Intro Wi-Fi Learning, Frequencies, Wi-Fi History, Interference duplex, Adjacent channel interference, 5Ghz, DFS, Channel bonding, Signal strength, BSSID, Other wireless, BLE, Ethernet vs Wi-Fi NICs, SNR and data rates, 802.11 frame types and timing, roaming, security, Wi-Fi Design & Ekahau design
So just a quick introduction, so you know who I am. My name is Joel. I am based out of Boise, Idaho. You’re looking at my home office right now. You might hear my family and stuff in the background. I hope you don’t mind.
I am CWNE #233. CWNE stands for Certified Wireless Network Expert. There are currently, I think, 379 or 380 of us worldwide, something like that right now.
We have at least, I know, at least one other CWNE on the call. So I’m not the only one here. That’s about a five year process to get the CWNE. And the biggest thing that I learned getting that was that I don’t know anything about Wi-Fi. That’s ultimately the number one takeaway that I got from that.
What I’m going to try to do today in this course is I’m going to try to just give you as much overall information about Wi-Fi as I possibly can in a couple of hours. So we’ll spend two hours talking about general Wi-Fi stuff. And then we’ll look at some specifics from some of our sales engineers here at Mist.
So if you want to get more education after this is done, we plan on doing more of these. We’re just going to kind of scratch the surface for a lot of this stuff. For example, I have one slide on security, one slide. And that is just barely a quick introduction to the different security methods that are available in Wi-Fi in 802.11.
So that’s just one example of something that we’re just going to skim over. We’re just going to gloss over that. And so we should have more content over time where we’ll try to fill in some of these gaps.
And what I also want to point you towards is that there’s another way you can learn more about wireless, if you want to really dive deep into this. And that is to that program where I got the CWNE. And that’s called the CWNP, or Certified Wireless Network Professional. CWNP is a vendor neutral certification organization. And they offer a whole bunch of different certifications that you can take that all focus on different aspects of wireless.
For example, the CWNA, the Certified Wireless Network Administrator. That’s kind of the entry level one. That’s– the book is massive.
It’s like that thick. It’s huge. It’s back here on my shelf somewhere and it takes up like a whole shelf. It’s massive.
CWNP is a good place to start. The CWAP is the Analysis Professional. If you want to learn more about the protocol, how the protocol actually works, then the AP is a great place to go. Then there’s the Certified Wireless Design Professional that will walk you through some of the fundamentals of how to do wireless network design. We’re going to talk about that today, by the way.
And then the CWSP, which talks about the Security Professional. So check out CWNP if you want to really dive in with this wireless stuff. I’m going to give you a taste today.
We’re going to offer more content from Mist to try to enable you with Mist products. But check those resources out. There’s lots of great, great resources there, as well.
Also, another thing that I want to do is I want to point you towards Twitter. The Wi-Fi community hangs out on Twitter. I don’t know why, but they do. That’s just where all the wireless people hang out.
So if you want to get involved in the conversation, if you want to ask questions, if you just want to keep up with the latest news and find the latest resources, Twitter is a fantastic place to do that. I probably, I don’t know, I would say easily 50% of what I know, what I understand about Wi-Fi, the good stories that I have to tell, things like that, they come from Twitter. Because that’s where those stories are.
So I highly recommend getting involved there. Go check it out. Get on Twitter. And if you follow a few right people, then you can get involved there.
So a question from C. Doyle here. Is there a podcast Joel might recommend, as well. Yes, there is. There’s actually several, lots of good ones. But I think I’m just going to point you at one for now. We’ll just start with one.
I would recommend checking out the “Clear to Send” podcast. That’s by Roel Dionisio and Francois Verges. They are awesome presenters. And they crank out so many episodes and so much content.
Back when I was working on my CWNE, I took the CWSP, the Certified Wireless Security Professional. And Francois did an episode on security. I just listened to that like five or six times on the way to work every day.
And it got to a point where I really had the content down. Because he would just lay it out verbally. It’s nice, because you can listen to it.
John Koch also recommends “The Contention Window.” That’s another excellent podcast. To be honest, I’ve only listened to a couple of episodes now.
It’s on my phone. It’s in my backlog. I need to get to it. Haven’t got there yet. But I hear from tons of people that it’s fantastic.
And also “Packet Pushers,” another one I haven’t listened to. But probably a really good recommendation, as well. So let’s go ahead and jump right in.
So the very first place that I want to start is I want to talk about the organizations that make this whole Wi-Fi thing happen. There’s two central organizations that are absolutely critical that are fundamental for wireless. And the first is the IEEE. That’s the Institute for Electrical and Electronics Engineers.
They create standards. That’s what they do. They make standards for things.
For example, 802.15.1, that is an IEEE standard. And looking at the numbers there, you might not recognize it. But I’m sure you use it every day.
That’s Bluetooth. 802.15.1 is Bluetooth. So they make the standard.
And then that goes over to another organization who applies branding to it. In the case of Bluetooth, that’s the Bluetooth sig. That’s the Bluetooth special interest group.
They apply the branding. They apply all that stuff to it. But the standard is called 802.15.1.
Another standard that you may or may not have heard of is 802.15.4. That one is Zigbee. Here in the house I have a couple of Philips Hue lights. So for example, this is a wireless light switch. If I hit on button, you see the blue glow behind me there.
This is driven by Zigbee. This is all– oops, I hit the off button. This is all driven by Zigbee. So I’ll talk about this a little bit later on. But 802.11 15.4 is the standard that defines Zigbee for low power communications, things like that.
Next is IEEE 1394. This is a bit of an older one. You might remember this one. This one goes all the way back to– back to the late 90s.
Remember the original Macintosh that Apple– or not the original Macintosh, the original iMac that Apple introduced? That had IEEE 1394, a.k.a. FireWire on board. And there were a few machines prior to that had fire wireless, as well. The original iPod, that was driven by FireWire. That’s another example.
Now we’re getting down to ones that are really pertinent to this presentation today. First off is 802.3. 802.3, that’s ethernet. That is where the ethernet standard comes from is 802.3.
And we see amendments, things added to that standard over time. For example, 802.3at, that’s power over Ethernet. Or that’s at least power over Ethernet plus. That’s one of the things that we’ve added to power over Ethernet over time.
And then finally, we get down to 802.11. That is the standard for Wi-Fi. There’s no branding associated to it. That is the standard for it.
And so that then goes over to the WiFi Alliance. So the Wi-Fi Alliance is basically an organization. It’s a non-profit organization that is composed of major players in the wireless space. Think companies like Apple, and Sony, and Qualcomm, and T-Mobile. These are all companies that are part of the Wi-Fi Alliance.
And that’s just a really short list of who is in the Wi-Fi Alliance. Basically, they provide branding for Wi-Fi. They are where the term Wi-Fi comes from.
If you see that little Wi-Fi logo on your– if you see that little Wi-Fi logo on a chipset or something like that, I’ve got a little Arduino board over here that has a Wi-Fi logo on it, that means that it is certified by the Wi-Fi Alliance. They provide that brand. They own that brand.
So they come up with Wi-Fi. They also come up with things like WPA and WPA2. Those, in fact, stem from 802.11i. Those both come from 802.11i. But the Wi-Fi Alliance applies the branding to it.
Wi-Fi 6, that’s also– that’s another thing that the Wi-Fi Alliance has done. The other really, really important thing that they do is that they ensure interoperability between vendors. So if you want your device to be Wi-Fi certified, you have to send it into the Wi-Fi Alliance and they will test it for interoperability with other vendors.
