Episode 2: Wave 2 Wireless

Facebooktwittergoogle_plusredditpinterestlinkedinFacebooktwittergoogle_plusredditpinterestlinkedin

In this episode of FortiCast, Ben Wilson and Koroush Saraf discuss Wave 2 Wireless technology.

Transcript

Download the Transcript PDF

EPISODE 2 – WAVE 2 WIRELESS
Dated: November 3rd, 2016
Host: Victoria Martin
Participant (P1): Ben Wilson
Participant (P2): Koroush Saraf

Host: This is FortiCast, the podcast for all things Fortinet. I’m your host, Victoria Martin.

Today’s episode is all about Wave 2 Wireless technology, which is also called 802.11ac. FortiAP’s use Wave 2 Wireless technology to increase wireless speeds, and allow more clients to share the same wireless access point, while avoiding bottlenecks.

Here to tell you more about it are Ben Wilson and Koroush Saraf, two of our wireless product managers.

P1: Hi and welcome to this FortiCast! I’m Ben Wilson.

P2: I’m, er, Koroush Saraf.

P1: We’re Wireless PMs, and we look after the wireless products within Fortinet, between us. We’re going to concentrate on what’s really happening, uh, within the world of wireless, and the main buzzword at the moment is ‘Wave 2’.

So, Wave 2, what is it?

P2: Wave 2, okay—we’re talking about the IEEE, WiFi standard that the industry has split into two, because it’s kinda difficult to get all of the standard [sic] implemented in the chip and get it all working in time to get to market right? So—

P1: So what we’re kinda saying is is that Wave 2 is actually really, purely a marketing term; there’s no such thing as the ‘AC-Wave 2 Standard’, it’s just something the industry has come up with to say that this is a next generation of chips with a few more features in, right?

P2: Pretty much, yeah I mean you can down the history lane of how the 802.11 standard has evolved from 11b—the 11 megabits back about 15 years ago—to 11n where you got [sic] the multiple antennas which was, I would say, fairly revolutionary—made WiFi quite stable and kinda started this all-wireless-office phenomenon and then you have the iPhones and iPads come along and all of a sudden wireless becomes a nice-to-have to—a necessity—

P1: Yeah, didn’t it just?

P2: And then—so the 11ac is the next thing that just makes everything go faster (inaudible 2:29) the term “Gigabit WiFi” was used, and there’s a lot of just complex algorithms, and math, and silicon, and firmware that’s involved to just make this whole thing work, so—I’m actually quite amazed at [laughter] how fast it operates and that it’s possible to do this kind of a thing.

P1: So I guess when people are talking to partners, or customers, or y’know people about Wave 2, when I talk about it and the kinda questions I hear and the feedback I get is really what is the difference between, y’know, Wave 1 and Wave 2.

I mean I guess the first thing that pops into my mind—or the main difference—is this thing called ‘Multi-user MIMO’, another great acronym and abbreviation that we need in the industry, ‘cause we don’t have enough of them, obviously.

Multi-user MIMO is the first time that WiFi has the ability to talk to more than one client at a time; hence the term ‘multi-user’ MIMO. MIMO was introduced in 802.11n; it’s multiple in, multiple out, and it means that you can have multiple antennas in order to transmit and receive traffic.

Now with Multi-user MIMO, you can talk to more than one user at a time, and although it’s not an exact technical analogy, a good way, or a familiar way that many people think about it, is like the difference between a hub and a switch. Y’know, WiFi-101 really is that, wireless is a shared medium, so when you think about an AP, it acts as a hub; the bandwidth is split between the users.

With Multi-user MIMO, because you’re able to talk to multiple clients at the same time, it acts a bit more like a switch. I’ve heard the terminology ‘wireless switch’ before, but basically if you think about it in the way that, it’s possible with something called ‘streams’—so in 802.11n, we had a maximum of three streams, but they all had to talk to the same client at the same time. With Multi-user MIMO, what we’re able to do is to split up to four streams to talk to different clients. And it’s a really important breakthrough and move forward—big part of why Wave 2 is certainly being so sought after and well received and why the product’s being developed so quickly, as you were talking about before.

P2: So I have a couple of data points on that actually and a few examples I think would be good to illustrate what the advantage is for multi-user, and where it comes into play.

