solutions_White_Paper.html As it mentions, this is more accurately called co-channel cooperation, rather than interference. And it's a huge problem in a convention. You're really kinda screwed in that scenario... unless you can use another spectrum, but it sounds like that's not much of an option given that you're trying to demo on consumer client hardware that's almost invariably 802.11b/g/n, 2.4ghz. Use A band if you can. Another thing you can try, and this would take some serious cooperation with other convention-goers (read:social engineering, perhaps?) is to get everyone to turn down the broadcast power on their APs. That's what the real problem is. If everyone talks quieter in the library, more people can carry on their conversations in their corner of the woods. You can also try to work out a pattern with nearby AP users to switch their APs to channels that don't overlap with yours. Be careful with this one, cause if you've got two APs on channels 1 and 4, they overlap enough it might still cause co-channel cooperation. This is probably pie-in-the-sky thinking that your fellow convention goers (and possible competitors even) might cooperate to that level, but there's no hurt in trying, right?
"For WWDC, this used to be channel 13, which is not licensed for use in the US, but is in Japan."
Does he have special WiFi firmware to go with it?
Channel 13 stops working on my MacBook if any access points with the country code set to 'US' are nearby, even though I'm clearly in Australia (where 13 is allowed).
Well sort of solved it. We were demo'ing a handheld wireless device that did not have a wired port. We opened the device, popped the nano sized connector from the wifi module to the PCB inverted "F" antenna and connected a very thin coax to the now vacant wifi module's antenna port. We connect all of the unit's coax leads to a RF mixer (think an analog version of a router) and also hooked up a generic wifi router via coax to the mixer so that the handheld units could talk to our on-site server (handheldmixerWiFi routerserver). Our demos worked perfectly. Nobody else had anything working and one of the main points of the show was to show off wifi capabilities. It took a bunch of cables and adapters/gender benders, etc as consumer routers have most of their pin's gender reversed so that you can't do this with retail parts and cables.
13 is allowed in Australia and Europe as well, but overlaps significantly with channel 11 and a few of the other higher channels.
There is a channel 14 that is only allowed in Japan and it is far enough above even channel 13 that there's virtually no overlap.
http://en.wikipedia.org/wiki/List_of_WLAN_channels
Specialist Mac support for creative pros, Melbourne
Go with the RTS/CTS and fragmentation threshold settings, typically I use something like 1024 for RTS/CTS and 500 for fragmentation threshold and also a hi gain directional antenna designed for Line of Sight transmission. you can apply this to a wireless bridge, some times called a WDS or Wireless Distribution System, thereby creating a good link to the network and also a high power very local wifi network for your booth. Good luck.
Wifi is inherently unregulated. It will have interference and lots of traffic from all the users and devices using the same ISM band. Everything from phones to toys to microwave ovens running in the green room of the booth next door which has a set dressing budget larger than your annual revenue.
You are all unlicensed users and get to accept whatever QoS or lack thereof that you get. There's nobody to whine to, or whine about. It is what it is. On the other hand, there are scary licensed users on that band who have actual priority over the unlicensed users and can use tremendously bigger output power which could obliterate the signal for everybody else. They aren't supposed to do that but they can. And if they do, good luck to you. Your typical wifi router will be useless.
The real question you need to ask is why you are apparently running mission-critical or at least very important business stuff on an inherently congested and unregulated radio band that has no QoS and no promises? What the hell are you thinking? To do what you are doing is naive at best and stupid at worst and makes me wonder why I should trust whatever it is you sell to perform when such poor planning went into the tools you use to sell whatever it is to make.
It's the same when a company selling high-end items chooses to print their marketing collateral on the cheapest paper they can find and use morons to do the production thus ensuring that every book has an upside down cover, or the company name spelled wrong. Gee what a great impression.
You know what impresses? When the people working a tradeshow booth know their stuff, and the demos work, and they have their act together and do not whine about how everybody's iPhone or iPad is using up all the bandwidth.
Now, please excuse me as I use the microwave in my booth to reheat last night's pizza. It will only take a minute or so but your wifi will crumble during this time. Meh. No need to apologize. My dinner has as much right to the ISM band as your apps do.
