3.0.1-RC9 lower speed than 2.6.2.1

Products ePMP
1

Hello,

I upgraded an ePMP 1000 Sector to ePMP 2000 Sector with beamforming antenna.

Everything seems ok but I noticed than upgrading SMs from 2.6.2.1 to >= 3.0 cause a lower downlink wireless link test result.

Here's the result with 2.6.2.1:

user_2621_performance.png
user_2621_performance.png1242×1243 78.7 KB

Here's the result using 3.0.1-RC9 (I also tried with 3.0, same result):

user_301_test.png
user_301_test.png1447×1252 121 KB

I then tried to downgrade to 2.6.2.1 and the downlink speed went up to 70Mbps, in the same moment.

It seems that using 3.0 cause a lower modulation, even if the signal and the wireless conditions are exactly the same.

Look at performances screen:

2.6.2.1

user_2621_performance.png
user_2621_performance.png1228×1246 214 KB

3.0.1-RC9

user_301_performance.png
user_301_performance.png1381×1250 244 KB

The Force 180 is very close to the AP:

signal.jpg

I didn't change anything but the software version on the SM.

On the AP I have 3.0.1-RC9.

Did anyone notice the same behavior?

Thank you

2

Hi giuseppe4,

I tried a few times and I cannot reproduce your issue on my setup, I have the same downlink and uplink on 2.6.2.1 and 3.0.1-RC9.
I can suggest you try to start Wireless Link Test a few times (for example 3 in series) than change firmware and make same tests on it.
If you want, you can upload or send on my mail (mykola.rudenko@cambiumnetworks.com) your configuration files from AP and SM, and I will try reproduced it again on my setup (maybe root cause in some specific configuration on your devices).

Thank you

3

Hi Mykola,

I tried to start Wireless Link Test more than 10 times always with the same result. Changing the firmware results in a throughput decrease if I upgrade from 2.6.2.1 to 3.0, and throughput increase If I downgrade from 3.0 to 2.6.2.1.

I'm sending you an email about that.

4

Thank you for operational efficiency, I will let you know about any results.

1 Like
5

I feel that I am having a similar issue. 

My upload stinks.

I have poor MSC and support says because of interference.

I thought these radios are supposed to handle a noisey environment?

My tests showed are -85 to -90.

6

Based on your screen shots, it seems pretty clear that v 2.6 is selecting MCS-15 and v3 is preferring MCS-12 for some reason. It would seem like the algorithm is maybe now erring on the side of caution?