So with today’s 802.11ac, Wave 1, and prior, when a client is talking, everybody else basically has to wait. So if we have a client that has the least capability and is talking, everybody else has to wait, so, if you could have a four-by-four access point, you could have 80MHz-wide channels, y’know it’s kinda like having a ten-lane highway, and if there’s a motor scooter going down, everybody else has to wait until this person’s—crosses before the rest of the traffic can go on this ten-lane highway.

In Wave 2, and in multi-user, it allows the AP to try to group devices that could simultaneously share the runways, so if you got a car that, for instance, needs to take four lanes, y’know, wide-load type of a thing, or if you have couple of motor scooters, maybe they can share a lane. So that’s what the AP is trying to do is trying to pack as many devices to simultaneously go at the same instant in time.

So we have some test-data that our R&D team has provided on our latest 802.11 chips, and obviously it—it depends on the client’s support for multi-user, but if you take a 20MHz-wide channel, which is the—really what I always recommend for high-density—20, maybe 40—and you take a one-by-one client, like a mobile phone, and a 20MHz wide—we’d run that with a Wave 1, we’d get about 70 megabits worth of bandwidth out of it. That’s essentially how fast a one-by-one client can go, in a 20Mhz-wide bandwidth, and you essentially need to add a second access point to get more bandwidth, and use more clients.

If we turn on the multi-user capability of that same access point, and we have three clients talking, instead of them all sharing the 70 megabit bandwidth, all of a sudden test results show we get over 200 megabits worth of bandwidth, so—

P1: Wow.

P2: —magically, there’s another tripling of this bandwidth, which really adds up—when I first looked at it, it seemed, er y’know, where did all this come from, and that’s what we’re gonna talk about next, um, how is this achieved, and how do you get tripling of the data in these Wave 2 technolo—

P1: It sounds almost too good to be true, y’know a ‘tripling of the data’, but y’know interestingly when you were talking about that just then, you talked about client support, and that’s one of the key things here, is that, one of the ways that is prevalent in making this achievable is the fact that there is something used called ‘beamforming’.

Now there are all kinds of different beamforming out there; there’s people who say they’ve got ‘smart antennas’ with dynamic and physical-beamforming, but what we’re talking about here is effectively—because you have to have client support as well, it’s an agreed matrix of parameters—RF parameters—which are agreed between both the client, and the access point, in order to boost signal-to-noise ratio, and make this multi-user MIMO work in the best way possible.

One of the key things about that is the fact that it is agreed between both ends, because there’s been some solutions out there, previously, where I’m sure our sales guys have come up against them, which have talked about beamforming, but that beamforming’s always been implicit; it’s been something that’s only happened on the AP-side; it’s been something where the AP is just being making a—effectively a ‘guesstimation’ of the client parameters, and trying to do the best that it can, whereas this is actually written in the standard, it’s explicit beamforming, it means that both sides negotiate, things like nought-packet, which means that the information is exchanged and it makes it work best for both sides, when the multi-user MIMO is, er, is turned on. Which is one of the reasons I think that you can almost see a ‘3X’ improvement in what’s going on.

P2: So we talked about (inaudible, 7:55) and how it’s achieved using this beamforming, and this table that allows the clients to differentiate what they need to listen to and that’s really where this extra-magical bandwidth comes from.

The ball (?) of Wave 2 is really efficiency of spectrum: “How do I not waste time, and organize?”, and so over the next year or so on the AP side, we’re gonna be improving how to group clients together, how to make sure that single-user-capable-only clients get separated out of multi-user so that we extract as much efficiency out of the air, right?

There is [sic] a couple of different aspects also that have been extended in Wave 2 that we can talk about, so Wave 1’s generally go up to three streams, I guess Ben could talk about what a stream is.

In Wave 2, you have an extra stream that single-user MIMO clients can use, and then the other way that you can get more bandwidth is by going the wider-spectrum. Normally Wave 1 supports 80Mhz-wide, but the wider the spectrum you have obviously the more you can fit into it and the more you can transmit.

So you go from 80Mhz to 160Mhz, and I think it’s worthwhile to talk about where that’s useful and where that’s not useful, because most people think they should just use everything that an access point gives them, but— [laughing]

P1: Yes! “Let’s turn everything on and see what happens!” [laughing]

P2: Exactly, exactly; click all the buttons and hope nothing bad goes wrong.

P1: Yeah!

P2: Yeah so—so let’s talk about that, let’s talk about the fourth-stream and 160Mhz and—

P1: Sure.