Like for example, my phone. This would have had to be sent into the Wi-Fi Alliance. And they would test it to ensure that it basically conforms to the standards and it behaves how it’s supposed to. And that it interoperates with other devices.
So you could see that there’s two organizations that are absolutely critical for the operation and function of Wi-Fi. That being the IEEE and the Wi-Fi Alliance.
– OK. Cool. So the next thing that I want to talk about is I want to talk about the two frequency bands that Wi-Fi, for the most part, operates in. And this is really, really important. This is very, very fundamental stuff. So if this seems basic, don’t worry. We’re going to ramp things up here as we go. We’re going to start– we’re going to walk, then we’re going to run, then we’re going to sprint.
So there are two primary frequency bands that Wi-Fi operates in. The first is the 2.4 gigahertz band. Let’s change our pen color here so you can actually see that. The 2.4 gigahertz band. Now, 2.4 gigahertz indicates basically the wavelength. How long is our wavelength from start to finish?
So we have our beginning of our wave there. We have our ending of our wave there. That is essentially our– that is our wavelength. Now, this ties directly to frequency, how fast that wave oscillates. And it also ties directly into the actual physical length of the wave. You can actually tie these into real-world values. I don’t have them here, but you can but you can calculate exactly what the real-world length of these waves are.
Now, since 2.4 gigahertz is a lower frequency, it has a tendency to punch through stuff really, really well, so it tends to go a long way. You ever notice that when you’re sitting in your car and somebody else is pumping some music in a couple of cars over or something like that, you can hear the bass, right? That’s because bass is– it’s a low frequency. And so we can hear that really, really well.
The same is true in RF, in the radio frequency realm as well. Lower frequencies tend to punch through stuff better. They tend to attenuate– they tend to attenuate less than higher frequencies. And that’s why 2.4 gigahertz tends to go a lot further than the 5 gigahertz band.
Skip, thank you for throwing that in the chat. Skip put in there that 2.4 gigahertz is– I can never remember these off the top of my head, but I don’t really need to, so I don’t really worry about it– is 4.9 inches. That’s the actual length of the wave. And then up in the 5 gigahertz band, it’s 2.5 inches.
So I think it’s a cruel twist of fate that the 2.4 gigahertz band is almost 5 inches long and the 5 gigahertz band is– that the wavelength is almost 2.4 inches long. It’s just a cruel twist of fate. So since we’ve got a longer wavelength, we typically get more range. It tends to go further.
And because of that, 2.4 gigahertz ends up being more crowded. There’s actually a number of reasons why 2.4 gigahertz has ended up being more crowded than the 5 gigahertz band. And we’ll get to that here in a couple of minutes. We’ll talk about why that is. Because it’s more crowded, we see a lot of non-Wi-Fi interference there.
Now, one important thing to keep in mind about both 2.4 and 5 gigahertz is that these are what’s called unlicensed spectrum as opposed to licensed spectrum. Now, for example, Verizon, T-Mobile, AT&T– just to name a few cell phone carriers out there– they all pay for licensed spectrum.
Essentially, they pay the IEEE for the privilege of using specific frequency bands– for example, maybe 1.9 gigahertz, 600 megahertz, those different frequency bands that are out there. They pay for a license. Those are licensed spectrum. And you really can’t do anything on their spectrum because they own it.
A few years ago– I think it’s probably six years ago now, something like that– there was a guy in Georgia who decided that he was sick and tired of people talking on their cell phones while they were driving, and he was going to fix that problem. So he went out and he bought a wideband jammer that jammed multiple cell phone bands. And he threw it in his car.
And every day on his way to work, he fixed the problem. Nobody could talk on their phones around him because he was jamming them all so all their calls were dropping. Well, all of the cell phone carriers noticed that up and down this one corridor every day, they would just drop call, drop call, drop call. And then they’d go back that evening. They’d all drop calls.
And they all submitted complaints to the FCC. And the FCC got out there, and they found him, and they identified who he was. And they wrote him a $250,000 ticket. You don’t mess with licensed spectrum. You don’t do that or else you will be in trouble with the FCC and you will be hit with a massive fine.
And so Wi-Fi, on the other hand, does not operate in licensed spectrum. Imagine if the average consumer had to go out and get a license to operate their devices at home, if they had to get a license to plug in their little gateway at home and have Wi-Fi at home. That’d be a huge pain. Nobody would use it. And so 2.4 and 5 gigahertz are designated as unlicensed spectrum, which means you don’t need a license to operate there. You just have to abide by some basic rules.
Now, both 2.4 and 5 gigahertz, they hold that as unlicensed spectrum. They’re both unlicensed. 5 gigahertz does have some incumbents that we’ll talk about later. But really, 2.4 gigahertz ended up being really, really popular– I think, personally, because of the better range. The range is really good. It goes a long way. Typically, we can get about 300 feet, something like that. It highly depends on the technology that we’re using, whether it’s Wi-Fi or an old cordless phone or something like that.
But we get good range out of it. And so it makes sense to put consumer electronics there. That’s why we see all of these consumer devices there. However, the problem is starting to spill over into the 5 gigahertz band now. We’re starting to see more and more things that are causing interference.
So the old– I don’t know. The old thing of saying like, yeah, there’s no non-Wi-Fi interference in 5 gigahertz. [GROANS] It’s starting to happen. We’re starting to see it. So 2.4 gigahertz is where 802.11b, g, and n live. We skipped it for 802.11ac. 802.11ac did not operate in the 2.4 gigahertz band.
But then I think the IEEE and the Wi-Fi Alliance realized, like, you know what? 2.4 gigahertz is not going away any time soon. We’re still using it, so let’s go ahead and bring 802.11ax back down into 2.4 gigahertz and try to improve things there with 802.11ax. So 802.11b, g, n, and ax, they all operate in the 2.4 gigahertz band.
Now, going up and looking at 5 gigahertz. It’s a higher frequency, so the wavelengths are much shorter– only 2 and 1/2 inches as opposed to almost 5 inches. Because that wavelength is shorter, we see less range. It just doesn’t punch through stuff nearly as well.
Now, the funny thing about comparing 2.4 and 5 gigahertz is if you put them outside, if you put them in an open field, they go about the same distance. It’s really about the objects in the building. It’s about the building itself. Even people in the building, those things all affect the range of 5 gigahertz.
Now, it’s also less crowded because it’s physically bigger. There’s a lot more frequency space for us to work with up there. We have a lot more channels to work with, and so it’s a lot less crowded. But we’re starting to push up against the limits of even what 5 gigahertz can do. In a highly dense Wi-Fi deployment, we’re starting to get to the point where there’s really nothing left for us to use there.
And there might be some solutions coming to that soon that we’ll talk about in a little while. There is less non-Wi-Fi interference there because the range thing that I mentioned earlier and also because it just doesn’t go as far. You just don’t hear stuff nearly as well. I mean, for example, if I look at a Wi-Fi scanner here at my house, I can see dozens of 2.4 gigahertz wireless networks. I can only see four 5 gigahertz networks, and three of them are mine, because it just doesn’t go as far.
And so because of that, we see less interference. It’s crazy how things have flip-flopped where it used to be all about range, range, range. We all wanted range. But now I don’t want range. I want things to be– I want things to fall off really quickly so that we can avoid interference. That’s what I’m looking for.
OK. So 5 gigahertz supports– 802.11a was the very first 802.11 standard that came out in the 5 gigahertz band. We’ll talk about that in the history of things a little bit more in a minute. 802.11n, 802.11ac, and 802.11ax are all up in the 5 gigahertz band. Now, one thing that you have to be careful of, one thing to watch out for is when you are shopping for wireless devices.
