Customer is using 10 Zyxel AP in 5000 square meter warehouse and they complain that MC3190, using terminal client application, roam gratuitously even if the signal strength is strong. They are using channels 1,5,9,13 . This is causing them disconnection with the server so MC3190 have to associate much more often. Maybe the Zyxel cannot handle properly the roaming algorithm? If so I have to prove it to customer that our device is working propoerly and the problem is in Zyxe. Any ideas are welcome. Thanks, Roland
Zyxel AP roaming issues// Expert user has replied. |
2 Replies
I have seen this before and with the information you provided it looks like they have several issues. The first being co-channel interference, the following channels will interfer with each other, 1/5, 5/9 and 9/13, they should change the channels to 1,6 and 11. The second issue is the number of APs at the site, 5,000 square meters is about 16,000 square feet and that small of an area can be covered with just two APs and even provide redunancy. They will need to rev the power way down to close to 1mW or shut down some of the APs. What is happening is that there is so much RF in the facility that the mobile units are seeing too many APs and can not roam properly.
Hi Fred, I'm not sure I entirely agree with some of the suggestions you've made here, as they seem to be a bit presumptuous. You cannot assume the need to drastically change the tx power output and/or the number of AP’s to deploy without first knowing the physical dynamics of the site. Also, the channels used (1, 5, 9, 13) appear to be for a European regulatory, so assuming this, keep them as is. Hi Roland, The key concern I have is: Are the MC3190’s newly introduced to this preexisting WLAN?
If the MC3190's are newly integrated whereas preexisting mobile's have been roaming with minimal concern/disruption to session flow within the same unchanged WLAN, then I would first suggest you perform a side-by-side comparison between the differing mobile's while walking about each cell to observe the individual signal detection levels, association count & roaming behavior of each.
In my experience (current sites I'm working with), the MC3190's tend to roam a bit more often due to their lower receive sensitivity & output power capabilities comparably to other Fusion (Photon) based (& competitor) terminals, assuming that is what’s also deployed.
If after this compare test you find a substantial difference in the SNR readings + a vast increase to the association count on the MC3190 when co-operating in same cells as other mobile's, then I would suspect a need to re-survey the site (specific to those cells) against the NEW weakest integrated mobile (the MC3190).