P2: —where it comes in handy.

P1: Sure, so basically, you know there are streams, and in 802.11ac we introduced something called ‘256 QAM’ which, I don’t really wanna ‘geek-out’ too much, but it’s a way of modulating in order to create more bandwidth, and effectively each stream is worth 433Mbps, so—

P2: (inaudible, 9:42) you’re not gonna geek-out, were you? [laughing]

P1: No. [laughing] No, no I’m not going to geek-out. But, but effectively that’s why you see kinda like the headline figures of roughly 1.3Gbps, on a three-by-three 802.11ac Wave 1, so for you to take something like a [sic] S321C, or a 320C, which is three-by-three, you would see like a headline figure of 1.3Gbps on the AC radio.

With the introduction of the fourth stream, you can add another 433Mbps to that, which is gonna give you kinda like 1.7, 1.8 Gbps, so there’s—there’s more bandwidth, but if you think about streams like lanes going to a town or a city, or something like that, well if you close off one of those lanes, then you’re not gonna be able to use it, and th—and what I’m trying to get at is, is if you’ve got a two-by-two client, or if you’ve got a three-by-three client, it’s not going to be able to take advantage of that extra fourth stream, when you’re in a single-user MIMO—MO, and obviously those people will only be talking to one client at a time.

So, when you look at these type of things, it’s great to look at the headlines, but you also as ever need to look at the type of clients that are in the environment that you’re specifying and designing the system for, because if they’re only ever gonna be one client-type, and it’s only ever gonna be two-by-two or three-by-three, then, y’know, the fourth stream will go relatively unused.

Host: I hope you’re enjoying this discussion. We have more information about Wave 2 Wireless available in our show notes, which you can find at our website, forticast.fortinet.com. Our website also has an archive of all our earlier episodes.

Our next episode’s going to be coming to you a little early, so look for it in mid-November, and it’s gonna be all about our new SOC3 chip. Here’s a quick preview of that upcoming episode:

Sneak peek: I think actually what’s striking when one looks at a [sic] internal schematic of the SOC3, is when I first seen [sic] one, I think I’m looking at the schematic of a high-end FortiGate device, because I see things on there, like a bunch of CPU cores, I see something called ‘NP6 light’, and I see something called ‘CP9 light’, and I think, “Yeah, okay, those are all the different elements of a big FortiGate”, and then of course I remember that I’m looking at the schematic of one chip! And of course, y’know, that’s the key feature; this is not multiple chips in one device, this is one chip with these multiple sub-systems in them. It’s quite striking, that, for someone coming from the higher-end devices…

Host: Now, back to Wave 2.

P2: So I think we should, uh then bring this all together and talk about what scenarios we’re talking about in terms of deployments, and where each one is useful.

So let me just, uh kind of summarize what we just talked through as far as the analogy that we’ve been using, which is this four-lane highway, so Wave 2 is basically a four-lane highway, and Wave 1 is the three-way highway.
Is that what you call it in England, as well? [laughing]

P1: Well, it’s correc—it’s correctly called a “motorway”. [sarcasm] [laughing] And we also run “betas”, not “betas”; don’t worry about that…

P2: What do you call traffic lights?

P1: …Traffic lights.

P2: Okay! [laughing]

P1: But we don’t have as many; we have roundabouts.

P2: “Roundabouts”, okay…Alright, so let’s imagine a four-lane highway which a Wave 2 brings, and then you have a traffic light. So if you don’t have multi-user MIMO, then every time that, uh the light turns green, only one of these vehicles can cross.

In a multi-user MIMO—if you got [sic] four vehicles in all four lanes, all four can potentially cross, and if you got [sic] let’s say two motor cycles, they could even share a lane and you could potentially even get more people across. So it’s about this sharing of the medium where you’re grouping devices that can fit together—cross at the same time, and that’s where this extra bandwidth comes from.

The extra pieces, like 160Mhz, and fourth stream, they’re generally useful for point-to-point links—uh, mesh links—so it’s not something that we recommend (inaudible, 13:31) enterprise application; it’s always better to have an extra AP; not only to s—have higher device capacity, but for redundancy…Ben, let me know if you agree with this, but—

P1: Yeah!

P2: —if we would be deploying an office building, and I had 30 people that had devices, I would take two access points and deploy it [sic] in a 40Mhz mode, maybe 80Mhz mode—

P1: Yeah…

P2: —instead of taking one AP, putting it on 160Mhz, and expecting to get similar type of performance.

