This week is Aruba Atmosphere 2019, #ATM19, in what I am assuming is sunny and hot Las Vegas. I’m not there for numerous of reasons, one of them not being I didn’t want to go, but since no one opened up a GoFundMe to get me there, I am sitting in what I can tell you is sunny Colorado enjoying the weather but missing the Wi-Fi talk. What I am seeing is a fair amount of talk on social media about 802.11ax or Wi-Fi 6 both from the technical side and from the marketing side. How can I tell which one is which? It’s easy.
Marketing guys talk about speed, and when they are done talking about speed, they talk more about speed. Speed is sexy, and sex sells. I didn’t go to any marketing classes and I’m pretty sure that is year 1 and 2 or marketing school.
Technical guys talk about how angry they are that 1) it’s called Wi-Fi 6 and 2) how the marketing guys got it wrong. See, now you too can distinguish between the two when you go to parties and talk about the latest and greatest 802.11 standard. Top Tip: if you are talking about 802.11ax at a party you are definitely the nerdy guy of your group.)
What I want to talk about, as a technical guy, is the letters that come after the 802.11ax part. Some people don’t know it, but since 802.11n there have been letters that came after signifying something:
- 802.11n HT
- 802.11ac VHT
- 802.11ax HE
HT stands for High Throughput, and VHT stands for Very High Throughput. Since 802.11 is indeed wireless, we can draw the correlation to the radio world that if we had HF (High Frequency) and then came VHF (Very High Frequency) the next set of letters on the old (and I mean OLD) TV dial was UHF (Ultra High Frequency). Why then did the IEEE decide to go with HE instead of UHT for 802.11ax? Turns out the crowd that runs in the IEEE circle (the REALLY nerdy ones at parties) understood what most Wireless Professionals do; speed is awesome but more speed isn’t the answer right now. What is the answer? HE, or High Efficiency. Keep this in mind, as it is really important for the next 4 years or so in the world of Wi-Fi.
While 802.11ax indeed has 1024 QAM and in theory and in labs can achieve some crazy throughput numbers on the old speed test, most of realize that with the exception of most corner cases, it’s 100% useless and is there to help the marketing guys put their first 2 years of school to use. What we professionals really want is efficiency and stability. Stability is in the hands of the vendors and their code but efficiency is in the IEEE wheelhouse, and they crushed it on this one! What does all this rambling really mean? Let me try to put it in a story you can tell at parties when you decide to bring this conversation up (and probably seal your fate of never being invited back!)
I have been at my current place of employment for almost 11 years, and have lived in my current house for almost 4 years. My drive to and from work, up until recently, consisted of a pretty long stretch of a 2 lane road. I go to work REALLY early in order to avoid the traffic but in the afternoon, the only good answer was the countries most expensive toll road or this 2 lane disaster. Driving 4 or 5 times a week, I generally found myself on the disaster more often than not. The difference in my commute times between the morning and the afternoon was easily 20 minutes. The problem was there was one lane heading north, and a bunch of vehicles wanting to use that one lane.
Let’s go ahead and correlate this to a Wi-Fi Access Point (AP). There is generally one channel that everyone wants to use (5 GHz) and no matter how wide that channel is, only one device can use it at a time. Whether that device (vehicle) is a jumbo frame (extra wide load tractor trailer) or the newest, shiniest device (hypercar) or the cheapest (scooters), by the rule of the road, only one can use that lane at one time, no matter the payload or speed.
Back to my commute. When this was a 1 lane road, vehicles would back up all over the place, and it was a real disaster. Entire exit ramps would back up, the road would be clogged, and almost useless, for hours every afternoon. It really made that super expensive toll road look AWESOME! I spent more money than I care to admit on that toll road to avoid the congested disaster that was that 2 lane road with 1 lane headed north.
