Wi-Fi 6 Release 2 revealed

[psychopomp1]

Austin, Texas – January 5, 2022 – With adoption of Wi-Fi CERTIFIED 6™ growing steadily, Wi-Fi Alliance® is evolving Wi-Fi 6 certification to meet increasing demands from today’s Wi-Fi® use cases. Wi-Fi CERTIFIED 6 Release 2 is now available, bringing new features that support increasing device and traffic density to deliver greater performance and power management with Wi-Fi devices and applications. Wi-Fi CERTIFIED 6 Release 2 adds support for uplink multi-user multiple input, multiple output (multi-user MIMO) to deliver smoother streaming services and video conferencing, faster uploads, and more reliable gaming. Additionally, three power management features improve Wi-Fi CERTIFIED 6 power efficiency, benefitting enterprise, industrial, and Internet of Things (IoT) applications. New features apply across all bands supported by Wi-Fi 6 – 2.4 GHz, 5 GHz and 6 GHz - bringing capacity, efficiency, coverage and performance benefits to residential, enterprise, and large public networks. Wi-Fi CERTIFIED 6 delivers the best experience with advanced applications, while providing strong WPA3™ security and promoting interoperability between Wi-Fi CERTIFIED™ devices.

Wi-Fi CERTIFIED 6™ Release 2 adds new features for advanced Wi-Fi® applications | Wi-Fi Alliance

www.wi-fi.org

I wonder if existing wifi 6 routers can be upgraded to 'release 2' through a firmware upgrade? Or does it require different (newer) hardware altogether?

 

[avtella]

TWT and Uplink MU if I recall we’re advertised by QCA back in 2017 for the IPQ807x series. In reality however Gen 1 807X for example, I think was missing uplink MU (ie Asus AX89U / Netgear RAX120v1), took a second revision for it (ie Netgear RAX 120v2). Not sure about Broadcom though in regards to what features were or not missing. Probably others with more knowledge here. Overall seems like this certification is just to ensure full AX feature set is supported.

 

[RMerlin]

Is Wifi Alliance certification even relevant nowadays?

 

[thiggins]

This is just the Alliance announcing they've upgraded their Certification test suite. As @avtella notes, some early chipsets may not support MU-MIMO or TWT or not support it properly. At any rate, they are not really important features for most users.

@RMerlin I would say it's still relevant. Wi-Fi has gotten so complex that standards testing is important. It doesn't catch everything, but can weed out really poorly implemented products.

Of course, Certification is done at one point in time and not with each firmware revision.

 

[TheLostSwede]

Is Wifi Alliance certification even relevant nowadays?

Very good question. We keep getting more and minor updates to every release, so now we have 802.11ax/WiFi 6, WiFi 6E, WiFi 6 UNII-4, WiFi 6 Release 2 and they don't seem to be quite combined into one thing. Obviously 6E isn't doable in all countries, which is causing further issues, but neither is UNII-4 it would seem. It's gotten to a point where your average consumer is going to be utterly clueless as to what they're looking at, so amazing job there WiFi Alliance.

 

[TheLostSwede]

This is just the Alliance announcing they've upgraded their Certification test suite. As @avtella notes, some early chipsets may not support MU-MIMO or TWT or not support it properly. At any rate, they are not really important features for most users.

@RMerlin I would say it's still relevant. Wi-Fi has gotten so complex that standards testing is important. It doesn't catch everything, but can weed out really poorly implemented products.

Of course, Certification is done at one point in time and not with each firmware revision.

Nor do they seem to be doing good old fashioned plug fests before hardware is released, so there are seemingly more and more compatibility issues between brands. Considering Intel and Realtek seems to rule the desktop and laptop client side of things, with MTK seemingly trying to win some market share by teaming up with AMD, plus Qualcomm and MTK being the two big mobile players, as well as Apple which seems to use Broadcom or Qualcomm depending on device, but Broadcom seemingly being the major higher-end router chip provider, with MTK being in most lower-end products and finally Qualcomm and Realtek in some niche products, unless of course you have a cable modem/gateway, then you might have MaxLinear...
Anyhow, due to there being a lot of proprietary features from each of these companies that aren't part of the WiFi spec, you'd think the WiFi Alliance main job would be to smack these big companies over the head every time there was an issue due to them not implementing the standards properly, but alas, that's not the case.