If you see something that says 802.11b/g/n, you might go, oh, cool. It has 802.11n. 802.11n works up in the 5 gigahertz band. That must be something that supports both 2.4 and 5 gigahertz. That must be a dual-band device. [GRIMACES] Nope. That’s not. 802.11n can operate in either the 2.4 or the 5 gigahertz band.
Just because it’s an 802.11n Wi-Fi adapter does not mean that it has to support the 5 gigahertz band. Now, if we find another device and it supports a 802.11a/b/g/n, 802.11a is a 5 gigahertz only standard. And so that is a dead giveaway that that device supports the 5 gigahertz band.
And you know what one of my favorite things about 802.11ac was? Is that it forced a whole bunch of cell phone manufacturers and things like that, it forced them to put 5 gigahertz chipsets in their devices and get us support in the 5 gigahertz band because 802.11ac was a 5 gigahertz only standard. Really, really nice that they did that.
– OK. Cool. So let’s move on. And the next thing that I want to talk about is I want to take a little bit of a look at the history of 802.11, where it came from, what happened at the different points in time. And a lot of people– ask me a lot of people when I first start teaching this, they go, why are we talking about this old stuff? Let’s talk about the new stuff.
Well, here’s the thing about 802.11. The thing about 802.11 is that the IEEE is obsessed with backwards compatibility. The reality is that you can take a modern 802.11– you can take a brand new 802.11 ax device, and you should be able to associate it to an old 802.11b access point. It’s probably going to work.
You might even be able to associate it to an original 802.11 Prime access point. I haven’t tested that. Love to get an 802.11 Prime AP to test that. Haven’t found one yet. But we have to talk about where Wi-Fi came from. We have to understand the original standard. Because the majority of time, that’s what we’re using.
If you look at a packet analysis tool, if you capture a bunch of wireless traffic on your network– maybe at your house right now. If you captured some traffic, you would find that a vast majority of the time, you would be seeing data rates. You would be seeing protocol data that is 802.11 Prime, the original standard. You would see that.
So we have to understand how this works. We have to understand where Wi-Fi came from to understand where it is now and where it’s going. So let’s just step through this really quick and just look at, OK, what happened over time? So the very first place that we started was with 802.11 Prime.
And you could call just 802.11 if you want. Most people like to call it 802.11 Prime because it’s a nice differentiator. It says, oh, we’re talking about the original standard here. This came out in 1997. It was only down in the 2.4 gigahertz band, and it had a maximum signaling rate of 2 megabits per second.
Now, notice the word that I used there. Notice I didn’t say data rate. I mean, technically, you could say data rate. But the problem with that is people think, oh, 2 megabits per second, that means I’m going to get 2 megabits per second, right? 9,600 megabits per second. That means I’m going to get that, right? Well, no. Not with Wi-Fi. We’ll talk about that more a little bit later on.
This is the signaling rate. This is not throughput. That’s not what we’re looking at here. OK. So we started out– and I distract myself way too often here. We started out with 802.11 Prime 1997 in the 2.4 gigahertz band. 2 megabits per second was our maximum signaling rate. And it used a modulation scheme called DSSS, or Direct Sequence Spread Spectrum. That’ll be on the test later, so remember that.
And it had a total channel width– how wide the channel was was 22 megahertz wide. So that’s the actual width of each channel in an 802.11 Prime network. So then a couple of years later– and by the way, I believe– somebody could check me on this somebody– somebody that’s been doing this a little bit longer than I have– but I believe that this was originally intended for wireless barcode scanners in warehouses.
If somebody knows whether that’s true or not, I’d love to hear it. Put it in the chat. That would be awesome to know whether that is true or not. But I believe that’s what it was for, was to unify communications for barcode scanners. So then in 1999, the IEEE refreshed the standard with a new amendment called 802.11b.
So 802.11b, 1999, and was only in the 2.4 gigahertz band. And so what they were able to do with that is maintain backwards compatibility. It was fully backwards compatible with the original 802.11 Prime standard. But now we had a maximum signaling rate of 11 megabits per second. Now, how did they make it faster? Well they introduced a new modulation scheme called High-Rate Direct Sequence Spread Spectrum. So there’s that little high rate in there, HRDSS.
Same modulation scheme– my lights keep turning off. That’s really annoying. Home automation? I don’t know. I’m not so sure it’s the best thing out there. High-rate direct sequence spread spectrum got us 11 megabits per second. Now, at the same time on the exact same piece of paper, they also introduced a new standard called 802.11a.
Now, what makes 802.11a interesting is that it was up in the 5 gigahertz band. This was totally new spectrum. Completely new slice of spectrum that nobody had been using for 802.11. And so they could start from scratch. They could reinvent the wheel here, and so they did. They introduced a brand new modulation scheme called OFDM. That stands for Orthogonal Frequency Division Multiplexing. You guys got that, right? Because again. That’ll be on the test later on.
But the big thing to keep in mind with OFDM, or orthogonal frequency division multiplexing, is that it supports up to 54 megabits per second. So it’s a lot faster. And it also had a slightly narrower channel of 20 megahertz wide. So you can see how OFDM, 20 megahertz wide, whereas, DSS was 22 megahertz wide, and the HRDSS was as well. You can see that relation starting to happen there.
OK. So now we have a bit of a problem. We’ve got two competing standards– 802.11b, which has nice range. We would see about a 300-foot range with 802.11b, but it’s slow. It’s only 11 megabits per second, which, I mean, for 1999 wasn’t too bad. Whereas 802.11a only had a range of around 90 feet.
Remember, that’s going to change a lot depending on your environment. But it only had a range of about 90 feet, but it was a lot faster at 54 megabits per second. So we have a format war. We have a competition going on here. This kind of reminds me of HD-DVD versus Blu-Ray.
My brother-in-law had a big stack of HD-DVD movies. And when HD-DVD died and Blu-ray won the format war, well he’s got 20 movies in an HD-DVD player and that’s all the movies he’s ever going to have for that player. Looking back a little bit further, I remember my parents had a Betamax player. They had VHS. They had Betamax. And they had like movies on Betamax. And VHS ultimately won the battle.
That’s kind of what happened with 802.11b and 802.11a. 802.11b made more sense because the range was better and chipsets were cheaper because we’d already been making chipsets for a while now. And 802.11a and– whereas 802.11 was faster, but the range wasn’t as good, so it was more expensive. And so 802.11b kind of ended up winning.
So then what do we do? Well, we want that fast range down in 2.4. But we want that fast speed down at 2.4, rather. But we want the range as well. So what do we do? Well, the IEEE then introduced 802.11g. And I think this is where a lot of us came into the picture here. This is where things really started to pick up. This is back in 2003.
And what they did is they made it work down to 2.4 gigahertz so it’s backwards compatible with 802.11b. And they brought down that fast, swanky modulation scheme, OFDM. they renamed it slightly, ERP– extended rate physical. OFMDM. Don’t worry. Memorizing that stuff is not important. And it had a maximum data rate of, you guessed it, 54 megabits per second.
So what we’ve done is we’ve taken that 802.11a technology and brought it back down into 2.4 and kept it backwards compatible with b. And so now we get the best of both worlds. So a whole bunch of time passed before 802.11n came into play. We got 802.11n in 2009. And one thing that’s really interesting that the IEEE did with this is they supported both 2.4 and the 5 gigahertz band with the same standard.