Sig for hire.
For WWDC, this used to be channel 13, which is not licensed for use in the US, but is in Japan.
If he can legally buy a human liver, this does not come as a surprise.
What a great country!
I believe you can get licensed radio's from tranzeo. If not they have some other band solutions as well. They are not that expensive or hard to set up if this is all that important.
Oh Crap, I'm an optimist.....
I just googled "you've got a choice speech" and the only result was this post.
Clearly not famous, but I'm interested in knowing more....
Channels 12 and 13 are technically allowed only with smaller output. A normal convention hall router would exceed these specs substantially, but for a limited-power small substation, channel 13 would be worth a try. While slightly overlapping with channel 11, you should have sufficient breathing room to convince a connection to stay enabled.
The USA regulate Wi-Fi to bands 1-11 but if you get a device from Europe or Japan it can use more slots which may give you less clutter.
http://en.wikipedia.org/wiki/Wi-Fi#Limitations
Totally not happening...
Which dept? Ask JR what happens. I'm on the team
Nope. You'll just have the network provider at your booth in 3 min.
Is the problem. It works fine before the show starts and goes to shit after because the AP's are now keeping track of thousands of wandering WiFi enabled cell phones. Consumer grade AP's just can't handle it, some of them will even crash due to their MAC tables filling up (they still have to keep track of MAC addresses even before they authenticate)
In the future, invest $100 in a used Cisco Aironet off ebay. They cost $650 new while your Dlink costs $50 new for a reason.
The fact that this was modded "Insightful" adds a sublime ridiculosity to the Slashdot Friday Night.
Wifi is a polite protocol. When your laptop hears others speaking, it waits for a dead space (and then a random amount of time) before transmitting. This makes for a highly innefficient protocol when there are lots of users. Wifi throughput drops like a rock when you add a second user (every try to transmit data between two laptops on one AP? - you never get that full 54 mbps do you :) )
Imagine only one person speaking at a time in the entire conference hall. When one is done, you wait a bit and then another speaks. It would be weird and inefficient, but so is wifi.
So, there's already multiple suggestions of using 5ghz, or channel 13 or 14. I second using 5ghz if possible.
But, if those are not possible, I would do the following:
1) Lock access point, and cards if possible, to 54mbps (or possibly a little lower, like 36-48mbps if you plan to move further away). I've had some situations with noise where the AP and/or card see packet loss and go absolutely apeshit, "death spiralling" down to like 1mbps, or if you are lucky 2mbps or 6mbps. I think what happens is it sees excessive packet loss, drops one speed (i.e. 54mbps down to 48mbps), since each packet takes longer it's MORE likely to be obliterated -- packet loss increases. So it drops speeds again. Once it gets down below 12mbps, the modulation is simple enough that it's not obliterated by the noise any more so packet loss decreases. But, "some" packet loss at 54mbps will still get you like 20mbps, whereas the nice no-packet-loss 1mbps will get under 0.5mbps. Newer Linux kernels use wifi rate control called "minstrel" that is explicitly designed to avoid this so your android devices are probably good to go. (Of course, you won't get 20mbps in convention conditions. But the point still stands.)
2) Set the wifi fragmentation threshold LOW on your access point and whatever wifi devices you can. By default it's set something above 2000, which since TCP usually uses 1500 byte packets means fragmentation is never used. In a noisy environment smaller packets are less likely to be obliterated, so set it to like 256 since it'll be VERY noisy. Unlike MTU on an ethernet card which sets the biggest packet that can be transmitted, this wifi frag threshold is transparent... your 1500 byte packet is broken up into 256 byte fragments on one end and reassembled into a 1500 byte packet on the other. Generally, the recommendation is to try 800bytes and go from there, but since your environment will certainly be VERY noisy, I think it's advisable to just go with the minimum 256 bytes.
3) Turn off frame burst, which is also called afterburner or a few other names. This sends 3 (or possibly more) packets back-to-back, BUT I think the loss of any of them causes the system to retransmit all of them.
