Laptop intermittently disconnects from wifi (about every 5 mins)

I am on a Lenovo Thinkpad P14s AMD Gen 4 and Fedora 38 has been well except for the wifi. It seems to disconnect every few minutes and dmesg shows the following:

wlp2s0: deauthenticating from XX:XX:XX:XX:XX:XX by local choice (Reason: 3=DEAUTH_LEAVING)

I have tried disabling wpa_supplicant or NetworkManager but the services seem to be dependent on each other. I am using:

02:00.0 Network controller [0280]: Qualcomm Technologies, Inc QCNFA765 Wireless Network Adapter [17cb:1103] (rev 01)
        Subsystem: Lenovo Device [17aa:9309]
        Kernel driver in use: ath11k_pci

Any help would be appreciated. Cheers.

Edit:

I am also getting this when trying to reconnect without restarting the radio with nmcli:

[ 3987.310720] wlp2s0: authenticate with XX:XX:XX:XX:XX:XX
[ 3987.336387] wlp2s0: send auth to XX:XX:XX:XX:XX:XX (try 1/3)
[ 3987.340079] wlp2s0: send auth to XX:XX:XX:XX:XX:XX (try 2/3)
[ 3987.342804] wlp2s0: send auth to XX:XX:XX:XX:XX:XX (try 3/3)
[ 3987.346694] wlp2s0: authentication with XX:XX:XX:XX:XX:XX timed out

I had a similar problem with a qualcom adapter. I replaced it with an intel wifi 6 AX210 adapter and the problem went away. Cost → ~$25

Unfortunately that is not a solution for the issue, the card is soldered. Also, this card has great support according to other users and is meant to work well as intel wifi on linux. It is able to work for five minutes before disconnecting. So I’m assuming it’s a software issue.

To confirm it is directly card/driver related you might try a wifi dongle that uses the intel iwlwifi driver as a test. If the problem disappears it would seem definitely directly related to the card or driver.

My qualcom card had worked well for several years, but with updating to the 6.4 kernel some time back there appeared to be some regression in the kernel or driver that broke its stability.

This related but unsuccessful troubleshooting effort might provide some insight.

I’ve had no issues today as of yet, but I will test out that theory when the problem resurfaces.