Now, that doesn’t mean we use both at the same time. That means we can use one or the other at the same time. Now, they called it HT, or high throughput. Because in theory, we thought we were going to be able to hit about 600 megabits per second, but we never really did hit that in the real world. It never quite did pan out, at least not that I saw.
And we also introduced a couple– we also introduced a new channel width. We could go from a 20 megahertz channel to a 40 megahertz channel. We’re going to talk about that a little bit more coming up. But we maintain backwards compatibility with 802.11g and 802.11a at the same time. So this was the unifying standard that brought everything together.
2013, we saw 802.11ac come out. This was a 5 gigahertz only standard. So while it was backwards compatible with 802.11n in the 5 gigahertz band, it was not backwards compatible with anything down in 2.4, only 5 gigahertz. We call it this one Very High Throughput or VHT. The theory was, seven gigabits per second? [GROANS] It didn’t pan out. We definitely did not see that.
But this standard focused on overall speed. We tried to see how fast can we go to a specific client device. And one of the ways that we did that was by introducing wider channels. We’ll talk about that some more in just a couple of minutes. Now, the new one that we’re all waiting for is 802.11ax. That’s the one that we’re all waiting to be ratified. Notice it is not ratified, at least as far as I know. It’s not ratified by the IEEE, so it’s not done yet.
It’s really, really close, though. And we are offering 802.11ax access points. Many vendors are. But we should see ratification on that very, very soon. What we’re doing with 802.11ax is very different, though. Instead of focusing on going as fast as we can to one client device, we’re focusing on being efficient within the coverage cell. And that’s going to be very, very interesting.
So we’ll probably do something dedicated on that in the future. We’ll talk about 802.11ax more in the future. Recently, the Wi-Fi Alliance added some branding to these. They called 802.11ax ax Wi-Fi 6. And then they applied– they went backwards and they said, OK. Well that, makes 802.11ac Wi-Fi 5. And that makes 802.11n Wi-Fi 4. And they didn’t go any further than that.
So you can infer from there what things look like. But you don’t see 802.11g stuff in the wild very often anymore today. So they didn’t worry about giving names for that stuff. OK. Cool. So now we’ve looked at– now that we’ve looked at the standards over time and how they’ve evolved–
– Well, let’s focus in on 2.4 gigahertz a little bit more. Now, you remember earlier, I mentioned that all of our channels are 20 megahertz wide. That’s the actual width of our channels. That’s true in both 2.4 and up in the 5 gigahertz band.
But you’ve probably– you’ve probably noticed that in a router configuration or something like a home wireless router configuration, maybe some vendors out there, you may have noticed that there’s actually a whole bunch of different channels to select from. You can choose channel one, channel five, channel nine, channel 11. But you’ve probably always heard people say, but yeah, but never, never, ever use anything more than channels one, six, and 11.
Why do we say that? Why do we say to only use channels one, six, and 11? Well, here’s the deal with 2.4 gigahertz. When we look at what we actually have to work with, when you look at each channel is 20 megahertz wide, so that’s our channel width, and the total space that we have, at least in the United States, to use for Wi-Fi is only about 60 megahertz wide. That’s how wide the band is, at least usable for Wi-Fi.
And so when you do the math there, you go, OK, there’s 11 channels, but it’s only 60 megahertz wide. And there 20– hang on. That math doesn’t work out at all. That math totally does not work out. Well, the reason for that is because in the 2.4 gigahertz band, Wi-Fi channels are marked by their center.
So if I draw a line up channel one here, there’s the center of channel one. If we go– if we want to find the center of channel two, all we have to do is go up five megahertz, so we just go up five megahertz, 1, 2, 3, 4, 5, and mark the center of channel two. And what you realize is that channel one and channel one overlap by about 75%.
There’s overlap between channels in the 2.4 gigahertz band. As it works out, the only three channels that do not overlap are channels one, six, and 11. Those are the three channels that can stand by themselves, and they don’t interfere with each other. They don’t overlap with each other.
And so that’s why you always hear people say that the golden rule of Wi-Fi is to only use channels– is to only use channels one, six, and 11. So the next question that you might have is, OK, why does it matter? Why does it matter if we are on partially overlapping channels? Well, let’s talk about that for a second.
So I think I got my slides out of order here. Here we go. Problem solved. So if we have to– if we have two access points, let’s say that we’re looking at– maybe we have an access point on channel five. So here he is right here. And we have an access point on channel seven. Here he is right here.
Those two are on partially overlapping channels. They overlap by about 75%. Now, when two networks overlap, remember this is not just APs, but this is access points. But this is also client devices as well. When they partially overlap, instead of politely taking turns, they’re just going to yell and scream over each other, and they’re going to corrupt each other’s transmissions, which is going to cause all kinds of bad stuff to happen.
And so because of that, we always, always, always, always only want to use channels one, six, and 11. We’ll talk about this a little bit more in depth here in a couple of minutes. But before we do that, I want to talk about the most important thing in Wi-Fi. If there was one thing that you get out of this presentation today, I want this to be it. I want this to be the thing that you take away.
The core thing that you need to know about Wi-Fi is that Wi-Fi is half duplex. Now, I’m sure you’re familiar with ethernet. With an ethernet cable, we have multiple twisted pairs of copper inside the cable, and that allows us to achieve full duplex connectivity. What that means is that we can have a switch or something like that on this side.
We can have a machine or a computer or something over here on this side. And those two devices can both talk at the same time. We can send traffic in this direction, we can send traffic in this direction at the same time. That is full duplex connectivity. It’s kind of like a two lane highway. We can send traffic East and West simultaneously.
I grew up pretty close to Highway 101. And yeah, you can– in most places, Highway 101 is two lanes. And so you can go North and South at the same time. No big deal, right. Wi-Fi is different, though.
Wi-Fi is half duplex, which means only one device can transmit on a channel at a time. On any given channel at any given time, one device can talk. And so let’s say that we have a– let’s say we have a MacBook here. And this MacBook needs to send some data to this access point.
So the first thing it’s going to do is it’s going to listen on the channel to see if anyone else is talking. And if no one else is talking, then it’s going to go ahead, and it’s going to transmit its data. Now, how does the laptop know whether the access point received the data or not? The laptop doesn’t know. It just transmits it out there into the wild, and it has no way of understanding whether the access point receive the data or not.
So what the access point does is the access point will transmit back an acknowledgment. It’s an act. And so it says, hey, access point, I received your data. I was able to decode it. It all make sense. Yep, you’re good. I understood that data that you sent.
Now, if there’s something else, let’s say there’s another access point, maybe over here somewhere. Let’s say there’s another access point. And that laptop, if that access point is transmitting to another client, if that laptop listens and it hears that something else is transmitting, it’ll just back off. It’ll just back off and wait for an opportunity to talk.
So we can only have one thing talking on the channel at a time. Now, there’s a couple of things that can go wrong here. There’s a few things that can go wrong with this. Let’s say that the laptop transmits its data, but the data never makes it to the access point. Maybe I don’t know. Maybe the access points a little bit too far away. Maybe there is some interference, something like that.
And by the time the data gets to the access point, it’s just garbled up. And the access point goes, I don’t understand that. So what’s the access point going to do? Nothing. It’s not going to do anything. It’s just going to chill out and just wait.
And so what happens here is that soon, the laptop is going to go, huh, OK. Well, I never heard an acknowledgment from the access point, so it must not have received the data frame. And so it’s going to retransmit its data. It’ll try to send it over, and over, and over again until finally, one of those data frames makes it through to the access point, and the access point can go, yep, check, I understood that data, and it will reply with an acknowledgment.
