We conducted tests in our office to assess the capacity of the MESH network. To do this, we chose two Ruckus R610 APs, which we usually use to create a WiFi network at various events. You can read more about this AP here:
http://www.3kgroup.ee/ruckus-r610-ac-wave-2-3x33/.
Mesh capability is available on all Ruckus APs except the 300 series APs.
For the test, we created a Mesh network between the two R610s, connected two computers running 1GbE port running Linux directly to the AP LAN port, and launched the iperf test. We set up the iperf test to send data in multiple streams at once using the TCP protocol, and the test was full-duplex. We placed the APs in different rooms, there was no direct visibility between the APs. As our office is located in the Tallinn area with a lot of data traffic in the air, it is probably not difficult to get better results in a better location.
As a result of this simple test, we obtained a data rate between the two computers of at least 300/300Mbps when the APs were "back to back", ie the upper parts of the AP were facing each other and 400/500 Mbit / s when the APs were facing each other or more at the same level.
Back-to-back positioning mimics a situation where APs are located on different floors, but offer better coverage to users on the same floor. The side is in a situation where the APs are on the same floor, but for some reason it is not possible to run the cable between the rooms.
Ruckus' MESH solution is very different from other manufacturers' solutions, as MESH APs communicate with each other with a two-way BeamFlex-controlled directional atenna. Beamflex allows signals to be routed and antenna polarization to be changed, resulting in more frequent spatial frequency reuse. Also, the data transfer rate of APs is usually higher, in this case half duplex TCP/IP 600Mbps, which is usually not realized between the client and the AP. Firstly, the customers are usually 2x2 MIMO and secondly, the distribution conditions of these customers are not very good. If the bandwidth of the whole system is calculated, then most likely the client's traffic on the two radios is 300 Mbps as TCP/IP traffic and the other 300 Mbps goes to MESH traffic.
Since the Wi-Fi protocol is still based on the ALOHA protocol, which is divided into time slots, and it is also necessary to acknowledge all packets, make sure the air is free before sending, etc., a very large part of the 1300 Mb / s theoretical maximum speed is managed by the system itself. Only the 802.11ax standard offers relief here through OFDMA bidirectional multiplexing. Thus, if we choose R730 APs (8x8 MIMO) with the current technique and put 10 APs in the MESH network, we get a very efficient MESH network, where the traffic between the APs takes place in parallel with OFDMA multiplexing. Most client devices are still likely to talk to the 11n and 11ac standards, but there is no bottleneck in this 11ax 6G Wi-Fi solution because, like our R610 test, the communication channel between APs is many times more efficient than the communication channel between APs and the client. In addition, there will be additional 500 MHz spectrum in the future to create MESH networks with 802.11ax standard devices.
Test results:
"Back-to-back" placement:
Martins-MacBook-Pro:iperf-2.0.5-macos-x86_64 martin$ ./iperf -c 192.168.3.76 -d -P 4 -w 128K Server listening on TCP port 5001 TCP window size: 128 KByte Client connecting to 192.168.3.76, TCP port 5001 TCP window size: 129 KByte (WARNING: requested 128 KByte) [ 8] local 192.168.3.23 port 52242 connected with 192.168.3.76 port 5001 [ 5] local 192.168.3.23 port 52239 connected with 192.168.3.76 port 5001 [ 7] local 192.168.3.23 port 52241 connected with 192.168.3.76 port 5001 [ 6] local 192.168.3.23 port 52240 connected with 192.168.3.76 port 5001 [ 10] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39720 [ 9] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39718 [ 11] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39722 [ 12] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39724 [ ID] Interval Transfer Bandwidth [ 8] 0.0-10.0 sec 85.2 MBytes 71.5 Mbits/sec [ 6] 0.0-10.0 sec 87.4 MBytes 73.3 Mbits/sec [ 7] 0.0-10.0 sec 72.5 MBytes 60.7 Mbits/sec [ 5] 0.0-10.0 sec 87.1 MBytes 73.0 Mbits/sec [SUM] 0.0-10.0 sec 332 MBytes 278 Mbits/sec [ 11] 0.0-10.0 sec 95.6 MBytes 79.9 Mbits/sec [ 10] 0.0-10.1 sec 95.5 MBytes 79.7 Mbits/sec [ 9] 0.0-10.1 sec 97.6 MBytes 81.5 Mbits/sec [ 12] 0.0-10.1 sec 95.0 MBytes 79.3 Mbits/sec [SUM] 0.0-10.1 sec 384 MBytes 320 Mbits/sec Martins-MacBook-Pro:iperf-2.0.5-macos-x86_64 martin$ ./iperf -c 192.168.3.76 -d -P 8 -w 128K Server listening on TCP port 5001 TCP window size: 128 KByte Client connecting to 192.168.3.76, TCP port 5001 TCP window size: 129 KByte (WARNING: requested 128 KByte) [ 13] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39726 [ 6] local 192.168.3.23 port 52254 connected with 192.168.3.76 port 5001 [ 12] local 192.168.3.23 port 52260 connected with 192.168.3.76 port 5001 [ 9] local 192.168.3.23 port 52257 connected with 192.168.3.76 port 5001 [ 5] local 192.168.3.23 port 52253 connected with 192.168.3.76 port 5001 [ 11] local 192.168.3.23 port 52259 connected with 192.168.3.76 port 5001 [ 10] local 192.168.3.23 port 52258 connected with 192.168.3.76 port 5001 [ 7] local 192.168.3.23 port 52255 connected with 192.168.3.76 port 5001 [ 8] local 192.168.3.23 port 52256 connected with 192.168.3.76 port 5001 [ 14] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39732 [ 15] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39734 [ 16] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39736 [ 17] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39738 [ 18] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39740 Waiting for server threads to complete. Interrupt again to force quit. [ ID] Interval Transfer Bandwidth [ 12] 0.0-10.0 sec 59.5 MBytes 49.9 Mbits/sec [ 9] 0.0-10.0 sec 55.9 MBytes 46.8 Mbits/sec [ 5] 0.0-10.0 sec 60.4 MBytes 50.5 Mbits/sec [ 11] 0.0-10.0 sec 59.4 MBytes 49.7 Mbits/sec [ 7] 0.0-10.0 sec 57.1 MBytes 47.8 Mbits/sec [ 10] 0.0-10.1 sec 41.8 MBytes 34.8 Mbits/sec [ 15] 0.0-10.2 sec 66.5 MBytes 54.6 Mbits/sec [ 18] 0.0-10.2 sec 67.5 MBytes 55.4 Mbits/sec [ 13] 0.0-10.2 sec 70.4 MBytes 57.8 Mbits/sec [ 14] 0.0-10.2 sec 66.9 MBytes 54.9 Mbits/sec [ 16] 0.0-10.2 sec 68.1 MBytes 55.9 Mbits/sec [ 17] 0.0-10.2 sec 66.8 MBytes 54.8 Mbits/sec [SUM] 0.0-10.2 sec 406 MBytes 333 Mbits/sec [ 6] 0.0-10.2 sec 1.75 MBytes 1.44 Mbits/sec [ 8] 0.0-10.2 sec 1.75 MBytes 1.44 Mbits/sec [SUM] 0.0-10.2 sec 338 MBytes 277 Mbits/sec
After being placed on the same plane, the results were much better, because the AP antennas are placed more favorably with each other:
Martins-MacBook-Pro:iperf-2.0.5-macos-x86_64 martin$ ./iperf -c 192.168.3.76 -d -P 8 -w 128K
Server listening on TCP port 5001
TCP window size: 128 KByte
Client connecting to 192.168.3.76, TCP port 5001
TCP window size: 129 KByte (WARNING: requested 128 KByte)
[ 13] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39726
[ 6] local 192.168.3.23 port 52254 connected with 192.168.3.76 port 5001
[ 12] local 192.168.3.23 port 52260 connected with 192.168.3.76 port 5001
[ 9] local 192.168.3.23 port 52257 connected with 192.168.3.76 port 5001
[ 5] local 192.168.3.23 port 52253 connected with 192.168.3.76 port 5001
[ 11] local 192.168.3.23 port 52259 connected with 192.168.3.76 port 5001
[ 10] local 192.168.3.23 port 52258 connected with 192.168.3.76 port 5001
[ 7] local 192.168.3.23 port 52255 connected with 192.168.3.76 port 5001
[ 8] local 192.168.3.23 port 52256 connected with 192.168.3.76 port 5001
[ 14] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39732
[ 15] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39734
[ 16] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39736
[ 17] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39738
[ 18] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39740
Waiting for server threads to complete. Interrupt again to force quit.