4) CTS/RTS. Clear to Send/Ready to Send. This may not help in this particular case -- really it is meant for the "hidden node" problem, where two or more devices can "hear" the access point but cannot "hear" each other, so they may both try to transmit to the access point at once. With CTS/RTS, the card sends a "ready to send" signal, and then the access point sends a "clear to send" telling the device when to send it's data -- and every other device also gets the "clear to send" and knows *not* to transmit during that time.
Step 1 won't hurt your speeds at all even in a clean environment. Steps 2-4 WILL cut your speed by about half if you had clean conditions, you'll be lucky to get 15mbps on a 54mbps link with these all on; but in crap conditions, it keeps your speeds from collapsing to zero.
You should try Airohive: it has nice APs and more importantly: an online wifi planning tool for optimally placing the APs for coverage and minimal channel overlap.
Simple, run a co-axial cable and a 60 dB attenuator between the two devices. Get it from L-com.
First put your antenna right on the table. Distance is king for radio reception. If possible, position yourself so the interference is coming from behind your booth. Metal window screen lining the back of your booth will block interference. If you can have little wings of promotional material lined with window screen to partially sheild the sides of your booth even better. Don't accidentally buy nylon screening.
Horizontal polarization: try making all your antennas horizontal. Depending on the reflections around the room's interior it could work.
They can have my command prompt when they pry it from my cold dead fingers.
... and configure your devices to work in the 13cm "high speed data" segment.
I have run the networking at several 600-1200 attendee conferences, and have a few things you might want to try...
If any of your devices can use 5.2GHz, make sure you deploy APs for that. 5.2GHz has way more spectrum, and in my experience it tends to work where 2.4GHz is pretty spotty. Try deploying with fairly narrow beam antennas like 90 degrees, so you are just covering your booth, ideally mount it up high looking down. Run at the lowest power setting you can. Use 802.11n equipment, which often seems to have better antennas.
In the end though, 2.4GHz at conferences can be very tough... There just isn't enough spectrum there. My primary recommendation to attendees at the conference I run the wireless for is: Use 5.2GHz.
Since your hardware isn't as flexible as it could be, you can't get away with tricks like special channels and using inbetween spectrum.
So with standard equipment and standard frequencies:
1) Use 802.11n, 5 GHz if possible (less crowded)
2) Try the wider modes (HT40) to see if they can hop around your interference better (just force it and see what happens)
3) Put metal bug screen around the back of your posters and such, (with your AP not behind them). This will decrease sensitivity to whatever is behind your bords but it will not help with anything infront of them.
4) Turn your beacon interval way down, so that if a device does loose signal it will find it relatively quickly.
5) More power for your AP will help, directional antennas are good *if* you know you're only going to be in a specific area (usually, at least a hemispherical panel works)
6) follow good RF practices (don't put antennas against other objects, for example)
7) operate multiple APs at once with the same SSID -- may help, might not too, depends on how bad your interference is. And how intelligent your wifi clients are. Some won't jump AP until the signal is almost uselessly low.
Hope that helps.
Try using Ruckus wireless for the AP. So far on an exhibition scenario it work wonder. We use it all the time on an exhibition floor.
Reroute the phase coils through the plasma regenerator, then boost the particle stream with a subspace flux inducer. If that doesn't work, try routing your signal through an anti-neutrino pulse, you should be able to generate one by modifying the main deflector.
It's amazing what you can learn about this stuff from a combination of Star Trek reruns and a complete lack of practical experience!
The world you experience is only a close approximation of reality.
I thought all of Apple's employees where in that huge sweat shop in china where they routinely commit suicide.
google for this: Either turn off your Wi-Fi (devices) or I give up
How many more years will slashdot have an off-by-one error on your Score in your profile?
Sorry, no useful advice for your particular situation, but it does remind me that exhibition hall interference is a problem that goes way back.
In the late 80s I worked for an exhibition company and we built a huge pond made up like Portsmouth Docks which ran a couple of remote controlled boats that punters could play with, and were meant to take part in a timed race. (Left unsupervised of course, they much preferred the sport of ramming each other until one sank). The first version used standard 27MHz radio control gear and was utterly hopeless - monitoring the channel using a CB proved that there was a positive noise-to-signal ratio. Version 2 ran the same basic gear but used infra-red transmissions from four strategically placed towers at the corners of the pond and an omni-directional pickup on the boats. Worked totally reliably. We never successfully solved the ram-and-sink problem though.
