I have WAPs deployed 3000 sq feet apart, channels 1, 6 11 inh use, with 7-9 WAPS per floor some channels touch, how should roaming work?
I have inherited the following wireless infrastructure: Cisco WDS control units and 1131APs. Using 802.11b and g. 7 - 9 WAPs per floor. Configured for channels 1,6,11. Had issues with roaming using authentication with VPN tunnels. Moving to WPA2 and Cisco ACS. 7-9 WAPS per floor. Some same channels do overlap with minimum signal strength. With three channels to pick from it is unavoidable? I am confused as to how roaming should work. It seems to me that WAPs with same channels would enable better roaming but there would be channel to channel interference (bad thing, so I am told). Roaming between channels should cause for some delays and therefore greater chance of 'drops'. I am also aware that companies like Meru Networks use only one channel. Can someone explain these channel issues in regards to roaming, channel contention, and delays?
Thanks
Thanks
ASKER
John,
Thanks for the link. I was not able to find this during prior to my post.
To be honest, the thread does not provide an absolute answer to my questions. Perhaps there is no absolute.
Especially when reviewing the testing performed later in the thread (i.e. test1, test2, test3). the tester indicates his best results were obtained when all the access points were on the same channel. Why he did not have channel interference is a mystery although I wonder how vendors such as Meru Networks handle this issue. I know they fool with the timeout value but do not know how they handle channel interference. The comment about WAP Meshing is also interesting in that again one channel was referenced.
So, Am I correct in stating that alternating channels (i.e. for 802.11b and g channels 1, 6, 11) is the 'best practice'?
I know that WDS will auto config this way.
Thanks for the link. I was not able to find this during prior to my post.
To be honest, the thread does not provide an absolute answer to my questions. Perhaps there is no absolute.
Especially when reviewing the testing performed later in the thread (i.e. test1, test2, test3). the tester indicates his best results were obtained when all the access points were on the same channel. Why he did not have channel interference is a mystery although I wonder how vendors such as Meru Networks handle this issue. I know they fool with the timeout value but do not know how they handle channel interference. The comment about WAP Meshing is also interesting in that again one channel was referenced.
So, Am I correct in stating that alternating channels (i.e. for 802.11b and g channels 1, 6, 11) is the 'best practice'?
I know that WDS will auto config this way.
To separate channels to have NO overlapping, the need to be 5 channels apart, meaning 1,6,11.
However, I've seen several environments that are using 1,4,8,11 so they have 4 channels to work with, and there is minimal interference.
However, I've seen several environments that are using 1,4,8,11 so they have 4 channels to work with, and there is minimal interference.
In my tests and on the same channel for all APs, distance was one of the factors whereby there was minimal overlap. But it worked just fine. A caveat...
Now, some APs handle overlap a lot better than others. I know nothing about your make AP.
And, yes, in almost all events, a five channel separation is best practices and preferred. Your APs may have an auto channel select. Mine do and it worked very well in the mixed channel mode so try that.
If you have the opportunity to experiment with a group of your APs in the real worl at your faclility, do so and gather some real world facts in your situation and with your APs.
John
Now, some APs handle overlap a lot better than others. I know nothing about your make AP.
And, yes, in almost all events, a five channel separation is best practices and preferred. Your APs may have an auto channel select. Mine do and it worked very well in the mixed channel mode so try that.
If you have the opportunity to experiment with a group of your APs in the real worl at your faclility, do so and gather some real world facts in your situation and with your APs.
John
PS.. In auto config/channel select, your APs may not select the perfect 1, 5 , 11 scenario. It will look at what channels are out there and configure themselves, "generally" for best performance.
John
John
ASKER
Thanks for the input. How many of you have given thought to moving data access to 802.11a to mitigate RF interferance at the 2.4Ghz band and reap the possibilities of greater channel selection. I beleive there ar 8 usable channels with no overlap.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
802.11a would require new hardware, unless your hardware is already a/g compatible. It operates in the 5.4 ghz range. None of the channels overlap, so you can use them all.
However, because of the higher frequency, range tends to be shorter. You should throughly evaluate 802.11a signals in your network before committing to a roll-out project.
Most equipment that supports 802.11a also supports b and/or g. This would allow you to replace your equipment and double your available bandwidth, by putting half the load on a, and half on b/g.
However, because of the higher frequency, range tends to be shorter. You should throughly evaluate 802.11a signals in your network before committing to a roll-out project.
Most equipment that supports 802.11a also supports b and/or g. This would allow you to replace your equipment and double your available bandwidth, by putting half the load on a, and half on b/g.
ASKER
Using Cisco 1131APs with WDS control and two radios, one for 802.11b/g and one for 802.11a.
802.11a roadio already enabled and broadcasting.
The standard that I inherited calls for each access point to cover 3000 sq ft, or approximately 55 feet apart.
I believe this would meet 802.11a standards; however, a survey would have to conducted or a least a walk around with the 802.11a client connected. I was considering using a recursive ping utility to check for latency and dropped packets. I also have NetStumbler. My operations team has a more robust product whose name escapes me (expensive)
I am in a hospital enviornment with a lot of RF interference. My plan is to move voice (Vocera and VoWIFI) to either 802.11b or g and guest access to what is left (probally 802.11g).
802.11a roadio already enabled and broadcasting.
The standard that I inherited calls for each access point to cover 3000 sq ft, or approximately 55 feet apart.
I believe this would meet 802.11a standards; however, a survey would have to conducted or a least a walk around with the 802.11a client connected. I was considering using a recursive ping utility to check for latency and dropped packets. I also have NetStumbler. My operations team has a more robust product whose name escapes me (expensive)
I am in a hospital enviornment with a lot of RF interference. My plan is to move voice (Vocera and VoWIFI) to either 802.11b or g and guest access to what is left (probally 802.11g).
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
To comment on the Meru Solution mentioned:
Meru gets by the Co-Channel interference problem by using the WLAN controller to coordinate AP's that would otherwise interfere with each other. Basically, the controller is made aware of what AP's can "see" each other, and makes sure that they do not transmit simultaneously. Meru has more information about its technology, and how it eliminates this, and many more traditional WLAN deployment problems on their website: www.merunetworks.com
Meru gets by the Co-Channel interference problem by using the WLAN controller to coordinate AP's that would otherwise interfere with each other. Basically, the controller is made aware of what AP's can "see" each other, and makes sure that they do not transmit simultaneously. Meru has more information about its technology, and how it eliminates this, and many more traditional WLAN deployment problems on their website: www.merunetworks.com
https://www.experts-exchange.com/questions/22963288/Wifi-in-public-areas.html
John