Since we have accumulated more than a few new people on the list... Back in 2016, I explained much that was wrong with wifi here, and at this particular time in the talk, I explained one of the biggest things wrong with 802.11n, ac, and ax, packet aggregation, and how the entire industry did testing, using my often hilarious method of "people as packets"....
https://www.youtube.com/watch?v=Rb-UnHDw02o&t=25m22s there was a LOT more wrong with wifi that could be fixed also outlined in that talk (multicast, 802.11e stupidities, etc - see the talk), but just fixing queueing as well as we did exhausted all the resources we had and it still took 4 years to get AQL into a more deployable state. Blatant plug: Toke's thesis is now in its second print edition and available for either download or a paper copy at: https://bufferbloat-and-beyond.net/ Get a copy. Get 10 copies! Give em to an IT manager, CTO, sysadmin, network designer, network operator, anyone that might listen. Post on your favorite social media site. Put one in your bathroom to read when you need a break from the internet itself. I gave a talk over here at broadcom about how to make queuing better in general, but I don't think they listened. http://flent-fremont.bufferbloat.net/~d/broadcom_aug9.pdf I'm pretty sure the next gen802.11ax mt76 chip can do gang scheduling right for ofdma, not so sure about anything else. -- Make Music, Not War Dave Täht CTO, TekLibre, LLC http://www.teklibre.com Tel: 1-831-435-0729 _______________________________________________ Cerowrt-devel mailing list Cerowrt-devel@lists.bufferbloat.net https://lists.bufferbloat.net/listinfo/cerowrt-devel
