Pages in this thread: 
Print Thread
zzing123
Plus the problem that the new feature of WiFi 6 is not actually enabled on most of the new routers:
https://www.smallnetbuilder.com/wireless/wireless-fe...
Yes, but the thread is about Wifi6 vs CAT6.
When you compare wifi standards, we have seen big increases (say from g to n) but the difference between ac Wave 1 and ac Wave 2 (which is MU-MIMO) is almost negligible. OFDMA is pretty much the same thing. Physics-wise it's properly interesting tech, and it goes to the previous argument I laid out, that it's all about spectrum management.
For a quick recap, MIMO is 'Multiple in-Multiple Out', which means that in a specific timeslot (remember each device can only shout at a single time only: think of a field of people holding megaphones), what MIMO does is that someone shouts at a high-pitched voice and someone else shouts at a low-pitched voice and using a DSPs doing FFTs you can get two perfectly distinct 'sounds' from the same noise, and thus deduce a message. The physics of the same noise being made in the same timeslot however, remains the same, but through proper engineering you get multiple clients talking at the same time.
OFDMA is slightly different, in that you spread across different frequencies. So while everyone knows channel 1, 6 and 11 is used in Wifi 2.4GHz (each being 20MHz bands), 5GHz has much more bandwidth, but at those frequencies, channels don't go through walls as easily (this is why 7-800MHz 5G using the old TV bands really matters and why many £billions are spent btw), but generally 5Gz wifi is operating at 20, 40, 80, or 160 MHz channel widths, with most choosing contiguous 40 or 80MHz, and operating at channels 52-64 and 100-128 across the EU (You ain't got contiguous spectrum cos ETSI and Ofcom said so). OFDMA allows you to use non-contiguous channels, meaning an AP can use channel 52, 100, 104 and 128 for example to form 4x20MHz channels and an 80MHz as a whole, giving more performance. Of course any Wifi 5 device or below will only see a 20MHz channel, so meh. It's complex, and involves even more DSP that I can only surmise in a metaphor as a person broadcasting an interleave of 4 different letters of a word at very inaudibly different pitches at the same time and the receiver deciphering it together to form a word at the end of it.
Evenso, there's a commercial argument and a chicken/egg argument. As the SNB article states, OFDMA is about RU's which corresponds to DSP capacity. That means silicon and chip design in laymans terms. Radio chips aren't entirely dissimilar to GPUs in that they're massive arrays of Execution Units each covering and coalescing areas of spectrum as quickly and efficiently as possible. Just like a GPU driver can only do so much to rendering performance, it's the same with this ethereal 'Wifi Firmware'. So why would you build a new chip when you don't have a standard ready? Because muppets will buy it. So current Wifi 6 is pie in the sky for all we know, but at least it's a Proper Engineer's best guess as to what the standard will be, and so long as the chip has enough leeway and the firmware programmers are smart enough, a 'Wave 1' device can be estimated. So basically OFDMA will be 'ax Wave 2' just like MU-MIMO was 'ac Wave 2'. MU-MIMO was always baked into the ac standard, but Wave 1 was the modem maker's best guess of what the standard would be prior to the standard's ratification, while Wave 2 is the full implementation of the standard. Same shenanigans with AX.
Either way, the thread is about Wifi 6 and CAT6. Increases in the science and engineering of managing RF (Radio Frequency) are all well and good, and this is by no means faulting the scientists. This really is Proper Engineering with capital P and E, but there is one constant: the spectrum hasn't changed. So while different Wifi standards will give a measurable, benchmarkable performance uplift that many, many column inches will be written about, pretty much all the low-hanging fruit has been solved already in Wifi 5, and the increases are really about capacity and working the spectrum as hard, fast and efficiently as it can be. At best, you will really only see a modest real-world uplift in ideal performance of Wifi 6 over Wifi 5, and then only if all client devices are shiny new, and in an area with no interference (so not urban) until the TalkTalks of this world decide their 802.11n router is really a pile of *ahem* even for them.
The fact of the matter, is the single best step you can do, to get the most linear benefit (as in +100% for each AP you add up till you cover every last Hz of spectrum reserved for WiFi) is to wire your backhaul and invest in a proper network topology with a good switch and router. Only then, as you swap APs in and out you can harvest the +15% or so every time the IEEE decides to ratify a new standard, but even then, wait until the standard is ratified, and the chipmakers have baked a fresh set of silicon.
In the meantime, Wifi 5, or 802.11ac "Wave 2" devices are the current standard, until this page goes green, and it's still only half-baked: http://ieee802.org/11/Reports/tgax_update.htm