PTP 550 link 12 km, LOS, antenna 30 dBi. Tx 23 dBm, Link Type 2+0 2x80 MHz, TDD DL/UL=75/25
LINKPlanner
Monitor-Wireless
PTP 550 link 12 km, LOS, antenna 30 dBi. Tx 23 dBm, Link Type 2+0 2x80 MHz, TDD DL/UL=75/25
LINKPlanner
Monitor-Wireless
UDP Duplex test by Bandwidth Test Mikrotik
TCP Download simplex test by Iperf
TCP Upload simplex test by Iperf
TCP Duplex test by Iperf (DL) and BT Mikrotik (UL)
Link Test simplex DL 820 Mbps +UL 260 Mbps=1080 MBps
UDP duplex, DL 835 Mbps+ UL 210 Mbps=1045 Mbps (DL+UL), RTT avg 11 ms
TCP DL simplex 835 Mbps, RTT avg 8 ms
TCP UL simplex, 230 Mbps, RTT avg 3 ms
TCP duplex, DL 810 Mbps+ UL 160 Mbps=970 Mbps (DL+UL), RTT avg 6 ms.
Tests were done by Unidata and WISP ArgoCom, Ukraine.
Conclusions
1. We see amazing throughput test results about 1 Gbps in 2x80 MHz Channel bandwidth, that we have never seen before in 5GHz.
2. Link Test, UDP and TCP tests results are almost the same. TCP tests, that take into account TCP packets losses, gives the simular results as UDP tests, that do not depend on channel packet's errors and losses. It means, that Cambium PTP550 link due to it's efficient Link Adaptation has low channel BER (Bit Error Rate)/ PER ( Packet Error Rate).
Very often ( in most cases) BWA equipment Internal Link Speed Test, channel Capacity, external UDP tests due to huge packet's losses in radio channel do not indicate real channel TCP/IP throughput. Only TCP tests of devices with high level channel BER/PER are able to give proper estimation of real link TCP/IP throughput.
Cambium PTP550 channel has low BER/PER ( I think BER may be 1.0E-5 - the simular as link at PTP450i ) and it's Internal Link test may be used to estimate channel real TCP/IP throughput.
PS
Competitors' 802.11AC and other products with support of high ( 1024QAM, 2048 QAM) modulation usually have TCP throughput test results much less ( very often in 2-3 times in case of 802.11 AC ), then their UDP tests results. Very often they may show Capacity ( and also high data rate/modulation) as indicator of high link throughput. But due to it's high channel BER/PER at high 256QAM/1024QAM modulation very often it is not real TCP/IP throughput of given link. It is max link speed, that theoretically is able to be achevied in ideal conditions, that do not exist in real life.
Thank you for this post. Do have a PTP 550 deployed in your network?
Yes,our end users WISPs have it a lot. Most of them are satisfied by PTP550 since v. 4.3.1 firmware and resolving SFP port issue in 4.3.2
Thanks -
Hi Vyacheslav,
We are glad you like the radios and also we would like if you be willing to share Your story via case study or solution paper. If you are interested kindly email me sagar.deshpande@cambiumnetworks.com
We can help spread the goodness and share your success story
Thanks
Sagar
congratulñations, just keep in mind that your spectrum is really free, I wish I can see PTP 550 in a noisy enviroment
Hi,
What happens if you run the TCP throughput test over the Mikrotik?
Bandwidth test Mikrotik ( any model) is not able to generate more then 330-350 Mbps TCP traffic. So for testing 40+20, 40+40... up to 80+80 MHz channels by TCP packets another test traffic generators should be used , for example Iperf, IxChariot. Also 4 Mikrotik routers (2 MT at each side of wireless link) may be used.