So what's the point of having a standards organisation that doesn't enforce the standards? Not saying the WiFi Alliance is unique here, as the USB-IF and HDMI Forum are just as bad imho.

It's obviously not just down to the chip makers, as I'm sure the router manufacturers are doing some dodgy things as well. It's just frustrating that so many consumers are having issues with their wireless products simply because the standards aren't followed properly and many of the issues are hardware issues that can never be resolved, so people have to go out and buy new hardware and hope that solves the problem.

 

[RMerlin]

@RMerlin I would say it's still relevant. Wi-Fi has gotten so complex that standards testing is important. It doesn't catch everything, but can weed out really poorly implemented products.

But with manufacturers selling products based on draft standards for years now (they did it with AC, and they did for years for AX as well), I don't know if manufacturers even bother anymore getting their products certified, as I assume the Wifi Alliance doesn't certify draft standards?

 

[RMerlin]

It's gotten to a point where your average consumer is going to be utterly clueless as to what they're looking at, so amazing job there WiFi Alliance.

It's the latest trend in the industry. Confuse everyone by ruining clear standards like USB or HDMI, and now Wifi is doing the same thing despite having made things clearer only recently (Wifi 5 and Wifi 6 is certainly less cryptic than 802.11ac and 802.11ax for the average customer). Now they are joining the party by adding this WIfi 6 Release 2 thing.

 

[TheLostSwede]

It's the latest trend in the industry. Confuse everyone by ruining clear standards like USB or HDMI, and now Wifi is doing the same thing despite having made things clearer only recently (Wifi 5 and Wifi 6 is certainly less cryptic than 802.11ac and 802.11ax for the average customer). Now they are joining the party by adding this WIfi 6 Release 2 thing.

You're not playing with UNII-4?

But yes, who is going to know if their USB4 port is Gen 2.1 or Gen 3x2? For some stupid reason, the USB-IF decided to continue with their daft naming. Also note that 2x2 and 3x1 are both 20Gbps, but not the same when it comes to the encoding or lane speed...

Also, now we're seeing things like AXE6600, which appears to be 600 Mbps in the 2.4 GHz band, 1200 Mbps in the 5 GHz band and 4800 MHz in the 6 GHz band, i.e. only the 6GHz band is 4x4... Talk about screwing over consumers just to be able to sell a product with a high speed rating.
Most people would gain much more from having 4x4 on the 5GHz, since I would hazard a guess that people own far more 5GHz than 6GHz devices.

 

[thiggins]

But with manufacturers selling products based on draft standards for years now (they did it with AC, and they did for years for AX as well), I don't know if manufacturers even bother anymore getting their products certified, as I assume the Wifi Alliance doesn't certify draft standards?

IIRC the first AX Certification suite came out before AX was released and was based on a draft (4, I think).
I understand your points, But without the WFA, there would be TOTAL chaos. They are by no means perfect, but have done a better job of putting certification suites together than, say, HomePlug.

Wi-Fi's feature set has become so complex that the best the average consumer can do is look for simple product generation indicators (Wi-Fi 6, 6 etc.), just as they do with mobile (3G, 4G, LTE). They don't understand the details. They just know the higher number is more current technology.

 

[sfx2000]

Broadcom seemingly being the major higher-end router chip provider, with MTK being in most lower-end products and finally Qualcomm and Realtek in some niche products

I would argue that QCA is the leader in WiFi6/6e chipsets, Broadcom is a close second in volume for AP/Routers

MTK is leading up the low end and doing a very good job down there, Realtek hasn't been a volume factor in a very long time as they've been more focused on the client silicon.