So I have three daughters. The oldest is eight, and the youngest is five. And they actually do this all the time. If I’m sitting on the couch or something like that, reading something, or looking at something on my iPad, or something like that, if my five-year-old comes along and starts talking to me and if– basically, what happens is– well, first off, she doesn’t wait to see if the channel was clear. She just starts transmitting immediately because she hasn’t figured out back off mechanisms and things like that yet.
But if she says something to me, and if I don’t reply right away, then she’s going to retransmit over, and over, and over, and over again until finally, something makes it through to me. And I go, oh, yep. Check. And I go, oh, yes, honey. And then she’s going to have a lot more data frames to send.
Any of you with kids, any of you with daughters, you understand how that is, right? And so it’s a constant process of sending data, receiving it, and an acknowledgment. Now, notice here that I’ve only been drawing arrows going from the laptop to the AP. You can flip these arrows back and forth.
It doesn’t matter which way the traffic is going. It behaves exactly the same. We wait for the channel until the channel is clear. We transmit our data. We receive an acknowledgment. And we do that over, and over, and over again. We will get to that more in depth as we go.
This has huge implications for Wi-Fi. Some massive implications. This basically ensures that only one device can transmit on a channel at a time. And there are some massive implications for this. Let’s take a look at a larger scenario here. Let’s take a look at a coverage cell.
So let’s say that we have our access point in the center here. And we have an iPhone, and a MacBook, and an iPad, or something and another iPhone over here. We’ve got a few client devices. And let’s say that they’re all on channel one. They’re all on the same channel.
Basically, what happens here is that whenever one of these devices is transmitting, everything else on the channel has to shut up, and be quiet, and wait for the opportunity to talk. Only one thing can talk at a time. Now, one thing that happens in wireless too is that we can have different data rates. We’ll talk about this a lot more later on. So we’re going to go with very basics for now.
But when we’re close to an access point, we can hit some pretty fast data rates, like really fast. I’m just going to take a look at 802.11g, just for the simplicity of data rates here. But this is exactly the same for 802.11n, ac, and all the new data rates they introduced. So let’s say that we’re really close to an 802.11g, access point.
We can hit 54 megabits per second. We can talk really fast, and we can get responses back really fast. And the devices, on the other end, since we’re really close to each other, we’ll be able to understand each other very, very easily. But as we move away from the access point, as our signal strength starts to get worse, then it’s going to be more and more difficult to understand what the access point is saying, and it’s going to be more difficult for the access point to understand what we are trying to say.
And so what we have to do is we have to use a lower data rate. This is just like when you’re talking to someone in a quiet room. Right now, hopefully you can hear me really well. Hopefully, this microphone is working pretty good and that you can hear me pretty well. And so I can talk pretty quick and get a lot of information to you really, really fast.
But if we were trying to have a conversation in a really noisy room, we have to talk slower. We have to enunciate more to help the other person understand what we’re trying to say. Wi-Fi devices, exact same thing. They do the exact same thing.
So as we move away from the access point, we have to drop down to 48 megabits per second. And then down to 24 megabits per second. Then down to 12. Maybe all the way down to one megabit per second when we’re way, way on the edge of the coverage cell. And when that happens, we talk slower.
And thus, we consume more time on the channel. And so maybe this iPhone over here. Maybe it’s so far away from the access point that it’s had to slow way, way down to a really, really slow data rate. Maybe this iPhone is quite a bit closer, so it’s able to maintain 54 megabits per second.
This iPhone is going to talk 54 times faster than this iPhone. And thus, it’s going to consume 54 times less time on the air. This device that’s only talking at one megabit per second, I like to think of this device as the slow kid in the back of the classroom. Because you know what I’m talking about?
There’s always this guy in every class you’ve ever taken, who every time he raises his hand, you’re like, oh, no. Here we go again. And he goes, uh, yeah, teach. Yeah, I have a question about this particular– and you’re like, dude, just spit it out. Oh, my gosh. He grinds the entire class to a halt because he asks his questions very slowly. It’s just the way he talks or something like that. We’ve all experienced that, right?
That’s what this device is. A device that talks slowly keeps everyone else from talking, and so it consumes more time on the channel. Now, this is also true of adding more devices to a channel. If you have three or four people in a conversation, everybody gets an opportunity to say what they want to say, right. We can all get an opportunity to say it we want to say. Everybody’s involved in the conversation. It’s really nice, right.
But if you have 10, 15, 20, 30 people in a conversation, well, everybody’s opportunities to talk get smaller, and smaller, and smaller the more devices you add to the conversation. The exact same thing is true with Wi-Fi in regards to co channel interference, and a half duplex Wi-Fi, and all that stuff. The more devices you add to a channel, the smaller and smaller everyone’s transmit opportunities get.
There’s only so much time on a channel, and we have what we have to work with. And there are some things that we can do to speed up those conversations, to keep everyone talking quickly. But when you run out of channel, you run out of channel. That’s it. There’s no more time left on the channel. Cool.
– So then we talked about adjacent-channel interference. Now basically, what I’ve outlined with co-channel interference is that, yes, we have to control access to the medium. Everything has to wait for an opportunity to talk, but this is an organized process. This is built into the protocol. This is how the protocol works.
The problem with adjacent-channel interference is that it breaks the protocol. It breaks it because instead of being on the same channel where we politely take turns talking, instead devices just yell and scream over each other and corrupt everyone’s transmissions. Skip says this is why anticipated airtime utilization is a critical design factor.
Yes, Skip, I want to talk about airtime utilization in this presentation so bad. I don’t know if we’ll have time, but I really wanted to. Maybe in an upcoming one, we can talk about that specifically because that one is very near and dear to my heart.
So if we have an AP on channel 6 and an AP on Channel 7 and maybe the AP on channel 6 starts talking and then the AP on Channel 7 starts talking, it is going to interrupt that transmission. It’s going to corrupt the transmission. The access point won’t hear the acknowledgment.
This guy probably won’t hear the acknowledgment either, and they’re both going to have to re-transmit, which just makes the problem worse. And so it just goes downhill from there. So this is why we always want to use channels 1, 6, and 11. And what’s scary is there are some Wi-Fi vendors out there that will do this on purpose.
They’ll actually initiate adjacent-channel interference on purpose, which is really nasty, and I really don’t like that. Try to avoid that at all costs. Miss, we only use channels 1, 6, and 11 worldwide. That’s what we use, and that means a good Wi-Fi for everybody. OK, cool.
So when we’re talking about co-channel interference and how we can only have a certain amount of devices on channel, there’s only so much time available on the channel to talk. Well remember that we only have three channels to work with down in the 2.4 gigahertz band.
The problem with that is that what happens when we have more than three access points? Then we only have three contentions to work with, and that’s a big problem. So what do we do about that? Well, what we have to do is we have to perform effective channel re-use.
So let’s just pretend that this is like a bird’s eye view of all the access points, all the access points in a building. And it just so happens that all of our coverage cells are these neat little circles. They’re just these pretty little circles that all overlap a little bit. This is never one it’s real and like in the real world by the way.
Coverage cells? They’re never pretty circles like that. They’re not hexagons. They’re not octagon. No, they’re nothing like that. They’re always just these big random amoeba blobs. But for the sake of simplicity, let’s just pretend that they’re little circles here.
So what we need to do is we need to keep the channel 1 coverage cells as far away from each other as possible, so that they don’t have to take turns talking. If we put two access points on channel 11 right next to each other, those two access points and all of their client devices that are connected to them are all going to have to take turns talking.
