Hello we have a P2P 40km link with a clear LOS.

# INTERFACE RADIO-NAME MAC-ADDRESS AP SIGNAL... TX-RATE
0 wlan1 000C42000000 00:0C:42:00:00:00:00 no -53dBm 90Mb...

when we do a bandwith-test we get very different values between the transmission and reception:

tool bandwidth-test address=10.??.??.?? duratio
n=20 user=admin password=mypsw direction=transmit protocol=udp
status: done testing
duration: 21s
tx-current: 27.1Mbps
tx-10-second-average: 26.1Mbps
tx-total-average: 24.3Mbps
random-data: no
direction: transmit
tx-size: 1500

tool bandwidth-test address=10.??.??.?? duratio
n=20 user=admin password=mypsw direction=receive protocol=udp
status: done testing
duration: 20s
rx-current: 51.8Mbps
rx-10-second-average: 54.4Mbps
rx-total-average: 43.6Mbps
lost-packets: 826
random-data: no
direction: receive
rx-size: 1500

the address=10.??.??.?? is a board over the wireless link.

The question is why we get so many different values in the bandwidth test?

If the signal is the same or similar on both sides it is some noise at the remote end.

That's what we thought too.

Site A
Site B

Test from Site A
tool bandwidth-test address=10.??.??.?? duratio
n=20 user=admin password=mypsw direction=transmit protocol=udp
status: done testing
duration: 21s
tx-current: 27.1Mbps
tx-10-second-average: 26.1Mbps
tx-total-average: 24.3Mbps
random-data: no
direction: transmit
tx-size: 1500

tool bandwidth-test address=10.??.??.?? duratio
n=20 user=admin password=mypsw direction=receive protocol=udp
status: done testing
duration: 20s
rx-current: 51.8Mbps
rx-10-second-average: 54.4Mbps
rx-total-average: 43.6Mbps
lost-packets: 826
random-data: no
direction: receive
rx-size: 1500

So if the bandwidth-test is better in receiving the noise will be on the site A or the other side?

cetalfio

Noise is on Site B.

Well. There could be also some traffic during the test, no? Then the test shows only its results, so it is lowered by the capacity taken by other traffic.
Anyway, you should monitor the CCQ while doing the test to see the link quality.