P1: Yeah absolutely, because effectively you increasing [sic] the medium that’s available for the clients to talk to, you’re also creating more redundancy, from a pure network-design—er y’know…

P2: So I think, in a [sic] upcoming podcast, we should actually talk about the mechanics of how this medium is acquired; all of [sic] inefficiencies that are involved—

P1: Sure.

P2: —and why it actually makes sense to deploy more APs with narrower channels—

P1: Yeah.

P2: —than one AP with a huge, wide channel.

P1: Absolutely; that sounds like a really good topic.

P2: So we could talk about that next time.

So, let’s go to scenarios; let’s pick an office building, let’s say, um I have er, 100, er accountants, that are in a building—

P1: That sounds like a tough office to me: 100 accountants, but okay, we’ll go for that! [laughing]

P2: And they each have er, let’s say a laptop, and a mobile phone, so…First I guess if I wanted to buy an access point, and I had a two-by-two Wave 1, a three–by-three Wave 1, and a four-by-four Wave 2, which one would I choose, and what—what should I take into account when I make my decision?

P1: As ever, with ‘tech’, from my perspective, if you can afford it, you should always go for the latest technology available because that’s gonna give you the most future-proofing.

P2: At Fortinet, though, that’s actually the interesting part; at Fortinet the four-by-four and our three-by-three is pretty much similarly priced, right?

P1: Well there ya go! I mean, from a—from a purely commercial perspective, then we can take that—we can take that off the table, y’know, it’s not gonna cost me any more to, really, in reality, to deploy Wave 2 than it is Wave 1. So that’s great.

P2: So then it comes down to a technical—something you’re trying to explain to somebody: “Okay, I’m buying the Wave 2 because…”

P1: “I’m buying the Wave 2, because it has”, for example, multi-user MIMO, okay, which I may not have client devices today, which will support it, but I know they’re coming in the future. And let’s face it: How often do we change laptops? How often do we change phones? How often do we change tablets? If you’re a real Apple-junky, you’re doing it every year, and as more of these devices get more of these advanced features, you’re going to see effectively the ability to do that tripling the amount of bandwidth. It’s not an exact math—because it’s not—it’s not an exact science by any stretch of the imagination, but by certainly enabling that, you are definitely getting multiples of bandwidth, or increasing the availability. That is for absolutely certain, and that’s why the technology was developed.

P2: Yep.

P1: However, if you’ve got a fairly static client base, if you’re only ever gonna have two-by-two clients, you’re not really gonna see a massive advantage from the four streams. There’s all kinds of different things that you need to look at as well as the normal things, like, “Am I designing for coverage?”, “Am I designing for bandwidth?”, “Am I—” what particular application…Yeah, client capacity: all those type of things.

P2: So let’s take this other example with—I think may make it clear, so let’s imagine we have a school cafeteria; a university cafeteria—

P1: Yeah.

P2: —Y’know, it’s lunchtime, there is y’know a thousand people that are basically there in this cafeteria, right? And then, y’know during that time, everybody has their Android or smartphone, iPhone out; they’re checking Facebook; they’re Snapchatting; pulling down content; gonna mixed up laptops, etc so you got a very high client density, you got potentially very bandwidth-intensive applications—

P1: Yeah.

P2: —and let’s say, it’s two years from now, right? (inaudible, 17:10) purchase and it’s two years from now, so in two years, I personally expect many Wave 2 clients that are two-by-two to be out there. So if I had a four-by-four access point that’s Wave 2 capable, and it’s going to group these two—so all of a sudden I will double the effectiveness of my wireless deployment—

P1: Absolutely.

P2: —by having this Wave 2 access point, and potentially I can get away with having less access points—

P1: Yeah.

P2: —if my client density er needs our address (?) so, yeah…

P1: I’d agree with that.

The other thing we probably wanna mention as well about that, is that, from a technically standpoint, the Wave 2 APs are actually specified chipset level to handle a higher number of concurrent connections—Wave 1 APs top-end limit was 128 clients, currently, and with the Wave 2 it’s 512 – Now, would anybody really do a network design where you had 512 concurrent clients?

P2: The rule of thumb that er, we basically give is, yeah the top-end you can have 120+ associated devices. Then (inaudible 18:07) if you’re doing an IOT application, you have, y’know these low bandwidth devices, then yes you could actually have 120 devices associated and transmitting, like a couple kilobits or megabits at a time.

