Self Interference


   We have a single tower setup with four(cardinal directions) 90 degree ePMP 1000 sectors and Connectorized GPS radios (ver 2.3.3). We have matched every setting for our East and West radios with the exception of Front/Back, which they obviously differ per best practice. The issue that we have is that when the West radio is configured in the same frequency has the East radio, the SM to the East will lose connectvity, and not regain sync until the West radio is taken out of the frequency. The SM we are using is a Connectorized radio with a Force 110. This is our first test customer using the GPS sync feature.

   It should be noted that we are in a relatively high interefernce spectrum (-64db), but when we have link, the down and uplink MCS values are between 11 and 13.  RSSI is -67db, with a SNR margin of -30db. Am I correct in assuming that the SNR margin is just a value calculated from subtracting RSSI from a clean frequency?

   Any help is appreciated, Thanks!

Hello Cory,

Regarding your question, the SNR is calculated by taking the carrier received level (-67dBm in your case) and subtracting that from the noise floor measured by the radio. The SNR calculation does not take into account interference.

Several follow up questions:

When the SM fails to connect, have you noticed the reason cause displayed in the Monitor->Wireless page of the SM?

At the same page in the SM, what are the signal levels displayed for the listed APs, as seen by the SM?

Do you have a Preferred AP or APs configured in the SM?



Thanks for the reply Luis,

   The reason is typically "Failed to Connect - No GPFs" when the SM can't connect. We have ONLY the intended AP set in the list of preferred APs. The other AP has a SNR of 20db with RSSI in the -72db range, considerably less margin and RSSI than our intended AP. I've read that the No GPFs is typically due to interference preventing the management packets from transmitting, but correct me if I'm wrong.

   We also have an issue of 50/50, and 30/70 ratios not providing a reliable link as 75/25 does.

How far away is the SM from the site, and how high are the APs?  If the client is reasonably close to the APs (especially if off-axis) it's entirely possible for the SM to see both APs at the same time.  Since they're on the same frequency, that means the best SNR you could hope for would be the difference between the two signals from the SM's perspective. (which might be 3dB for a customer a few hundred yards away)  If the customer is farther out, or structural facts help block the opposite AP from their view (like APs mounted on railing of a water tank) then it shouldn't be a concern.

If you are seeing interference at -64dB then I can't imagine you'll get a connection to work at -67dB.  (I'm guessing that -67 is downlink RSSI reported at SM, and -64 is 'uplink' noise floor seen at the AP?)  I wouldn't even try to deploy a PMP sector on a channel with noise floor stronger than -85 to -90dB.

With good GPS sync going then both APs will transmit at the same time, and receive at the same time - this keeps one AP from crushing the other by transmitting when the other is listening to SMs.  But an SM (or far end PtP, whichever) that can see both APs will not be able to separate their signal, they'll be shouting each other down.

Also, are you using the built-in GPS, internal (self-generated) sync or external (CMM3/CMM4/SyncPipe/etc) for your sync?


We are using the built in Cambium GPS sync feature with the external GPS antenna included. The APs are about 230' high, and this SM is 2.4miles away at about 35' LOS. To the best of my knowledge, the SM is aimed appropriately. 

Your suspecions about the noise, and RSSI are correct. The AP is seeing a noise floor of -64db and the SM is reporting a -67 RSSI to the AP. We are having pretty good success with the RSSI being -67db without the back sector transmitting in the same frequency. Wireless link tests in a 10MHz channel are coming back with 27M/4M. As soon as the back sector is put in the same frequency, the link drops and will not reconnect.

  Unfortunately a clean channel that you recommend isn't possible in this suburban area that we are. In 5.1/5.2 and 5.7/5.8, -64db is the cleanest that we've found with the spectrum analyzer.

   I've included a picture of our tower, which is on a rooftop.

Hello Cory,

Yes, the No GPF error is  due to SM failing to demodulate management packets transmitted by the AP. Normally these messages are transmitted using MCS1, but there is an option in the AP Configuration->Radio page to force them to MCS0 for noisy environments. Have you tried that already?



   I sure have. I still get the same results. When the back sector is configured in the same frequency it does not get link, but when it is moved to another frequency, link is established with great MCS values in a 75/25 ratio. When I change to any other ratio, the link isn't as reliable and bandwidth is cut down considerably.

Hi. This thread died out... I'm currious if this was this resolved?


I'm still working with Cory on this. His site is located in a very congested RF environment and this could to be causing his problems.



2 Cory Ditzel, 

Make sure, that all AP sectors  in  ABAB frequency scheme have the same DL/UL ratio, and setting equal max distance  to CPEs.

Did you notice , that  "Failed to Connect - No GPFs" messages appear at CPEs, that  locate  next  to ( up to 500-600 meters distance )  AP?

Try to increase setting "max distance"  at all APs up to 65 km  ( 40 miles).

This is interesting observation. I can confirm, that this happen time to time on links with this distance even if the signal seem to be okay. Never tried to increase the Max range. Will do next time ;-)

I'm working with Luis on testing some theories with my setup. At this point, it looks like that other Cambium gear in the area might be the culiprit here. When we get this fixed or figured out, I will share my experience.