Interestingly enough, Quantenna is still out there - see a lot of them in Operator Prep Equipment (CPE's)..

 

[sfx2000]

Wi-Fi's feature set has become so complex that the best the average consumer can do is look for simple product generation indicators (Wi-Fi 6, 6 etc.), just as they do with mobile (3G, 4G, LTE). They don't understand the details. They just know the higher number is more current technology.

Spot on...

 

[RMerlin]

You're not playing with UNII-4?

UNII-4 was only approved in Canada back in May or something like that. And the only device that I have that supports it is a router that's currently sitting on a shelf gathering dust. No client for it (and no need for it either, quite frankly).

Most people would gain much more from having 4x4 on the 5GHz, since I would hazard a guess that people own far more 5GHz than 6GHz devices.

I've seen quite a few times - I don't expect much out of Wifi 6e personally. Limited country support, limited client support, adds a significant cost to a router to support since it requires an extra dedicated radio... I don't see Wifi 6e ever becoming mainstream before Wifi 7 actually makes it into a "close enough, let's ship draft products" status. I always considered Wifi 6e to be a niche "side product" rather than an evolution on Wifi 6.

A Wifi 6e standard would have gained more traction IMHO if either combined with the gradual deprecation of the 2.4 GHz band (so the total BOM of routers would remain fairly the same), or if more manufacturers had started releasing Wifi 6e-only APs. Plug a single radio AP to your existing network, and instantly gain Wifi 6e support at a reasonable investment in cost.

Yes, I agree that going 3 stream or even 4 stream on the 5 GHz band would make much more sense as an evolution. The fact that virtually nobody is selling 3 or 4 stream clients right now shows how little market interest there seems to be for increased wifi throughput. If I remember correctly, Apple was the only laptop manufacturer to ever support three streams.

Wi-Fi's feature set has become so complex that the best the average consumer can do is look for simple product generation indicators (Wi-Fi 6, 6 etc.), just as they do with mobile (3G, 4G, LTE). They don't understand the details. They just know the higher number is more current technology.

And this is where the Wifi Alliance just completely screwed up their own thing. I loved the idea of Wifi 4, Wifi 5, Wifi 6, as it was more indicative than a bunch of cryptic letters. The customer instantly knew that wifi 6 was faster than Wifi 5. But now they will expect customers to distinguish between Wifi 6, Wifi 6e, and Wifi 6 Release 2. Is Wifi 6 Release 2 faster, or slower than Wifi 6e? Can my Wifi 6 router work with my Wifi 6e phone, or do I need a new router?

They are back at square one.

 

[RMerlin]

I would argue that QCA is the leader in WiFi6/6e chipsets, Broadcom is a close second in volume for AP/Routers

Qualcomm benefits from the increasingly large phone/tablet market. They are also quite popular with many AP/Router manufacturers. We probably have a skewed vision here because Asus is very Broadcom-centric, more than their competitors.

Realtek hasn't been a volume factor in a very long time as they've been more focused on the client silicon.

I think Realtek has a good market share in the low-end IoT/embedded market, which is probably a fairly significant volume.

Interestingly enough, Quantenna is still out there - see a lot of them in Operator Prep Equipment (CPE's)..

Asus used to be working on a product based on Quantenna's 10G platform, and it never went anywhere.

Wasn't Quantenna bought by another company a few years ago?

 

[TheLostSwede]

UNII-4 was only approved in Canada back in May or something like that. And the only device that I have that supports it is a router that's currently sitting on a shelf gathering dust. No client for it (and no need for it either, quite frankly).


I've seen quite a few times - I don't expect much out of Wifi 6e personally. Limited country support, limited client support, adds a significant cost to a router to support since it requires an extra dedicated radio... I don't see Wifi 6e ever becoming mainstream before Wifi 7 actually makes it into a "close enough, let's ship draft products" status. I always considered Wifi 6e to be a niche "side product" rather than an evolution on Wifi 6.

