Hacker News Reader

Comment by ksec | Hacker News Reader

Other than the 1200Mhz [0] from the 6Ghz Band, the requirement of 6Ghz Certification ( New WiFi Chip, Antenna, and FCC Certs ) means that WiFi 6E brings opportunity to fix everything that went wrong with WiFi 6.

The Current WiFi 6 clients and router have OFDMA problems, some are not turned on by default, some require firmware upgrade which you can guarantee wont happen with lots of users, many do not support 80+80Mhz / 160Mhz.

Without some of these features, current WiFi 6 / 802.11ax is nothing more than WiFi 5.5 or a slightly improved 802.11ac Wave 2.

I hope next generation WiFi6E client [1] and Router fixes all these problems.

[0] 1200Mhz only in US, Europe and UK are only getting 500Mhz and are on similar time frame with US. No idea on the scheduled in rest of the work.

[1] https://www.broadcom.com/products/wireless/wireless-lan-blue...

View on HN

For what it’s worth, there are practically 0 clients capable of handling WiFi 6 in those bands as well. Further, as a person in the industry, I think it’s completely irrelevant to need anyway.

At this juncture, interference and noise are more of the problem with WiFi today. It’s not larger bands capable of faster speeds. We need more spectrum to mitigate interference and provide more landscape for multiple AP’s in tight geographical spaces to CO-exist without competing too much for airspace.

Bonding 80Mhz channels for 500+ Mbps is overkill and hogs airspace. 40Mhz channels are more than adequate for 90% of devices out there if that channel is stable and free from congestion.

I’m super against the FCC granting more spectrum only for routers/APs manufacturers to default to hogging more chunks of the airspace. It’s a race to the same shitty experience we see today. We should parcel out chunks that are allowed for high channel bonding, leaving the rest of the space to max out at single 40 MHz bands.

Can't the spectrum be allocated dynamically, so routers don't "hog airspace" unless someone is actually using it and their TCP connection is actually going up to those speeds?

Well... that’s about software and less about spectrum and the regulation of it.

The problem is the box in the aisle of BestBuy that say “FAST 2.5 Gig WiFi” sells better than “Dynamic Signaling” so the incentive to just make really crappy radios that have more throughout outweighs the incentive to actually make decent and smart software defined radios.

That’s why I think the FCC needs to just set some band rules. They do this already in the commercial WISP space.

>For what it’s worth, there are practically 0 clients capable of handling WiFi 6 in those bands as well. Further, as a person in the industry, I think it’s completely irrelevant to need anyway.

Yes I thought it was clear with the WiFi 6E mention, but turns out not everyone is up to date with things as shown in other replies, so I apologies for the confusion.

I agree with everything you said, but my opinion is it should be down to Wifi to enforce Dynamic Signalling as mandatory in WiFi 6 ( too late now may be WiFi 7 ) spec rather than having regulation. And I thought OFDMA is part of the solution?

If I remember correctly that is what the WiFi 7 / 802.11be committees were looking into. But then again judging from their Opera happening in WiFi 6 I am not too sure if they are capable of doing it.

I live in a high density urban living area and my MacBook currently reads at least 20 AP excluding their 5GHz naming counterparts. So I do feel the pain with crappy WiFi.

We could put a requirement that your signal have a certain meaningful information density.

That's exactly how wifi works. If there's no data to be transferred, the radio's transmitter is off and it is not using the spectrum. That said, periodic beacons and keepalives defined in various protocols will cause data to be sent from time to time.

Interpreting an SSID closes to me, with 0 activity (noone connects there). According to wireshark, beacon takes 516µs at 6Mbit/s with 40 beacons/s, which equals 500e-6 * 40 = 0.02 = 2% airtime goes to a single SSID with 0 traffic. Typical settings are 10 beacon/s which would be 0.5%, but 2.4Ghz has 1Mbit/s default rate, so about 3%.

With 20 idle SSIDs on 2.4Ghz with default settings, roughly 50% of airtime is spent on SSID broadcast.

It just hit me that all of what is now sold as WiFi 6 does not in fact have a 6 GHz radio. The incompetence is breathtaking, it's been a year since they revamped all of their naming and it's already fucked again.

I’m not sure that ghz and version number were ever intended to be interchangeable, were they? I think if it happened it would be a happy coincidence.

True, it is coincidence IEEE 802.11 --> Wi-Fi 1

IEEE 802.11b --> Wi-Fi 2

IEEE 802.11a --> Wi-Fi 3

IEEE 802.11n --> Wi-Fi 4

IEEE 802.11ac --> Wi-Fi 5

IEEE 802.11ax --> Wi-Fi 6

IEEE 802.11ax in 6 GHz --> Wi-Fi 6E

IEEE 802.11be --> Wi-Fi 7

Has nothing todo with GHz frequency.

The intent of "Wi-Fi 6" is to start versioning Wi-Fi numerically (i.e. the next major rev will be Wi-Fi 7) instead of the IEEE's ever-lengthening 802.11 alphabet soup.

It's just a coincidence that we're also talking about 6 GHz at the same time. No formal relation.

The alphabet soup still makes more sense than names like USB Gen 3.2 2x2. Then again, because the door has been opened now by marketing guys to stray away from engineer driven names. It won't be long until WIFI 7 ULTIMATE ULTRA EXTREME 4 PRO EDITION is a thing.

Then why is GP referring to 6E? Sounds like more alphabet soup, to me.

That’s actually fairly standard naming in networking, like cat5 cables, cat5e is cat5 but after fixing a few bugs in the standard. Cat6 came out later and supersedes cat5 and cat5e.

Think of it like 6.1 instead of 6E. Minor revision to the existing standard.

Partially true. 6E is specific to 6GHz and not a revision of Wi-Fi 6.

It could be worse. You could have a business facing physical layer made by the same people and spend weeks trying to decipher what someone means when they says 802.3bj as opposed to 802.3-2015 when referencing 100GBASE-CR4, or it's medium-less counterpart: CAUI-4.

Reply on news.ycombinator.com