Cambium Epmp 4500c 8x8 mu mimo

Hi, I bought an AP 4500C. I mounted two pairs of 4x4 RF element horns, pointing in the same direction, so that they work in 8x8 (I think I understood in the radio configuration and I selected sector). What conditions must be created to activate the MU MIMO? I have 7 clients connected with the Cambium 4525L, each of them exchanges very well from 8ds onwards. Can you explain this situation to me better? Are there too few clients for this reason I don’t see it active? Has anyone had experience with it? Thanks a lot.

Are you using the newest firmware 5.10.3? Are you using TDD fixed or flexible mode? Did you enable MU-MIMO?

Are your SM’s spread across the entire azimuth? Have you tried running a multiple SM link test to those SM’s that are on opposite sides of coverage area?

I’m using the flexible frame. Maybe this could be the problem. I can try doing some simultaneous tests with clients at different angles. But how much different should the angles be?

MU-MIMO works with TDD-flexible mode, so that shouldn’t be an issue.

I personally have not had good luck getting any RF elements horns and MU-MIMO to work properly and I don’t think I’ve seen any MU-MIMO success stories in these forums using the 4x4 horn array. I’d hit up RF Elements for support on how their horns need to be arrange for best MU-MIMO performance.

No, because I read that with the flexible TDD frame there could be a problem with MU-MIMO activation. I’ll do some tests now, and it’s likely that it won’t activate because the customers aren’t yet in the beam grouped toward one direction. I don’t know.

Is there anyone in this forum who has managed to get these RF elements to work properly? AH60, 4x4

Not quite the exact same but I did get MU-MIMO to work with 4500c and RFE horns. It was in an omni setup w/4x90deg horns, but it did work. I am pulling it down though due to some things I did not like about that type of setup though.

4500c Omni Setup - Weird Signals and MU-MIMO - Products / ePMP - Cambium Community

Can you show me where you found this documented?
Last I read, and my news could be outdated I’m sure, MU-MIMO only worked with the U/D ratio in fixed mode.

I’m currently testing sector mode. I’m using flex mode. There are currently 9 sm transmitters connected. I’ll let you know if the MU-MIMO activates. The two 4x4 RF element pairs point in the same direction.

While MU-MIMO does work for flexible mode, when it was released in 5.10, its status was that of “Performance optimization is pending for Flexible TDD ratio configurations.” I’m not sure how much has been optimized since then in the newest release, 5.10.3. Eventually it will be fully optimized as it was with e3k. There are a bunch of MU-MIMO performance optimizations planned for the upcoming 5.11 firmware release. You can read about this and more in the official firmware new feature release post:

That being said, I don’t think you’ll be able to see MU-MIMO gains using the RFE horn antenna arrays. When we tried using the RFE 4x4 horn array using a ePMP 3000, we found that it did not work well at all. You can read more about my experience HERE.

You could temporarily try TDD fixed mode to see if it makes a difference in triggering MU-MIMO.

Thanks for your reply.
TDD fixed mode triggers MU-MIMO for me in the tens of megabits, while flexible sees a single megabit’s worth of activity.
This is a 4500 on 5.10.3 with 20 subscribers, about 50Mbps of ambient traffic, up to bursts of 200Mbps, with great signals and modulation, 40Mhz channel width.

Well, one of the things to remember about MU-MIMO is that it’ll only be triggered if link needs it to deliver the requested data. So for example, if you are set on 75% / 25% in 80 megahertz, you may see very little MU-MIMO because all the data requested maybe serviceable in SU mode. If you change your settings to 50% / 50% in 40 megahertz, the MU-MIMO may jump up dramatically in order to deliver the requested data.

So likewise, going into flexible frame may perhaps lower the need for MU-MIMO as the link may be able to deliver the data in SU mode with a flexible duty cycle without needing to envoke MU-MIMO. .

2 Likes

Yeah you’re 100% right. Thanks for noting my oversight.

1 Like