The firmware should be available in the next few working days. It’s currently going through cnMaestro’s acceptance testing process, which ensures quality but can take a bit of extra time. We appreciate your understanding.
Hi Andrii, are there steps being taken in regards to the lower mu-mimo performance that some users are reporting? I know the current work around is a reboot but has this been seen and replicated in the lab? Also reports of lower speed test results compared to link test results between 4500 and 4525. Thank you for your time!
This release includes just one fix, but we’re continuously improving the platform.
Can you clarify what you mean by MU-MIMO performance with a reboot workaround?
Some users have seen uplink MCS fluctuations where a reboot or config change restores ePMP 4500 performance. It does impact throughput, but it’s not directly tied to MU-MIMO. If you’re seeing the same thing, I’d like to review your setup — we haven’t been able to reproduce it in our lab, and it hasn’t been reported on the ePMP 4600.
As for speed tests, that’s a separate topic. It’s less about radio performance and more about how TCP handles microbursts when speeds shift (like 10G to 1G before the AP) and when TDD aggregates TCP ACKs in the uplink.
We’re adding a feature in 5.11.0 to help with this — details are coming soon.
If you want more consistent speed test results, I’d suggest as well:
Using a modern TCP congestion control algorithm like BBR on your speed test server.
Adding a Cambium QoE device as a TCP proxy for better traffic shaping.
Hi Andrii, thank you for your response! I will look forward to the 5.11.0 release, as for tcp I cant find what I saw in the forum but I do know there has been chatter with TCP speed test results vs link test results.
this is what I was reading the other day in regard to MU-MIMO. Thanks!