Anyone else seeing better performance at MCS9 vs MCS11 on 80mhz 4625 SMs? These are SMs whose signals are good and the performance tab of the SM and AP show that this SM is doing MCS11 98% of the time or more so it is not modulating down during the performance tests.
DL Retrans % is 1% or less for each SM and I am seeing this on most SMs on different towers and APs.
Example would be an SM got 571 down 149 up on a 20s link test with the SM set to Auto on DL and MCS10 on the UL However the same SM got 620 down and 146 up when set to MCS9 on the DL. This was repeatable.
Another on another AP. This SM is the only SM on the AP. Performance table shows 98.9% it is showing MCS11 and 1% it is MCS 10. On the link test it gets 582 DL 256 UL on a 20s link test. It gets 628 DL and 253 UL when the SM is set to MCS9 on the DL. When I set back to MCS11 for the DL the SM gets 538 DL and 254 UL on the link test.
I assume you’re running 5.9.1 on everything? Under best case circumstances, e4k @ 1024QAM can do about 10/BpHz… so 20MHz = about 200mbps (aggregate).
I have noticed that larger channel widths like 80MHz and 160MHz in PtMP mode seem to struggle a bit getting close to the 800mbps and 1600mbps (respectively) aggregate theoretical max.
Yes 5.9.1 and 80mhz channels. I have been questioning the 80mhz performance I was getting and today decided to try forcing the SMs to MCS9 (256 QAM) to see if I got the same or worse performance, but to my surprise I got better performance on all link tests.
I am going from like 650mbps aggregate to 791mbps aggregate just by telling the SM to only go up to MCS9 on the DL. UL numbers seem consistent and do not change much if I use Auto MCS or force the UL MCS.
This is also interesting. This performance ‘increase’ only seems to affect single SM link tests. If I do a dual SM link the AP gets more performance at MCS11 than it does at MCS9 but single SM link tests are better at MCS9 vs MCS11.
