802.11a Wireless Bridge - DFS
Last Post: January 21, 2010:
I've installed a wireless bridge at a customer site in the UK using 802.11a band B (channels 100-140). All of these channels are subject to DFS. i have run the wireless link on channels 100, 116 and 140 and all have detected radar and shutdown the radio for 30minutes making the links unusable.
I seem to have run out of options. I've read Band C UNII channels can be used outdoors in the UK providing the a license is obtained, but the Trapeze Networks kit I'm using does not allow these channels on a GB countrycode - so maybe i've been misinformed.
Can anyone shed some light on how they have configured outdoor links on 802.11a?
What is the distance of the link and what antennas are you using?
It's around 50m, just bridging between two school buildings.
What gain are the antennas?
The reason I ask is this:
I don't know if this would work, but if you use high gain antennas there is a less likely chance that the bridges will detect radar. Unfortunately I can't guarantee that it will work.
A couple of suggestions:
1. I???¡é?¡é?????¡é???¡éd suggest you try and find out if it really is radar that is the problem and not just general inteference. You can contact the following:
There are others, but that is a good start. Are you near any airports, military bases etc ?
2. If you can get hold of a spectrum analyzer, try to do a sweep to pinpoint the direction the signals are coming from. You may find that the interference might not be radar, but could be just general interference from another system. Sometimes you can get lucky and find the people responsible and have them help you out. I???¡é?¡é?????¡é???¡éve done this a few times with radio systems. Radars have a very distinctive pattern on a spectrum analyzer.
3. There is a slim possibility that one of the radios may be faulty. Try swapping the radios and see if you get the same results. Have you tried the complete system anywhere away from your current location [ in a lab, car park etc ] ?
4. From the results of your spectrum analyzer sweep, you may be able to determine another frequency range that is clean and talk to Trapeze to see if they can help you out [ don???¡é?¡é?????¡é???¡ét know if the kit is used or still in warranty ]. Manufacturers will often try to help out customers by doing a swap in order to keep future business [ not all though !! ]. You maybe able to go to 2.4 Ghz.
5. GT has a good suggestion. The higher the gain of an antenna, the narrower the beamwidth. That is why it is important that you know the exact direction the interference is coming in on. In some antennas with poor sidelobe performance, interference can be picked up ???¡é?¡é?????¡?¡°off-axis???¡é?¡é?????¡é???¡é provided the inteference is coming in at the ???¡é?¡é?????¡?¡°correct angle???¡é?¡é????????. Also, even with antennas with good sidelobe performance, if the power is blasting in, you can still pick up bad stuff.
Further to item 3: there is a complex algorithm that the radios use to detect whether radar energy is present in the range being checked. This algorithm cannot detect every type of radar [ loads of different pulse shapes ]. Simple firmware corruption can cause problems. It is possible that it is not radar at all, but something else. Radar would usually cause problems only in one direction of the link [ excluding the possibility that a reflecting object is close enough to "bounce" that initial pulse back along the same direction - moving targets usually only cause a momentary blip. When I was working in the Falkland Islands, an F-4 would come in at 50 feet above the water each day and cause a blip in our satellite link ]. You need to establish which direction the interference is coming from.
And to add to Dave's thoughts, DFS requirements are strict enough that vendors are extra cautious about detecting radar... so it is quite likely that non-radar RF is causing false positives. Double check your firmware to make sure it's the latest. Hopefully all vendors are working to improve radar detection and avoid false positives, otherwise UNII-2 and 2e aren't much use to us.
Thanks for the replys.
I did wonder if the radios were 100% accurate at distingushing radar from general rf noise. I received a 5Ghz Wi-Spy kit today so I'll check the spectrum and see if there's anything obvious.
Some general info on Radar. Note how the frequency bands [ C,X etc ] were named [ very few publications explain this ].
That's an interesting read. I was confused what all this K band and X band was all about.
Am I right in saying radar is a transient thing, i.e. a satellite can pass overhead every now and again thereby triggering the DFS algorithms in wifi access points. Or is it the case that if radar is present in a particular location it's always present.