Luckily, just recently, they completed an expansion of that disastrous 2 lane road, and now it’s a whole FOUR lanes, 2 headed north and 2 headed south. Now, on my road, 2 vehicles can occupy what used to be one lane, at the same time. The difference? It’s beyond extraordinary! I can now breeze through what used to be a solid 5 minute backup without even stopping! The difference between my morning and evening commute, the one that used to be 20 minutes? Gone. Takes me the same amount of time now. The toll road? It’s still there but I don’t have to use it any longer.
On my road, the speed limit wasn’t changed from what it was before. There isn’t any more, or less, stop lights than before, and if anything, the number of vehicles using that road has actually increased instead of decreased, but my evening commute was almost cut in half by the addition of one single lane going my way. What used to be 20 minutes of stop and go traffic has now turned into a normal drive that I can accomplish sometimes without even stopping, and at the same time, handling more vehicles.
It can do it, because just like in the trunked radio world that I spoke about here, an addition like this is not a linear addition, it’s logarithmic, meaning it is more than the simple sum of the 2 parts. Taken in a single snapshot in time, it might appear to be simple multiplication, but over time, the ability of the traffic (frames or vehicles) to adjust and better utilize the resource (channel or lane) means that the overall impact is massive when you go from a single to multiple.
With 802.11ax HE, the HE part is allowing what used to be a single lane road to be utilized as if they made it into a NINE LANE road? Can you image what your commute would be like if all of a sudden every single lane road you drove was now NINE LANES? That simply crazy talk right there! The best part about this new technology that is allowing this to happen, OFDM vs OFDMA (as pictured above), is that the AP can finally be in charge of something and can dynamically adjust the lanes as needed, on the fly, giving those oversized loads (1500 byte frames) the space they need while the rest of us on our mopeds (300 bytes) only use what we need, and allow more of us to get through at the same time, reducing congestion, and improving everything overall.
This is accomplished by utilizing something that has always been there, but never utilized efficiently in Wi-Fi, called subcarriers. These subcarriers are grouped into what are termed Resource Units, or RU’s. Each RU will consist of no less than 26 subcarriers per RU, but if needed all subcarriers can be utilized, dynamically, to accomplish the task. Very efficient you say? Maybe those IEEE guys are on to something with their naming convention!
What this means the end result will be faster speeds that what you experience today, even if it isn’t the 1.21 jigabytes per second that 1024 QAM can do. But be honest, wouldn’t you simply be happier driving your Honda Accord at 60 MPH in a 45 MPH zone? Do you really need to do 150 MPH on your evening commute? So when will this construction be over and your evening commute speeded up? That is still being debated within many the circles. I was going to write a second post, but I’m going for the gold now, and in a down and dirty fashion.
I have heard some people say we won’t really see the expected results that I have explained above until nearly 40% of all clients in a given space can use this new eight lane freeway that comes with 802.11ax/Wi-Fi 6/OFDMA. My prediction?
12%, or at least 2 clients, of saturation in a given space and you will see a noticeable improvement
Put in the books, write it in the clouds, I am saying it here, the number of clients needed to see an improvement utilizing this new technology is WAY less than everyone else I have heard say. My reasoning is just like with my evening commute. When things are very inefficient, like they are today, it doesn’t take much to improve everyone’s life greatly. The next step in efficiency after this jump it will take a lot more, but for now, this is a huge addition.
Am I advocating that anyone run out and buy these brand new shiny AP’s to see this improvement before your neighbor? Not even close. As most of you know, vendor code has just as much to do with device performance as chip manufacturers, the IEEE, and the Wi-Fi Alliance ever thought about. Don’t believe me? I would like to introduce you to Lee Badman and his blog site and you can read all about it! For now, if you are in the cycle to buy new hardware, do your research and you should be fine. If you can wait, or you just refreshed your hardware, then for goodness sakes, WAIT! I believe that sweet spot will happen around the spring of 2020. You have time, and maybe, even by then, we will be talking about 802.11ax Wave 2!
So the next time you hear a sales guy start talking speed, nod quietly to yourself and know that while technically they are correct, and it will make things faster, it has nothing to do with 1024 QAM and 1.21 jigabytes per second, but all about being efficient.