02-23-2012 02:15 AM - edited 07-03-2021 09:38 PM
Hi all,
I've tried to test the backhaul throughput using a mesh network consisting of two 1552 APs operating on channel 100. For this I've used iperf to simulate a transmission between one laptop connected to AP10 and a server connected to the same switch as AP09 and the WLC.
During this transmission I've used an USB spectrum analyzer from Metageek in order to observe channel utilization. I could saw a duplicate signal with atenuated amplitude near channel 52. No other wireless networks were in place in 5 GHz band and this results were found using another analyzer, so I'm
sure that it was caused by my transmission.
I don't have too much experience in wireless field so I couldn't find a good explanation for that (I thought that the signal could be a harmonic of the original one or it could be caused by a broken AP). Do someone have a technical explanation for these results?
Thanks/Regards,
Andrei Muresan
02-23-2012 06:04 AM
I don't think its a bad AP... how far do you have these AP apart from one another? Are they already mounted outside when your doing this testing and are you seeing this from AP1, outside inbetween the AP's or near AP2? If you power down these ap's, does that signal at channel 52 go away or stay.
02-23-2012 06:39 AM
It was a lab setup so both AP were placed in the same room (around 3m between them) and the spectrum analyzer was running on the laptop connected to the MAP via GB port (the 2.4 radio was down). During these tests I also want to observe some throughput limitations caused by signal saturation (this explains the high SNR and signal power). In the picture above for a signal level around -45 dBm the parasite one had an amplitude around -75 dBm. Running same tests but using attenuators for antennas I obtained a parasite signal around -80 dBm for a signal level of - 56 dBm. The parasite signal was present exactly when I was using iperf to transmit. If no transmission was in place the entire 5 GHz spectrum was quite.
02-23-2012 06:52 AM
Well it could just be something from the laptop, but not sure. The hard part, is that you don't know if its from your laptop or the AP. Did you disable your wireless NIC by chance to see? How much TX power did you set on the AP and what attenuation did you use for the antenna?
02-23-2012 07:22 AM
I didn't tried to disable my wireless NIC because, until now, I was almost sure that the signal was caused by the AP. The TX power level was set to 1 and the attenuators which I've used had a 20 dB attenuation. I will try to reload the test maybe I will get more information about this. Do you believe that this issue can be explained mathematically using signal theory?
02-23-2012 07:29 AM
I don't know the answer to that... Maybe someone with a military rf background can explain that. They usually are very knowledgeable with various types of rf. as far as your AP's are concerned. There is alway overlap to the adjacent channels, but in the 5ghz, the channels are separated enough to not overlap. A radio can be set to use various channel width. The Cisco Mesh can do 20mhz and the 1550 can do 20mhz or 40mhz. The channel 52 is far away from channel 100 you are using and I don't think that the AP is generating that. It would be nice to see how your test goes when the AP's are mounted and you don't have the attenuators on anymore.
Thanks,
Scott Fella
Sent from my iPhone
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide