You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more
Hey guys.. Any networking/telephony guys in the house.?
I'm struggling to get Cisco wireless phones (8821) running in our environment.
Very large and noisy environments, ducts, shelving units, machines, forklifts, the lot.
We have undertaken a wifi survey from an external company to find we have full coverage, good as we spent a fortune installing antennas and Cisco LAPs to flood the area.
But these Cisco 8821 phones are still very unreliable. The survey has come back good, but looking on our wifi controller the phones are still between 90 and 70 -dbm. Bad and contradicts the survey.
We've upgraded controllers, phones, configured all the Cisco best practices, logged Cisco TAC cases, but we still struggle to find an acceptable level of service.
And the business leads are kicking ass over it, quite rightly so.
Move back to DECT?
Has anyone got experience of this type of thing?
Probably not much help. But the antennas (as far as I am aware) are designed to radiate out from the unit - ie you ceiling mount them and they radiate down from there. Are you mounting them this way (ie the "standard" way)?
You could use your mobile and a wifi app that shows you signal strength to see where your signal is good and where it is bad - that might help explain things?
Was the wireless survey a full spectrum analysis?
As above, make sure survey was for 2.4 and 5 GHz frequencies. Assume youve looked into non-ovelapping channels? Any machinery running in the 2.4Ghz spectrum? Are the AP's covered or blocked? Was the survey done in real world conditions with machinary on etc?
What sort of a survey? I have seen people run a survey by filling out a spreadsheet of the signal % seen once on a cheap external NIC. Ideally you want a full survey using a professional grade tool such as Ekahau including spectrum analysis (as per the comment above).
However, that is only part of the story, you then need someone to review and 'translate' it.
The end device is also potentially running on a compromised antenna snuggly sat in your hand - hence the -90 to -70 reading on your handsets (I assume this is dB, but might be a 'percentage' which is less helpful).
What was the network designed to give you and what was the survey results measured against?
FWIW - We often survey to -65dB RSSI with a SNR target of 25dB. Even then thats does not tell the full story.
The LAP's and antennas are all mounted correct, and in areas that an original survey suggested.
loads of Cisco 3802E with 2544 omni antennas. Connected to a Cisco 3504 controller.
We then organised a 2nd survey to confirm we were covered.
We are awaiting the full typed up survey results but from their initial findings, they found nothing of concern. Both 2.4 and 5 and 'full'. And during production hours so a fully working plant. LAP's are all situated correctly and installed properly.
They found a few LAP's (lightweight access points) that were competing on channels but these arnt in the critical areas so although not perfect, they wont be contributing to the problem elswhere.
I'll dig into the full survey write up when I get it in the next few days.
It's of concern that they say (pending full survey results) nothing was notably wrong with the coverage and spectrum, but when checking our controller, most of the WiFi phones are showing way over -70dbm, 10-15 of them showing above -80 and into the -90's!
Cisco TAC cases, software upgrades on the controllers and phones, config best practices, software config adjusted as per Cisco's tech advice, and coverage looks good....
I'm starting to doubt that the Cisco 8821 handsets are not fit for our needs.
(anyone in the market for 160 Cisco 8821 handsets!?? 😉 )
It maybe that the handsets are not roaming quickly enough. Try removing some of the lower data rates to force devices to roam quicker.
Did you get any data on how many AP's can be seen overlapping with the required RSSI?
Something like the heat map shown here In the Number of Apps section
Yes, we are hoping to get the heatmaps when we get the full written report in the next few days. That should help us with the bigger picture regarding overlaps.
If they used Ekahau for the survey, see if they will give you the survey file (.esx)
It’s easy to do the data collection part of a survey, but it’s much harder to interpret the results. In addition, it easy to leave the settings on ‘default’ which can make the heatmaps look way better than real life.
BentandBroken, lets see what the survey brings up.
Ive got the survey in paper form, and it was done in Ekahau. It's in French... [facepalm]
Which is fair enough as it's a French factory and a French survey company.
It's over 400 pages so will take a while to digest.. I've asked for the .esx
Can you post an image of one of the RSSI pages with the scale on it and also the 'walked paths' page for the same area?
Ideally a part of the building that is not a perfect square/rectangle..... PM me if you would prefer