How does Beamforming works - explanation

[RMerlin]

Pretty good in-depth explanation on how beamforming (including SU-MIMO and MU-MIMO) works:

https://www.oreilly.com/library/view/80211ac-a-survival/9781449357702/ch04.html

 

[sfx2000]

Pretty good in-depth explanation on how beamforming (including SU-MIMO and MU-MIMO) works:

https://www.oreilly.com/library/view/80211ac-a-survival/9781449357702/ch04.html

Should note a couple of things...

11ac cleaned up beamforming for 5GHz - much better than the 802.11n approaches with explicit and implicit beamforming approaches.

Also note that in 11ac - SU-MIMO, beamforming is optional, where MU-MIMO is totally dependent on Beam Forming.

Marvell did a great job with 11n, but not many clients support 802.11n beamforming - they've done a good job with 11ac for SU, esp in their 4*4:3 chipsets (like in WRT1900ac) with the closed source vendor drivers - the FOSS drivers for Marvell are basically Wave1, and with reduced capability compared to the closed source - but attention there has been more focused on stability.

Broadcom SU-MIMO beamforming was decent enough for 11ac in 5GHz, but early issues with MU and interop with client chipsets was a problem.

QC-Atheros ATH10K is probably one of the best in 11ac Wave2 for both SU and MU when using the closed source (QSDK) drivers - in MU operation, no impact to SU clients across multiple client chipsets, unlike Broadcom, and MU compatibility is good - with the FOSS drivers, there are two branches - one for Wave1 and one for Wave2 on ATH10K.

 

[Trip]

Also a good adjunct to this is looking at PD-MRC. It's a large part of why I still like Ruckus over most everything else, but I would expect most of the tech to commoditize and appear in everyone else's stuff, even consumer, in the next few years. At least I would hope...