[ ID] Interval Transfer Bandwidth
[ 12] 0.0-10.0 sec 59.5 MBytes 49.9 Mbits/sec
[ 9] 0.0-10.0 sec 55.9 MBytes 46.8 Mbits/sec
[ 5] 0.0-10.0 sec 60.4 MBytes 50.5 Mbits/sec
[ 11] 0.0-10.0 sec 59.4 MBytes 49.7 Mbits/sec
[ 7] 0.0-10.0 sec 57.1 MBytes 47.8 Mbits/sec
[ 10] 0.0-10.1 sec 41.8 MBytes 34.8 Mbits/sec
[ 15] 0.0-10.2 sec 66.5 MBytes 54.6 Mbits/sec
[ 18] 0.0-10.2 sec 67.5 MBytes 55.4 Mbits/sec
[ 13] 0.0-10.2 sec 70.4 MBytes 57.8 Mbits/sec
[ 14] 0.0-10.2 sec 66.9 MBytes 54.9 Mbits/sec
[ 16] 0.0-10.2 sec 68.1 MBytes 55.9 Mbits/sec
[ 17] 0.0-10.2 sec 66.8 MBytes 54.8 Mbits/sec
[SUM] 0.0-10.2 sec 406 MBytes 333 Mbits/sec
[ 6] 0.0-10.2 sec 1.75 MBytes 1.44 Mbits/sec
[ 8] 0.0-10.2 sec 1.75 MBytes 1.44 Mbits/sec
[SUM] 0.0-10.2 sec 338 MBytes 277 Mbits/sec
Martins-MacBook-Pro:iperf-2.0.5-macos-x86_64 martin$ ./iperf -c 192.168.3.76 -d -P 8 -w 128K
Server listening on TCP port 5001
TCP window size: 128 KByte
Client connecting to 192.168.3.76, TCP port 5001
TCP window size: 129 KByte (WARNING: requested 128 KByte)
[ 13] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39774
[ 7] local 192.168.3.23 port 52288 connected with 192.168.3.76 port 5001
[ 12] local 192.168.3.23 port 52293 connected with 192.168.3.76 port 5001
[ 6] local 192.168.3.23 port 52287 connected with 192.168.3.76 port 5001
[ 5] local 192.168.3.23 port 52286 connected with 192.168.3.76 port 5001
[ 9] local 192.168.3.23 port 52290 connected with 192.168.3.76 port 5001
[ 8] local 192.168.3.23 port 52289 connected with 192.168.3.76 port 5001
[ 11] local 192.168.3.23 port 52292 connected with 192.168.3.76 port 5001
[ 10] local 192.168.3.23 port 52291 connected with 192.168.3.76 port 5001
[ 15] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39780
[ 14] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39788
[ 16] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39782
[ 17] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39784
[ 18] local 192.168.3.23 port 5001 connected with 192.168.3.76 port 39786
Waiting for server threads to complete. Interrupt again to force quit.
[ ID] Interval Transfer Bandwidth
[ 12] 0.0-10.0 sec 50.1 MBytes 42.0 Mbits/sec
[ 6] 0.0-10.0 sec 68.4 MBytes 57.3 Mbits/sec
[ 5] 0.0-10.0 sec 64.4 MBytes 53.9 Mbits/sec
[ 8] 0.0-10.0 sec 68.6 MBytes 57.4 Mbits/sec
[ 17] 0.0-10.1 sec 87.0 MBytes 72.2 Mbits/sec
[ 9] 0.0-10.0 sec 71.5 MBytes 59.7 Mbits/sec
[ 15] 0.0-10.1 sec 86.9 MBytes 72.1 Mbits/sec
[ 14] 0.0-10.1 sec 88.2 MBytes 73.2 Mbits/sec
[ 18] 0.0-10.1 sec 86.4 MBytes 71.6 Mbits/sec
[ 13] 0.0-10.1 sec 90.8 MBytes 75.1 Mbits/sec
[ 16] 0.0-10.1 sec 87.1 MBytes 72.1 Mbits/sec
[SUM] 0.0-10.1 sec 526 MBytes 436 Mbits/sec
[ 7] 0.0-10.1 sec 2.25 MBytes 1.87 Mbits/sec
[ 10] 0.0-10.1 sec 2.25 MBytes 1.87 Mbits/sec
[ 11] 0.0-10.2 sec 69.9 MBytes 57.7 Mbits/sec
[SUM] 0.0-10.2 sec 397 MBytes 328 Mbits/sec
