The release notes mention this:
450m CPY-18210 The 450m scheduler runs in the prior frame to eliminate lost downlink slots, which can improve downlink throughput in some scenarios.
Would someone at Cambium be able to:
(1) enumerate the scenarios for which this change is intended
(2) enumerate which schedulers can utilize this new change
(3) explain the expected impact of the change regarding AP statistics such as frame utilization, etc
Thanks!
This is a minor improvement (that took major work) that prevents wasting some downlink slots. This can be seen especially with busy APs processing multiple groups per frame. You should see an increase in downlink throughput. Of course, it depends on traffic and where SMs are and MU-MIMO groups, so hard to quantify in the field. Under ideal circumstances, maybe as high as 5% improvement in throughput. Improvement may be more visible on 2.5ms frames than 5ms since 5ms has a lot more slots. PMP450m 5GHz AP’s will also probably see a more visible increase then PMP450m 3GHz AP’s.
Credit where credit is due… I stole mostly @Charlie’s words 
Ah I didn’t see the 22.1 release thread - is there a link to it?
No, I’m roughly quoting from behind the scenes R22.1 development correspondence.
Oh ok - thanks for your help. Hopefully Charlie can expand a little bit more on that in this thread 
This feature is only for 450m, and is for MU-MIMO downlink scheduling scenarios, even with multiple MU-MIMO downlink groups per frame.
It applies to both Legacy and Proportional Schedulers.
Regarding Frame Utilization, both the pre 22.1 and 22.1 could run 100% utilization, but the 22.1 version runs in the prior frame can utilize a few more slots for acknowledgements and data. So you could see a slight increase in downlink throughput in 22.1 even with 100% utilization in both cases.
Of course this is very difficult to predict with MU-MIMO and what SMs have traffic demands on a given frame without a constant traffic generator.
Hope that helps.