They’re all going to have to take turns, because they’re all in the same channel. So what we do is we effectively re-use channels. We alternate between channels 1, 6, 11, 1, 6, 11, 1, 6, 11 as we move across the building, so that way we try to keep our same channel access points from being able to hear each other.
And since they can’t hear each other, then they don’t have to take turns talking, and so that’s really nice. The problem is that in the 2.4 gigahertz band, remember, the range is really good. Oh, it ruins it, because the range is so good in the 2.4 gigahertz band. Because we only have three channels to work with.
It’s very, very difficult to keep our same channel coverage cells from being able to hear each other. It’s almost impossible. So you will never achieve a perfect channel plan in the 2.4 gigahertz band no matter how hard you try. You’re welcome to take a track at it, but I’ve never been able to do it.
Maybe somebody here has. I don’t know. So seed oil says, hey, is there no channel 13 in Japan? That is a great question. So basically, I’m only showing you the US channels here, but I think it’s probably worth mentioning that there are more channels over in Europe.
We also get channel 12 and channel 13, and so they look something like this. And then way up above here, it’s a little bit further up here, so I can’t draw the whole thing. I run out of space on the slide. But there is also a channel 14 up here by itself in Japan, but that is only available for 802.11b use.
We can only use– remember the modulation scheme, it’s called HR-DSSS or high rate direct sequence spread spectrum. That’s a data rate of 1, 2, 5.5, or 11 megabits per second. As far as I know in Japan, channel 14 is the only place that channel is available. And we can only use 802.11b up there. So yeah, a bit of a problem.
So good to know about, but probably not something that most of us will interact with. So we have this problem where 2.4 is way too tiny. There’s not enough channels to work with. It’s a big problem. What do we do? Well–
– Fortunately, the 5 gigahertz band gives us a lot more space for a lot more channels and, thus, a lot more throughput. Now, you remember how I said that on any given channel at any given time, only one device can talk. That means that each channel has a limited amount of capacity inside the coverage cell. And 5 gigahertz really helps this out a lot. Because up in the 5 gigahertz band, we have a lot more channels to work with. In fact, we have 25 channels to work with.
Now, don’t be fooled by the scale of this slide. These are all still 20 megahertz wide just like in the 2.4 gigahertz band. I just have to make them smaller to make them fit on the slide and so does everybody else. So they’re always represented, they’re always represented like this.
Now, the 5 gigahertz band is split up into a few little sub-bands called the UNII bands. It stands for Unlicensed National Information Infrastructure. You don’t need to remember that. But what you do need to remember is that there’s all these little separate frequency bands up in the 5 gigahertz band and they kind of have some different rules associated to them that we have to be aware of.
– Now UNII 2 and UNII 2 Extended. These are the two that we need to pay close attention to, because these are home to the DFS channels. DFS stands for dynamic frequency selection. Selection– there we go.
Dynamic frequency selection. What that basically means is that, on these channels, it’s possible for Doppler weather radar to exist on these channels. Specifically, Terminal Doppler Weather Radar at airports.
And I don’t know, somebody decided that not crashing airplanes is more important than Wi-Fi or something like that. I don’t know, I think that’s kind of dumb, but whatever. And so, we cannot operate Wi-Fi on these channels without checking to see whether there is radar present on that channel first or not.
So what we have to do is– let’s say that you take a missed AP, and you add it to your dashboard and everything. You get it all on-boarded and it comes online. The very first thing that has to do if it’s going to broadcast on a DFS channel, is it has to listen for 60 seconds. It has to just chill out, listen on the channel, and see if it hears any radar on the channel.
If it hears radar, then it’s going to go to one of our non-DFS channels. Somewhere down here in UNII 1, somewhere up here in UNII 3. It’s going to have to go to one of those channels instead.
If it’s already serving clients on that channel. If the AP is already online, it’s already talking to clients and all that stuff– if it hears a radar ping, then what it’s going to do is it’s going to basically transmit a channel switch announcement. It’s going to say, “Hey, all the clients on this channel, I detected radar, which means we are moving in 5, 4, 3, 2, 1,” and then, we move to a different channel to get away from the radar.
I remember one of my good friends in the wireless industry, we were giving a presentation at a conference down in Las Vegas, and we need an access point on channel 120. We need an access point on that channel, specifically, because we needed to do some stuff with Wi-Fi during our presentation, and we really needed it to work, and that was the only channel that was clear.
The conference Wi-Fi was using every other channel except for channel 120. So we’re up in our hotel room in the Mandalay Bay, trying to configure our AP. And we configure channel 120, it doesn’t show up. We’re like, what’s going on? We reconfigure it, it doesn’t show up.
We’re tearing our hair out trying to figure out, why won’t this stupid access point show up on 120? And then all the sudden, we turned over and looked at the runway at the airport. And we’re like, oh, we’re right next to an airport, there must be radar on channel 120. No wonder the conference isn’t using 120. There’s radar on that channel and so they are not able to use that.
So you have to be careful. Just be cognizant of this. It doesn’t affect everyone everywhere. It just means that you might lose a channel here and there and you might need to figure out which channel your stuff is on. Fortunately, Mist will do this stuff automatically– will automatically work out channel plans and all that, which is really nice. And it will of course, automatically avoid all of this stuff.
Let’s see– question here from [INAUDIBLE] is how will the AP know if the DFS utilization is a radar or another access point transmitting? That’s a great question. Basically, APs have a set of signatures that they have to look out for.
I believe is the FCC– it’s either the FCC or the FAA, I’m not sure which– basically, it gives access point vendors a set of signatures, and says, hey, if you hear any of these, you need to move. We don’t actually know which one is radar. It doesn’t matter. We just have to stay away if we see one of those signatures.
Some vendors do have false positives. I have a deployment with another vendor here in Boise for a nonprofit organization, kind of a budget vendor. And it gets DFS hits all the time. We don’t have terminal Doppler weather radar and 5 gigahertz here in Boise, so I have no idea where that’s coming from. I’m pretty sure it’s a false positive. So it kind of depends on the vendor. Some do it better than others, but that is subjective. That’s just my opinion about that.
– Cool. So let’s go ahead and talk about channel bonding. This is one that’s really important to talk about. Now basically what channel bonding is channel bonding was introduced with 802.11n. And that’s where we take two 20 megahertz wide channels, and we just mash them together to make a 40 megahertz wide channel.
And so instead of a two lane highway, now we have a 4 lane highway, and so we can get faster data rates. This just basically allows us to hit faster data rates on our wireless. Now in sort of an end days, we call this channel bonding. So two 20 megahertz channels are bonded together and a wider channel gives us higher data rates.
I don’t really like to call it channel bonding when I’m talking about 802.11ac though. Because of 802.11ac, we have a lot more options for a lot more different channel widths. Of course, we can still use the old standby, the 20 megahertz channel. That’s what we were just looking at in the other chart. Or we can do the 40 megahertz channel.
But they also added a new 80 megahertz channel. So this is where we take four 20 megahertz wide channels, and we mash them all together to get this massive 80 megahertz wide channel. There’s also a 160 megahertz wide channel out there as well. But seriously, there’s no room for us to use 160 megahertz wide channels.
It’s in the standard I’ve never found an AP that actually supports using it. Here at Mist, we don’t allow to use it. There’s just no environment where that would make sense, and so we just kind of keep that turned off. It’s not available. So with a wider channel, we get faster bandwidth, which is really nice.
And the way that these channels are organized is we basically have primary and secondary channels. So for example, when you set up an 80 megahertz wide channel, you could say OK, channel 40. That’s going to be my primary channel. And then 36, oops, I forgot this switch this should be 44, and this one should be 48.
