Distributed Wireless Performance
The second part of wireless testing used the open air test process I've been using with all DWS products. But the time has come to streamline testing, so I've done away with the set of tests with the middle node in the "Hallway" location. From now on, the middle or second test node will be placed only in the Living Room location. This reduces test time and gives products their best shot at high performance due to stronger signal levels in the Living Room location.
The downlink bar chart shows results for all DWS products tested so far. Despite differences in backhaul method and bandwidth, it's fair to compare all these products because the same 2x2 test client was used for all tests and all products have at least a 2x2 AC radio for device connection. Note the eero results shown are for eero retested with V2.0 firmware.
It's obvious Velop didn't do very well at all in this test, turning in the lowest performance in all tests except Kitchen Reconnect, where it ever-so-slightly edged out Ubiquiti Amplifi, and Office, where it beat eero and Google Wifi.
DWS throughput summary - downlink
Uplink shows Velop coming in dead last except in the same-room Office test.
DWS throughput summary - uplink
So why did Velop do so poorly? Bad backhaul. The next chart shows throughput for Google Wifi, eero and Velop under same conditions. Except instead of connecting the test laptop via Wi-Fi, it was plugged into the Living Room and Kitchen nodes via Ethernet. This yields throughput as a wireless bridge to each point, which should accurately reflect the throughput available to that node's Wi-Fi radio.
Velop has a measly 27 Mbps available at the Living Room node and only 22 Mbps available at the Kitchen. Since throughput is so close between the two nodes, I'm guessing the Living Room - Kitchen link is using a different radio than the Office - Living Room.
Wireless bridge performance - downlink
Uplink results show similarly low throughput, with only slightly lower throughput at the Kitchen Velop than Living Room.
Wireless bridge performance - uplink
As to why backhaul throughput is so bad, I just don't know. The eero and Google Wifi results clearly show much higher throughput is possible, even without a second 5 GHz radio!
Perhaps Velop's odd choice of 2.4 GHz channels is a cause. The inSSIDer channel map below shows Velop (ssid = "DUT") using dhannels 1, 3 and 10. The channel 6 signal is a beacon from my Roku box, which is connected via Ethernet. This could account for the channel 3 use vs. channel 6, but channel 11 certainly could have been used instead of channel 10.
Linksys Velop channel use
Another possibility is that MU-MIMO, which is enabled for backhaul and client connections, is somehow lowering performance. Regardless, it appears Linksys has some backhaul tuning to do.
I was happy to see Linksys break out of the DWS 3-pack pack by including a second 5 GHz radio in Velop. Properly used, a third radio can provide more options for separating backhaul and client traffic and improving overall system performance.
But it appears Linksys has more work to do to get Velop in shape to be a viable competitor in an increasingly crowded field of distributed Wi-Fi systems.