I wrote code for the 802.11b stack, and have gotten a lot of feedback from the test team as well, and here is my 2c:
1. Stacks should handle at least 127 radios on one channel, but most implementations crash with as few as 8 radios alive. Make sure you use a stack that handles many radios. Test your router and your gear (netgear and D-link passed, but check with your current router anyway)
2. Nearby channels appear as noise. If you have many TX on nearby channels, you may not have enough signal/noise ratio. Make sure all gear is MI-MO, and maybe add directed antennas to your router that keep your signal strong in your area.
3. Use channel 14 (you may want to check legality of this in your area) Standard US HW is limited in FW to use ch1-11. Ch 12,13 and 14 is all in virgin territory, and you would be alone at those frequencies, unless of course, you traveled to Spain or Japan or other where you this would look different.
don't cut it off www.mgmbill.org
Quite a few routers will use "other" channels with after market firmware such as DD-WRT.
Set up the public network on 2.4 Ghz and your own network on 5Ghz. Done and done????
HE'S APPLE!
They use PROTOTYPE equipment all the time... He owns the peeps that can set the channel to whatever Steve wants.
It actually makes sense that Steve would set the network up that way.. It would also keep "nuisance" hackers out of his demo. It does make it extra funny to know Steve was using a channel people in the room should not have been able to access... Not a slightly geeky crowd much?
But on topic, short of having firmware access to unlicensed bands, there's not much you can do. Perhaps convention halls should get infrastructure level equipment put in and force everybody to use their channels? That would fix the problem of battling networks kicking each other out. Then YOU, the presenter, would still have to use VPN fir your products to prevent snooping.
... i cannot believe you went throu the trouble of wiring low loss coax instead of going ethernet.....
a power contained cooperative system would obviate ISP's for the last mile, by creating an intelligent mesh
this isn't really a presentation of research, but rather a call for some:
http://nimbusgarden.com/ben/_UWB_/FriendSpace07.pdf
I recommend cooperation and cognition, but disagree that "the real egoistic behavior is always to cooperate," as sophisticated cognition among trusted peers should always be a step ahead of blind cooperation. WiFi is more-or-less straightforward, but the realm of radio is slightly mystical, delightfully so.
MU-MIMO with software UWB antennae, and cooperative channel allocation, sophisticated AI pathfinding, albeit power-constrained for mobiles, is the way of any future where this doesn't become a real problem.
That will block out all the crazies. In this case it needs to be a little bigger than the traditional one. More seriously, I'm guessing that putting your router in an aluminum foil dunce cap above your booth would be sufficient to allow the clients below to connect. Put another router underneath for decoration. You'll have to experiment with how to make it look good but I'm thinking that you could take a cone or hemisphere made with standard lightweight tradeshow construction and cover it with aluminum foil from the grocery store and save yourself the cost of the copper.
Put the router in a grounded metal box with the side facing the stage open. Ensure the people on stage have line of sight while the audience doesn't. This will block the RF from the audience and give you exclusive use.
Even in a crowded environment you might to well to remove the antenna from a router and place it very close to the mobile device (inches). This would reduce the gain at the receiver and allow the nearby mobile device a significant SNR advantage over the rest of the access points in the room.
Linksys, Netgear and D-Link are generally consumer grade equipment (crap)
Try a higher end product like an Engenius ECB-3500 if you desire omnidirectional coverage, or the EOC-2611P for coverage from the sidelines or in a particular direction.
Both models have excellent receivers, and they have up to 600mW of output power. That will punch through a noisy environment. 600mW is 10 times the power that most "consumer" stuff offers, and their receivers are far better.
The ECB-3500 is an indoor model which has two 5dBi antennas with diversity, and the EOC-2611P is an indoor/outdoor model and has a 10dBi panel antenna that is configurable for horizontal or vertical polarization, or diversity between the two.