Just ignore that, is no big deal. Just pretend that was something else. But when we have channels 36, 44, and 48, these are all secondary, and the channel 40 is the primary channel. And so this is where all of our traffic kind of defaults to. This is where all of our management overhead goes.
But if we need to branch out and use the wider channel to move a ton of data really, really fast then we can do that. It’s really fun to do this and look at a device like a spectrum analyzer which will actually show you these shapes happening in the spectrum while they’re happening.
It’s really fun to see like, OK, you’ll see these little 20 megahertz wide blips. And then you start off a throughput test or a file transfer, and all of a sudden, you get this big wide shape that takes up all of this channel width. So when should we use wider channels?
Well how about in 2.4 gigahertz, is that a good idea? Should we ever do that? No. No, there’s no space down in 2.4 gigahertz for us to use a 40 megahertz wide channel. If we did a 40 megahertz wide channel, that’s 2/3 of the spectrum, boom, eaten up gone. We don’t want to do that.
Five gigahertz though is a little bit more permissive. We have a lot more space up in the 5 gigahertz band. And so the way that I look at this is I say use wide channels until you can. Use wide channels until you can’t anymore.
In an environment where we only have a handful of APs, were there’s not a lot of neighbors things like that, yeah, 40 megahertz channels they totally make sense. Although 80 megahertz channels in enterprise, you’re almost never going to see that. But yeah, use it use 40 megahertz in lower density environments.
At some point, you’ll run out of channels to reuse. You might have to fall back down to 20s, and don’t ever use 160 megahertz wide channels if you can even find an access point that supports that. So let’s take a look really quick at what happens with the wider channels.
So here on the bottom, these are all of our 20 megahertz channels we have. We have 25 of them available. You can see that we basically cut that in half. We move down to 12 channels if we use 40 megahertz wide channels. 80 megahertz, it’s the same thing here. Just ignore that.
But I had to get all my channels numbers assigned in here. I knew there was something I was missing on this slide deck. But don’t worry you can find examples of this online very, very easily. Maybe we can fix this and publish it later on.
And then here are 160 megahertz wide channels. We only get two of them, and so they’re just isn’t space to use them. And so my recommendation is don’t. So use 40 megahertz channels until you can’t. That’s what I recommend. That’s just my personal recommendation for that.
Now re-using channels up in the 5 gigahertz band, way easier. Super duper easier to do, because we have so many more channels to work with even if you’re using 40 megahertz wide channels. For example, a channel 36, there’s no other channel 36 is on this network that I’ve got here.
Channel 100, yeah, there’s no other no other APs on channel 100 here, so everybody gets their own channel. It really depends on the density of your environment. If you’re doing like a single story building, it’s very, very easy to do a channel plan in the 5 gigahertz band. Super, super duper easy.
Once you start moving to multiple floors though and if you have a venue or something like that with a whole bunch of APs in the arena something like that, you can run out of channels really, really fast. I’ve done a few designs that it was very, very difficult to keep my same channel coverage cells from being able to hear each other in the 5 gigahertz band. It just depends on the density of the environment that you’re in.
– OK. So let’s talk about signal strength for a minute because signal strength is just one of those mystical things. OK. So signal strength is a mystical thing. And sometimes, it helps us to have a quick explanation OK, what exactly does this mean?
Basically for signal strength, we use decibels to represent changes in signal strength or a change in signal strength. And we use dBm to represent absolute values in signal strength like what is our signal strength at this location. That is represented using dBm.
So basically, dB is a very tricky scale. Because first off, it is not a linear scale. It is a log scale. So we have to abide by the rule of 3’s and 10’s when we’re thinking about changes in signal strength. So let’s say that you’re sitting at negative 70 dBm, and you go up to negative 67 dBm.
And you’re thinking like oh, hey, hang on a second. We went up to negative 67. We’ll remember, we’re working in negatives here. These are negatives. Negative 90 is a lower number than negative 30. And so Wi-Fi is transmit power is so low. Our receive signal strength in Wi-Fi is so low that dBm dips down into the negatives, and that’s where we live for everything in Wi-Fi.
So keep in mind, it’s inverted here. We have to think of this kind of backwards. So if we go up 3 dB, if we have three dB of gain from negative 70 up to negative 67 dBm, that is double the signal strength. If we go down from negative 70 dBm down to negative 73, that’s 3 dB of loss, then that is half the signal strength.
If we go down again to negative 76, we’ve just cut our signal strength in half again and so on and so forth. Every time you go down 3 dB, you cut your signal strength in half. Every time you go up 3 dB, you double your signal strength. That’s the rule of 3’s.
Similarly, the rule of 10 says if you go up 10 dB, that is 10 times the signal strength. So if we’re sitting at negative 60 dBm and we go up to negative 50 dBm, that’s 10 times more signal strength. If we go down to negative 70, that’s 10 times less signal strength. You see what’s going on there.
Now you might be thinking to yourself like, so this is so confusing. Why would we use dBm and dB? Why can’t we use something that’s a little bit more linear, a little bit more easier? Well the other one that we could use is milliwatts, but the problem with milliwatts is that Wi-Fi transmission power is so so teeny tiny.
It is so low that there’s a ton of decimal places involved. So let’s say that you’re on the phone with somebody and you’re troubleshooting a coverage issue, and you got somebody on the phone and they’re like yeah in this room, the signal strength, it’s pretty good actually. It’s 0.000001 milliwatts, and you’re like oh, OK, cool.
Well what happens if you move into the adjacent room? It’s like oh, it drops sharply. Now our signal strength is 0.00000001 milliwatts. Nobody’s going to do that. Oh my goodness, that is very difficult to keep track of. That is why we use dBm. Even though we have to think about the rule of 3’s and 10’s, it’s down the negatives, and there’s kind of some things to think about there, it’s way easier than using milliwatts to define signal strength. So that’s why we use it.
So what is a good signal strength, anyway? Well these are just totally my opinions. These are not set in stone. These are not in any books anywhere. These are Joel’s opinions, so please take them with a grain of salt. If you disagree with me on these, cool, that’s fantastic.
You should disagree with me on these because these are just my opinions. Allow your own experience and your own opinion to just change these to match your own opinion. Negative 30 dBm is like ridiculously high. That basically means that you’re standing underneath the AP or something like that.
Negative 65 dBm, that’s what most people consider to be good enough for voice over Wi-Fi applications maybe mobile devices. It’s a really nice high signal strength, so we can maintain fast data rates. We can transmit once and not have to re-transmit because of device on the other end couldn’t hear it, so it’s going to limit our re-tries.
That’s a really good thing. It keeps things very efficient. Negative 67 dBm is kind of the standard for streaming video, high speed data, that kind of thing. You can push it all the way down to negative 75 dBm for things like web browsing, email, instant messaging.
Those are all applications that don’t need a lot of bandwidth. Things don’t need to be delivered in a quick and timely manner, so that’s OK. Negative 80 dBm is where things are going to not work very well. At that point, your access points and your client devices are going to have a hard time hearing each other.
They’re not going to understand each other all the time. There’s going to be lots of re-transmissions. Once you hit about negative 85 dBm, that’s when most client devices give up and say, OK, this network sucks. I’m going to go find someone else. My Nokia phone that’s when it gives up is about negative 85 dBm.
It will just disassociate from the network even if it can’t find anything else to associate to. Negative 90 dBm is where you’re basically approaching the noise floor. Basically the noise floor is any background noise in the environment.
