you didn’t explain how you made those tests - from where to where you were sending traffic, what bandwidth generator you used? Please post wireless registration details.
Sorry, I was doing a test between 2 SXT5HnD board in bridge mode. I had them separated enough to have good signal but not too much. The test was using the b/w test in the MK.
The devices distance about 100m. With the built in bandwith test tool 100 mbit/sec one direction UDP, 54 mbit/sec both direction UDP. 54 mbit/sec one direction TCP, 25-30 mbit/sec both direction TCP. Tx/Rx CCQ is about 98-100% so the link quality is ok.
The real full duplex UDP is a dream I think. On TCP test the processor usage was 100%.
Advices:
try different spektrum, in my situation the default spectrum was so dirty
nv2 protocol can give you some gain
try nv2 TDMA change to xxx (I use 5 - better throughput and minimal increased latency)
Don’t run the bandwidth test on the SXT or you will not see the maximum throughput that it is capable of. If the CPU on the SXT is busy than you cannot expect it to have enough time for 802.11n protocol.
For a more accurate bandwidth test, you must plug another device into the SXT on each end and run the bandwidth test on those.
Also, why not set the radio mode to 5GHz-only-N? For maximum speed, you don’t use 802.11a.
Got 120 Mbps (Tcp) download & 52 mbps (Tcp) upload on SXT 5hnd.
Configured it using this link http://wiki.mikrotik.com/wiki/Transparently_Bridge_two_Networks_using_MPLS
getting this result by keeping the devices 5 meters away from each other.
Will be testing the device on an outdoor link of 7.5 KM next week.
thanks
At 5.17 KM the link is giving a signal strength of -81/-83 with both the channels enabled
& tx/rx ccq remains 36/12 literally pulling my hair as to why this is not working at all