Frequency Re-Use on 4600s

Products ePMP
1

I just implemented Frequency ReUse on two 4600s last night on one site. We were getting some severe interference from something (maybe another ISP using 6 Ghz) on AP3 on the site and so I decided to try putting that AP on the same frequencies as AP1 (AP3 is pointing at 180 deg. and AP1 is at 0 deg).

To say that I’m impressed is an understatement. The equipment that we’re replacing was supposed to have this capability, but it never has worked well, to the point that frequency re-use with it was unusable on any of our sites.

The Cambium is absolutely amazing. There doesn’t appear to be any interference of any kind between the APs that are on the same frequencies on this site. And this is on a tower where the APs are only around 4 to 5 feet apart.

I just enabled Frequency ReUse on AP2 and AP4 and the results look great.

I watched for any noticable change in RSSI numbers on any of the APs and any noticable change in DS numbers. Everything looks pretty much the same with Frequency ReUse enabled as compared to not enabled. And throughput tests to a few customers look good.

With all of the above said, I tried implementing Frequency ReUse on another site and I’m not getting results as good as on the first site. This second site has quite a few clients that are very close to the site, some within a few hundered feet of the APs and underneath them (the APs are at about 140 ft above ground).

This brings up a question - if an SM is so close to a pair of APs on the same frequencies, then can there still be interference from that SM or to that SM?

John Rayfield, Jr.
Rayfield Communications
Springfield, MO

4 Likes
2

On the AP’s, what do you have the Subscriber Module Target Receive Level set to? By default it should be -60, but if you have clients that are very close with hot signals, you may want to decrease that value to -65. On the SM’s, if you find that they’re able to perform an AP scan and they’re able to hear the AP pointed in the opposite direction at strong levels then you may want to check out the following items:

ALSO, what antennas are you using? Again, if you have AP’s collocated on the tower very close to one another and/or clients very close to the tower, you want to make sure you use AP antennas with excellent front to back ratio design… that or you need to physically separate the collocated antennas adding either horizontal or vertical distance between them.

3 Likes
4

yeah, physics.

Executive summary : physically separate your APs as much as you possibly can, the same applies to antennas. Provide any shielding between them.

Few hundred feet really is too close. Decrease TX Power from clients where necessary - you only need to achieve certain level, more is not better in this case. -60dBm will not provide you maximum throughput, but depending on your speed options, it might suffice (-60dBm and corresponding high modulation is able to transfer around 500Mbit/s, that is combined physical rate for all clients together, actual AP throughput will be around 80% of that).

What I want to say : check any relocation options for AP. Just moving away one AP from another AP to different mast / tower and separating them by 20 ft (they should not be back-to-back) makes big difference. If you can hide AP+antenna “around a corner”, that will be a miracle.

Remember, these devices themselves leak signal quite a lot. Putting each AP in a faraday cage, which is very small, cheap and installable everywhere, should make tremendous difference when reusing the same frequency. Especially if APs are next to each other or back-to-back. Don’t use solid “cages”, you need cooling for your device, it should be wired mesh : https://mountainhouse.com/cdn/shop/articles/diy-faraday-cage-featured-image_35df0fcf-597f-4ba2-b74b-f14c6b324e01_1024x.jpg?v=1709066767

Antennas : the best front-to-back ratio is your friend, but still - any antenna will “leak” signal, sidelobes and backwards. Few hundred feet is the problem here : I would bet your closest clients can hear the opposite-facing AP almost as well as their dedicated AP, depending on your antennas and their direction the difference might be 10dB only. Also, installing antennas on extended arm / standoff and separating them by three meters (9 ft something) should provide additional 56dBs [!] of signal separation between them.

At three locations, where allowed, we’ve built a small single-brick wide wall and separated APs that way :slight_smile:

Maybe the last hack : put any grounded sheet of metal between APs and between antennas. Secure it perfectly as they tend to fly off. Simple, cheap, light, quick and … depending on conditions, extremely effective.

5

Sync is amazing, we have been using it since the old Canopy days and felt the pain of not having it when Ubiquiti first came out and they didn’t have it and then ePMP came along and we immediately started replacing Ubiquiti with ePMP and GPS Sync was 90% of the reason.

You are correct though, Sync only helps the AP’s and by helps I mean makes them invisible to each other. It does nothing to stop client radios from interfering with AP’s or other client radios on other APs or AP’s interfering with the client radios of other AP’s. Good front-to-back and , in places where you don’t have a lot of local interference you have to overcome at the AP, reducing / controlling your clients TX are your best tools.

I personally think DiY and most/all aftermarket “shielding” is usually bad.
I don’t think a lot of people realize that metal “shielding” to RF is a mirror. Metal “absorbs” a negligible amount of rf, most of it either passes through or gets reflected. It would be like if you had big light bulbs on a tower and lots of spot lights on the ground shining back at them. If you just took some big flat double sided mirrors and stuck them behind each light bulb.. well yeah, now they aren’t seeing much from the other lights behind them but now you are literally reflecting light in every possible direction, the APs/Lights are blasting themselves in the back, their own clients are hitting them , bouncing off the “shield” behind them and hitting them again. You spent all that money on a sector that radiates power in a very specific way then stuck mirrors up there and created multipath reflection, refraction, diffraction, scattering , attenuation disco ball.

Waveguide/Horns seem pretty great for back to back channel reuse. I wonder how much it affects front-to-back when the horn is downtilted 2° since that uptilts the back lobe 2° . Most sectors today have electronic downtilt and I believe the way they do that also causes the back lobes have downtilt.