A Wifi 6e standard would have gained more traction IMHO if either combined with the gradual deprecation of the 2.4 GHz band (so the total BOM of routers would remain fairly the same), or if more manufacturers had started releasing Wifi 6e-only APs. Plug a single radio AP to your existing network, and instantly gain Wifi 6e support at a reasonable investment in cost.

Yes, I agree that going 3 stream or even 4 stream on the 5 GHz band would make much more sense as an evolution. The fact that virtually nobody is selling 3 or 4 stream clients right now shows how little market interest there seems to be for increased wifi throughput. If I remember correctly, Apple was the only laptop manufacturer to ever support three streams.

The only UNII-4 device I've seen is the upcoming model form Synology, but I'm sure we'll see more. It seems like a more sensible extension than 6E imho, even though it only adds one more 160MHz channel for the 5GHz band.

It'll be hard to kill off 2.4GHz, since as you know, it has longer range, better wall penetration etc. That said, your idea is interesting, as if the main router has 2.4GHz and the AP's are 5GHz or even 6GHz only, that seems like a sensible compromise, since its the higher frequencies most people are having issues with when it comes to range. That would allow people to use all their 2.4GHz IoT devices, while getting better overall WiFi coverage in their homes.

I'm not sure if most people are aware of how MU-MIMO works and it's something most people should be able to take advantage of now, if they somewhat modern devices. It obviously has its limitations, but with additional AP's, some of those limitations would also go away, assuming devices are connecting to the closest WiFi AP/router. But as no-one has come up with such a solution...

 

[thiggins]

Wasn't Quantenna bought by another company a few years ago?

https://www.onsemi.com/site/pdf/Quantenna-Acquisition-Announcement.pdf

 

[thiggins]

. I don't see Wifi 6e ever becoming mainstream before Wifi 7 actually makes it into a "close enough, let's ship draft products" status. I always considered Wifi 6e to be a niche "side product" rather than an evolution on Wifi 6.

Wi-Fi 7 mainly brings 320MHz channel support. Where do you think those channels are going to come from?

Unless the CTIA manages a bandwidth-grab coup, 6 GHz Wi-Fi is here to stay. Too bad Wi-Fi lost out on the CBRS band. The lower frequency would have meant better range.

 

[sfx2000]

Wasn't Quantenna bought by another company a few years ago?

On Semi bought them for $1B USD back in 2019... rumor is that they're looking to sell.

Anyways, all the 11ax Residential Gateways provided by the local cableco - CoxHSI - Technicolor with Quantenna radios for both bands.

 

[RMerlin]

The only UNII-4 device I've seen is the upcoming model form Synology, but I'm sure we'll see more. It seems like a more sensible extension than 6E imho, even though it only adds one more 160MHz channel for the 5GHz band.

My bad, I confused it with Wifi 6e channel extensions. I don't know anything about UNII-4.

 

[RMerlin]

Wi-Fi 7 mainly brings 320MHz channel support.

Hm, I thought it brought other notable improvements.

Adding 320 MHz channel support is kinda odd considering pretty much nobody even uses 160 MHz channels today. I used it for a while, but I eventually got annoyed by the longer boot times while the router was doing DFS scans. Granted, I do reboot my router more often than a regular user.

Does Zero-Wait DFS apply only to when you get bumped off the channel, or also for the initial boot time scan?

Where do you think those channels are going to come from?

Maybe Wifi 7 SoCs will be able to consolidate all these channels so a single radio can cover them all, rather than requiring a separate radio to handle the extended Wifi 6e channels. That would resolve the issue, by allowing manufacturers to go back to dual SoC designs.

On Semi bought them for $1B USD back in 2019... rumor is that they're looking to sell.

Buying a chip manufacturer right before a major chip shortage. Talk about bad timing. They're unlikely to recover much of their initial investment if they do offload it that quickly...