Either one can be had for under $100.00. I've been using Engenius/Senao gear for a long time and I will say that it is worth three times the money.
Out of order? Fuck! Even in the future nothing works! - Dark Helmet (Rick Moranis) "Spaceballs"
Free tin foil hats for everybody!!!
Did... did you just suggest to re-calibrate the main deflector dish?
While the N specification does indeed enable the use of 5GHz, the vast majority of equipment I've seen can only operate in the 2.4GHz range.
Thus, spend the money, get equipment capable of operating at 5Ghz, and you're probably gold even past the point when the rest of convention goers start doing it, due to the vastly larger signal space.
I don't read AC A human right
Channel 14 isn't that helpful though as it is only allowed for 802.11b which tops out at 11Mbps. Some devices will connect at 54Mbps on channel 14 but most g and n equipment doesn't even list it in the driver.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Sometimes switching to Google's DNS servers at 8.8.8.8 or 8.8.4.4 and/or using a local DNS caching server rather than using whatever DNS server is provided to you via DHCP can solve the problem (because sometimes the heavy traffic is causing greater problems with local DNS server overload than with wireless packet collisions etc.).
I have been wondering lately if switching to using a tiny IP packet size would fix the problem, by minimizing the number of aborted packets due to random wireless collisions. My theory is that if my packets are much smaller than anybody else's, they will have a higher chance of getting through. Has anybody tried this?
I have been wondering if switching to using a tiny IP packet size would fix the problem, by minimizing the number of wireless collisions. My theory is that if my packets are much smaller than anybody else's, they will have a higher chance of getting through. Has anybody tried this?
Two factors matter: Signal power and Noise... resulting in the magical "Signal to Noise Ratio" or SNR.
Increasing your Signal does NOT reduce other people's signals. Other signals show up as increased Noise to you. So you're increasing your individual Signal output, and competing against the Collective Noise of everybody else. And when they turn up their Signal in response to yours, you see it as a massive Noise spike.
Basically what you have going on is a massive DDOS, where everybody participates and targets everybody else.
The only way to solve this problem is either through the use of an isolation system (such as a Farady cage, RF absorbing/dampening materials, etc.), or a reduction in the number of AP's which are transmitting in the area. In other words, it's a reduction in Noise which is necessary, not a Signal increase.
The ideal solution would be for the hosts of the event to provide a single blanket wifi network and disallow AP's at the booths entirely.
Yes, there are also some issues due to the large number of clients and how they are behaving, but that's a completely different issue and there's a lot of ways to mitigate it... but again the best solution is to have a single, well-managed wifi network for the event.
Cisco sells/rents software to allow you to model a conference hall for wireless coverage and number of wireless clients and then adjust power, channels, and max range depending on the crowd moving through. If the provider of your conference hall is worth a damn, they should be doing this for you as part of their price for wifi coverage. If not, well find a university that is already using the software and a good wifi admin to keep coverage working for the lecture halls. This admin can do the modeling and adjustments remotely.
And those of you saying "High gain antenna" "More power", you're causing the problem.
High gain and more power work if you only have a handful of APs and a couple of wireless devices and a lot of area to cover.
In a conference hall, you want the minimize the number per AP and have them use the AP closest to them.
wifi doesnt work with a lot of people on it. Can you say interference? It doesn't help that they want to turn 12 usable channels (b) into 3 (g) into 2 (n) into one (i forget). They can't get appreciable improvement in the bits per hertz so they take more hertz (5 to 20 to 40 now 80 mhz in the latest version). Note that taking all the ISM band won't leave any room for other things like bluetooth, zigbee, microwave popcorn (make the popcorn before starting netflix) or your home phone...
there isn't an answer so this is just a rant.
Copper as a reflector will create multipath problems. It is best to use something designed to absorb the radiation to attenuate it. Visit the microwave cookware booth nearby and offer to display all their browning dishes on your backdrop. This will convert the signal to heat and attenuate it. This works for signals entering your area as well as your spill out to your neighbors. You will want a lot of dishes to fully tile your wall.
The truth shall set you free!