Our rule of thumb is about 30-50 devices to be—for planning purposes, if you are connecting, y’know HD cameras as part of the deployment, then you just need to take into account the total available bandwidth for the spectrum that you’re using.

So, planning for about 30-50 with Wave 1s is what was our rule of thumb, I think we can basically safely go and say 50-80, for Wave 2 um—logically that—the chips are more capable, the airtime is used less because the clients are basically getting their data across a lot faster, so you have this doubling of bandwidth, er benefits, if you have a Wave 2 capable device, you have a fourth stream that can give you a higher signal-to-noise ratio, even for single-user devices, you have beamforming which can, again, increase the signal-to-noise ratio, and then you have the higher association capability, and the cost is pretty much the same as a Wave 1, right?

P1: Sounds like a no-brainer to me! (laughing)

P2: So that’s er, that’s what we were hoping to get across in this er, in this podcast that, okay, it’s a lot of different things going on, but it’s pretty straightforward that, y’know the Wave 2 has quite a few benefits for future-proofing your purchase, and at least, er to me, it—it makes sense to just go ahead and augment the deployments.

And you can mix and match, so if you have Wave 1 APs you can add a few Wave 2s; they’re backward-compatible with Wave 1, and that sells—a great way to do greenfield (?) as well as augment existing deployments.

P1: So really, to kinda-like ‘wrap-up’, I guess what we’d say is is that there’s a couple of words; two or three terms that you might hear when you’re out there, y’know you’ve got “multi-user MIMO”, which is the ability to talk to multiple clients at the same time rather than clients individually, which is way [sic] you can generate extra bandwidth, you’ve got “beamforming”, and that basically means that as part of multi-user MIMO, each of the client and the AP share information about the RF environment, another items to create a matrix so they can agree on the best way to communicate, you’ve also got [sic] this option of 160Mhz-wide channels, but in any kind of enterprise scenario, you’re not really gonna be using 160Mhz, er—

P2: Let’s dedicate that for the mesh applications.

P1: —Yeah, absolutely.

And that’s really where we’re up to, and from a scenario point-of-view, as ever, design is pushed by what the customer requires, whether it’s client-density, whether it’s specific application, but y’know, everybody should be aware that there is higher capacity looking today [sic] on the Wave 2 access points.

P2: Yeah: higher capacity, not because we are widening the highway, or doing higher modulation—in fact the modulation is still the same, it’s efficiency of usage of the existing spectrum that—which is the theme that you’re going to be hearing for the next few years, is that, as the technology improves, (inaudible, 21:00) improve, y’know the laws of physics are still the same, so it’s—how do we use mathematics to get more information across in the same unit of time.

P1: Make more of what you’ve got. Because—

P2: There we go!

P1: —the laws of physics aren’t gonna give us much more spectrum, so what we have to do is optimize the use of it to make it as good as possible.

P2: That’s right.

P1: Well I really enjoyed that!

P2: That’s a good, er, good note to end on, I think!

P1: Excellent, excellent, well er, thanks for your time, and we hope you found this interesting today. I’ve been Ben Wilson.

P2: …I think you’re still Ben Wilson! And—(laughing) I’m er, Koroush Saraf, and we hope to be able to do more of these for you.

P1: Thanks very much, thanks for listening to the Wireless FortiCast.

Host: Special thanks to our guests, Ben Wilson and Koroush Saraf.

FortiCast is hosted and edited by me, Victoria Martin. Our show is produced by Michael Strickland and Bill Dickie. Our executive producer is Darren Turnbull. Our music is from bensound.com.

You can listen and subscribe to FortiCast in iTunes or any other podcast app. If you have any questions about the podcast, our e-mail is forticast@fortinet.com.

Thanks for listening!

More Wave 2 Wireless Resources:

Subscribe to FortiCast

         

Victoria Martin

Victoria Martin

Technical Writer & Head Cookbook Chef at Fortinet
Victoria Martin works in Ottawa as part of the FortiOS technical documentation team. She graduated with a Bachelor's degree from Mount Allison University, after which she attended Humber College's book publishing program, followed by the more practical technical writing program at Algonquin College. She does need glasses but also likes wearing them, since glasses make you look smarter.
Victoria Martin

Latest posts by Victoria Martin (see all)

  • Was this helpful?
  • Yes   No