Voice over Wireless LAN Solution Guide v1.0 December 2005
______________________________________________________________________________________________________
Page 28
other words, the user device will not be aware of any problem with the WLAN (other than that
traffic flow has stopped) until the AP 2330 detects the WSS 2300 failure.
After the detection, the AP resets and associates to a surviving WSS 2300 according to bias
settings. The entire time for the AP to become operational again from the point of WSS 2300
failure can be anywhere from 45 seconds to a minute. Any existing calls on an AP 2330 will be
impacted. If a neighboring AP is associated to a working WSS 2300, then the handset will attempt
to roam to it when the current AP resets. However, this roam will very likely take place after the
call has been lost due to the lengthy interruption.
When the WSS 2300 is repaired, the APs do not automatically return to its high bias switch. This
minimizes disruption in the network until a convenient time of the administrators’ choosing, when
APs are reset and returned to their proper WSS 2300.
Figure 12: Poor redundancy planning example
Because of the previous scenarios, it is critical when designing a fault-tolerant WLAN to
intersperse APs among multiple WSS 2300s. If a network is deployed as in Figure 12 (Pink =
WSS 2300 No. 1-controlled APs, Blue = WSS 2300 No. 2-controlled APs) such that large clusters
of APs belong to the same WSS 2300s, then an outage will leave a very large hole that cannot be
covered by either devices roaming to neighboring APs or by auto-tuning increasing power of
neighboring APs.
Figure 13: Better redundancy plan
Figure 13 illustrates a better design that yields a much higher number of devices with alternate
APs to which to roam. Note that a third WSS 2300 (yellow) is added to the pattern. This is