Cisco Cisco Aironet 3500p Access Point 백서
Copyright © 2010 Miercom
Cisco CleanAir Competitive
Page 15
Self-Healing
Given the negative impact of non-Wi-Fi interference on a wireless network, access points need
to avoid this interference to protect the end user Quality of Experience (QoE). We conducted
this test using the 2.4GHz band.
to avoid this interference to protect the end user Quality of Experience (QoE). We conducted
this test using the 2.4GHz band.
Cisco Equipment:
With the camera enabled in Location A, all five clients immediately lost ping. The access point
switched from Channel 1 to Channel 6, and clients recovered ping in 49 seconds. When the
camera was engaged at Location B, the AP took 39 seconds to change channel and client to
recover ping. With the camera in Location C, the access point took 1:04 to change channel
and recover ping. As the Cisco AP has persistent avoidance, we reset the access point between
tests to clear it so that alternate channels would not get locked out by the feature. In normal
operations, persistent device avoidance automatically ages out the interference source to make
the channel available to the system once again. A second run at each location took 30 seconds
at Location A, 41 seconds at Location B, and 48 seconds at Location C. As expected at the 100
feet location, only the farthest client was failing ping. Although video quality was impacted on all
clients, the access point detected interference and changed channels.
switched from Channel 1 to Channel 6, and clients recovered ping in 49 seconds. When the
camera was engaged at Location B, the AP took 39 seconds to change channel and client to
recover ping. With the camera in Location C, the access point took 1:04 to change channel
and recover ping. As the Cisco AP has persistent avoidance, we reset the access point between
tests to clear it so that alternate channels would not get locked out by the feature. In normal
operations, persistent device avoidance automatically ages out the interference source to make
the channel available to the system once again. A second run at each location took 30 seconds
at Location A, 41 seconds at Location B, and 48 seconds at Location C. As expected at the 100
feet location, only the farthest client was failing ping. Although video quality was impacted on all
clients, the access point detected interference and changed channels.
Aruba Equipment:
The same test was run on the Aruba AP125. With the camera in Location A, Aruba reported a
noise level of -87dBm while a spectrum analyzer reported the noise level at -52dBm. As the
channel was completely jammed, no errors were reported. Since the noise level and error
thresholds were not crossed, the access point did not change channels and all clients were
disconnected.
noise level of -87dBm while a spectrum analyzer reported the noise level at -52dBm. As the
channel was completely jammed, no errors were reported. Since the noise level and error
thresholds were not crossed, the access point did not change channels and all clients were
disconnected.
With the camera in Location B, clients far from the access point were affected and near clients
were not affected due to the signal-to-noise ratio. The noise level threshold was triggered, and
the access point changed channels in 2:01 minutes.
were not affected due to the signal-to-noise ratio. The noise level threshold was triggered, and
the access point changed channels in 2:01 minutes.
At 100 feet, the noise level read -75 to -77dBm and was not high enough to trigger. Clients far
from the AP were affected most, and high latency and degraded bandwidth was experienced for
from the AP were affected most, and high latency and degraded bandwidth was experienced for