Thanks to @slh for the command.
Simple command to remove lines from mac80211.sh file that @quarky recommends to solve this problem.
Connect to the device via SSH and enter the following command:
sed -i '/SHORT-GI-160/d;/vht160_hw/d;/$vht160/d;/# supported/d' /lib/netifd/wireless/mac80211.sh
Yes. I edited the script on my 23.05.0-RC2 installation.
One observation... after I applied the changes in the .sh. When changing the Width to 160 MHz on AX, my laptop with Intel(R) Wireless-AC 9560 160MHz card, connects to Wifi4 (n) with 300 Mbit/s.
AX@80Mhz is fine with 866 Mbit/s connection.
WIthout the changes in the mac80211.sh, with AX@160 setup on the router, the laptop connects with AC@160 MHz.
Maybe it removes too much, have you tried with the mt76 patch if this also happens there?
Well not really.
so the client connects to n (300 Mbps) and ac@80MHz (866 Mbps) at the same time? - Or only wifi 4? Do you have to manually switch it from n to ac?
What are the speeds you normally would expect with wifi 4?
Hi,
Router AX@160 - client (Intel(R) Wireless-AC 9560 160MHz) receives n@40
Router A@160 - client ( Samsung S23) receives ax@160
Router AX@80 - client (Intel(R) Wireless-AC 9560 160MHz) receives ac@80
What I believe should happen ( but I am not expert)..
Router AX@160 - client (Intel(R) Wireless-AC 9560 160MHz) receives ac@80, if the intend of the setup is to remove ac@160.
PS: Btw I executed sed -i '/SHORT-GI-160/d;/vht160_hw/d;/$vht160/d;/# supported/d' /lib/netifd/wireless/mac80211.sh on 2x RT3200
You could show the link connection modes for the connected wifi stations with this command:
iwinfo <device> assoclist
Or in Luci in the wireless network tab. Example output:
432.3 Mbit/s, 80 MHz, HE-MCS 4, HE-NSS 2
612.5 Mbit/s, 80 MHz, HE-MCS 6, HE-NSS 2, HE-GI 1
The first line is the uplink and the second line is the downlink connection of one wifi station.
AX@80 to Intel(R) Wireless-AC 9560 160MHz: 6.0 Mbit/s, 80 MHz 866.7 Mbit/s, 80 MHz, VHT-MCS 9, VHT-NSS 2, Short GIt
AX@160 to Intel(R) Wireless-AC 9560 160MHz: 6.0 Mbit/s, 40 MHz 300.0 Mbit/s, 40 MHz, MCS 15, Short GI
AX@160 to S23: 1361.2 Mbit/s, 160 MHz, HE-MCS 7, HE-NSS 2, HE-GI 1
AX@80 to S23: 960.7 Mbit/s, 80 MHz, HE-MCS 9, HE-NSS 2 1020.6 Mbit/s, 80 MHz, HE-MCS 10, HE-NSS 2, HE-GI 1
PS : seems I didn't copy the full info AX@160 to S23, but it is 160 Mhz uplink as well
The link connection needs to be monitored while you are doing a load or performance test, for example with a local iperf3 server.
On wifi stations with idling connections it could be power saving that you see slow link connection speeds. My examples above are link rates that are more or less constant during an iperf3 network performance test.
Some observed connection rates are not much indication. What matters is the actual achieved layer 3 network throughput, which is lower than the raw interface speed.
The test with s23 were while running test speed in waveforum. I ran the same test with the laptop, as I can not install third party soft - it is company laptop:
AX@80 to Intel(R) Wireless-AC 9560: 866.7 Mbit/s, 80 MHz, VHT-MCS 9, VHT-NSS 2, Short GI 866.7 Mbit/s, 80 MHz, VHT-MCS 9, VHT-NSS 2, Short GI
AX@160 to Intel(R) Wireless-AC 9560:300.0 Mbit/s, 40 MHz, MCS 15, Short GI 300.0 Mbit/s, 40 MHz, MCS 15, Short GI
Yes, MCS15 is 802.11n in 40 MHz channel, with a connection rate of 300 Mbit/s and two streams.
Do you have Apple iPhones or IPads in your wireless network? Do you have other 5 GHz neighbor wireless networks around which you will interfere with when you use a 160 MHz channel?
Well not now. My wife have, hence I was trying to improve her connection ( albeit she didn't complain), however she is on a business trip and that's the reason for me having time for testing:)
I will keep the setup like that until she returns to see what happen on the Iphone.
Btw: how can I revert back to the previous state ( just googled that there is no way back with the "sed" command)? I am using snapshots for my RT3200, I just need to update ( I am using auc) them to the latest, right?
On your question about other 5Ghz networks - answer is no - where I test. I have 2 Dump Aps - one on Channel 36 - interfering with others on 160, and the one I am currently testing on Channel 100, which do not interfere with any other networks on 160.
Thanks
Thanks to @neheb for the command.
With this command you return to the unmodified mac80211.sh file:
cp /rom/lib/netifd/wireless/mac80211.sh /lib/netifd/wireless/mac80211.sh
Another test using my kids laptop using Linux:
Network:
Device-1: Intel Wireless 8260 driver: iwlwifi v: kernel pcie: speed: 2.5 GT/s lanes: 1
bus-ID: 08:00.0 chip-ID: 8086:24f3
IF: wlp8s0 state: up mac: <filter>
Device-2: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet vendor: Lenovo driver: r8169
v: kernel pcie: speed: 2.5 GT/s lanes: 1 port: 3000 bus-ID: 09:00.0 chip-ID: 10ec:8168
IF: enp9s0 state: down mac: <filter>
AX@160:650.0 Mbit/s, 80 MHz, VHT-MCS 7, VHT-NSS 2, Short GI 780.0 Mbit/s, 80 MHz, VHT-MCS 8, VHT-NSS 2, Short GI
AX@80: 780.0 Mbit/s, 80 MHz, VHT-MCS 9, VHT-NSS 2 866.7 Mbit/s, 80 MHz, VHT-MCS 9, VHT-NSS 2, Short GI
I can test with Win 10 on the same PC. Let me know if it is needed.
Kind regards
even eaasier:
cp /rom/lib/netifd/wireless/mac80211.sh /lib/netifd/wireless/mac80211.sh
Over 600 Mbps uplink connection rates are fine. The issue here is mainly about AX capable Apple iPhones and iPads dropping to the lowest uplink connection rates with 32 Mbps in AX mode with some obstacles in line of sight.
The fix is in now!
And this commit maybe could alleviate the problem on AX160 with AC160 clients?
Thank you OpenWrt community for raising awareness of the issue and fix and thank you devs for listening! Should be plenty of time to get this tested out in snapshot and backed into 23.05 now. This is great news.
I see that we are looking at the functionnality on of the mt76xx driver
Can't they also look to make it work in AP/STA bridge mode like with the MTK driver ?