Those of you that are familiar with Spectrum Analysis in general usually expect to see something this bad (high duty cycle) in the 2.4GHz spectrum but not the mid-5GHz spectrum! Having just reloaded my laptop with Windows 7 and installed Service Pack 1, I was in the 'let's test it all' mode to make sure nothing unexpected happens. At this point, I was pretty blindsided by the obnoxious noise happening and the 'Generic - Fixed Frequency' tag wasn't helping me any. At a loss for what this could be since I live an acre away from my nearest neighbor and several miles from the nearest airport, I pinged a few of my friends. They suggested the usual suspects - MRI machine, TDWR, neighbors, etc all of which I explained away by location. Being that TDWR is in the 5470-5725 frequency, I changed my card over to 5.725 - 5.850 and after some time got this equally disturbing read:
At this point, I started to suspect my Spectrum Analyzer since I was using a non-Cisco branded Spectrum Analyzer card with the Cisco Spectrum Expert software (the card I was using had the Cognio components that Cisco purchased and re-branded as their own). So I grabbed a copy of the card manufacturers software to rule out in compatibility and I got the same results.
At the end of the day, I was able to swap in a Cisco branded SA card and my results normalized. Clearly I have a flakey (old) SA card that was giving me improper readings. Lessons learned:
- Always test your tools and keep them in good working order
- Don't assume that your tools are telling you the truth. If you see something suspect, dig into it and validate against another source
Now I'm sure that I have a good card in hand I can go confidently into my week and knock this survey out of the park!
Hi Samuel,
ReplyDeleteGreat wisdom ! Always test your tools. I got the original Cognio card and sometimes I find discrepancies in results between Cognio and another SpecAn brand (name disclosed), leading me to question which one tells the truth. But when VoIP over WLAN calls keeps dropping, I know there is something wrong, even though both SpecAn tools do not classify the culprit the same way. Thanks for sharing.
Thanks for the comment! I find it interesting that our tools give us only their perceived opinion. :) At the end of the day, it's down to the engineer to accurately read (or interpret) what's given to them.
ReplyDelete