That’s basically, at that point, the signal strength is down below the background noise, and so we’re not going to be able to hear anything at all. We have background noise in the audio world, right? If you’re sitting in a room right now, if I stop talking for 5 seconds, you’d eventually hear the noise floor in your environment.
If there’s no one talking, if there’s nothing going on, there’s still some background noise there. Maybe it’s the HVAC, maybe it’s the heater, maybe it’s the fans on your computer running or something like that. It fades into the background noise. John says negative 95 is the standard for hotel Wi-Fi. That’s fantastic. I love that.
And Chris says negative 50 could be poor signal depending on SNR. Yes, Chris, you’re right. I’m going to get there. Hang in there. I will get to it. SNR is coming up really soon, but we’re going to introduce topics one at a time, and we’re building on it as we go. So cool. Cool.
– OK. So working through some more fundamental stuff. Next thing I want to talk about are the difference between a basic service set, an SSID, and an extended service set. And so let’s talk about these really briefly for just a couple of minutes here.
So a basic service set identifier is– essentially, it’s the MAC address of the access point’s radio. For example, a BSS might be, D4:20:B0:A2:EF:B7. That could be the MAC address of your BSS.
In fact, if you hold down the Option key– if you’re on a MacBook, if you hold down the Option key and click on the Wi-Fi logo– I’m not going to do that because I’m on my Wi-Fi right now, and I don’t want to keep my Wi-Fi adapter– I don’t want to tie it up scanning for their networks and mess up our communication here. But if you hold down the Option key and click on the Wi-Fi logo up in your menu, right up here on the top, then you’ll actually be able to see the BSS that you are in.
That is essentially– the BSS is one coverage cell that’s composed of the access point, the radio on the access point, and all the clients that are associated in that coverage cell. That is a BSS, or a basic service set. The BSSID is the MAC address of the radio on the access point.
Now, one question you might have is, well, why do you have two MAC addresses here? What’s up with that? Why do you have two MAC addresses there if that’s just one coverage cell from one access point? Well, remember that pretty much every modern access point out there has two radios for serving clients.
And so one of them might be on the 2.4 gigahertz band. The other one might be on the 5 gigahertz band. So one access point essentially provides two BSSes for devices to connect to. Just happens that they’re in two separate frequency bands.
Now, an SSID or a service set identifier, this one’s probably going to be really familiar to you. This is basically the friendly name of a wireless network. For example, it might be Aperture Science. That is the SSID that we could be broadcasting. That’s a friendly name that allows users to find the network. When we branch this over into an extended service set, an extended service set is multiple APs with the same SSID.
And so an extended service set, that involves all the access points that have the same SSID and all the clients that are connected to them as well. That is an ESS, or an extended service set. Really good to know this terminology. This is going to come in really handy in a few minutes here when we start talking about things like roaming. Now, really quickly, before we move on to the next section–
– I just want to touch on other wireless technologies very, very quickly. We’re not going to spend a significant amount of time on this, but I just want you to be aware of these. And we might be able to turn out some focused content for these in the future. First off is Bluetooth Low Energy or BLE. BLE operates down in the 2.4 gigahertz band,. It is not Wi-Fi. It is a different technology.
But it uses the same frequency space. It’s often used for things like location services or quote, unquote, “blue dot navigation” inside a building. And it is a current technology that is available today. And in fact, I think it’s one of the coolest things that Mist has done is we have awesome, awesome, awesome, really flexible, really interesting location technology that works with Bluetooth Low Energy.
It’s not our main selling point, but it’s really good. It’s a really, really cool technology. It’s just not the first thing that most customers are looking for when they’re deploying a network. But if you want location, Mist does it the best. We will talk about BLE a little bit more in just a second. Next is Zigbee. You remember I talked about my little remote here in my– let’s see. I can turn my Philips Hue Lights off, and I can turn them back on.
That is Zigbee. It also operates in the 2.4 gigahertz band, but the big thing to keep in mind about Zigbee is it’s very low power and very low throughput. The maximum throughput that we can hit with Zigbee is about 250 kilobits per second. Although, I think most devices out there really only do 100 kilobits per second.
But remember, when we’re doing something very simple like turn Joel’s office lights off, we don’t need throughput, and we don’t need a ton of bandwidth for that. We just need low power. This light switch has a coin cell battery in it. I have never replaced the battery in this, and I’ve had this for two years.
So it’s very low power, very low throughput. And it uses a mesh technology. It’s available now. It’s been available for a long time. Next, this is one that, to be honest, I don’t know much about. I need to do some research on this one. But CBRS or Citizens Broadband Radio Service is an up and coming technology that’s looking very, very interesting. It works up in about a 3.5 gigahertz range between 3.5, 3.7 gigahertz.
There’s a lot of different channel widths available 5, 10, 15, 20, 40 megahertz channels. And what’s interesting about it is it uses dedicated clean spectrum with a tiered licensing model. This is where my understanding of this runs out really quickly. So we could probably follow up on this later on to talk more about CBRS. But it is an upcoming new technology.
But there is some implementation today in a few handsets like the iPhone 11, a couple of Google Pixel devices. And this is supposed to say Galaxy S10. I wrote S8 on accident there. Ignore that. Just another typo. Ignore that. But it is available in a limited fashion today. And what makes it interesting is, because it’s using dedicated clean spectrum, the way that CBRS devices get access to the medium is very different from Wi-Fi.
Wi-Fi is very Wild Wild West. CBRS is very organized. So we’ll probably talk about that more maybe in an upcoming training session or something like that. And then, of course, there is Wi-Fi 6E. Or as I just like to call it, just the 6 gigahertz band. There’s currently a lot of work being done on opening up the 6 gigahertz band for Wi-Fi. There’s potential for a lot of new channels, which would give us a lot more throughput and a lot more capacity.
And what’s really interesting about 6 gigahertz is– you remember back in the day when we introduced 802.11a in the 5 gigahertz band, right? We’ve got to start from scratch. We’ve got to do a new protocol, kind of do things from scratch, get rid of all the old legacy stuff. There is an opportunity for that to happen with 6 gigahertz, but it all depends on what the Wi-Fi Alliance and what the IEEE decide to do.
So we’ll see– we’ll see what happens. It’s an exciting thing though. And I’m really excited about it because 5 gigahertz is getting quite full. And having new spectrum to work with would be really, really great.
– Cool. So very briefly, extremely briefly, I just wanted to mention Bluetooth Low Energy and kind of how it works and where it lives. Bluetooth is traditionally a frequency hopping technology. Basically, when you take two Bluetooth devices like– I don’t know– a mouse or a phone, and you pair them to a computer, they pick a set of channels to hop around on. And so they don’t stay on a channel for hardly any time.
We’ll land on this channel. Then after one millisecond, will hop to another channel. Then we’ll hop to another channel. We’ll hop around all over the place. That’s how Bluetooth traditionally works. BLE for beacons uses Bluetooth channels 37, 38, and 39. And it stays put on those channels. But the big thing that I want you to notice here is notice that they are just barely outside of where our Wi-Fi channels are.
And so that way, we avoid any interference at all. There’s no potential interference issues. Although, the reality about Bluetooth is that it moves so quickly. It hops around so fast. It typically does not cause a noticeable impact on a Wi-Fi network even if it is using the same channels as Wi-Fi. It’s a very quick look at Bluetooth and how it can coexist with– how it can coexist with Wi-Fi.
– OK. Cool. So the next thing I want to talk about is the difference between an ethernet NIC and a Wi-Fi NIC. There are some really big differences between the two that we need to understand to really get more in-depth in this wireless thing, this 80.11 thing and see exactly how they work.
First off, let’s take a look at an ethernet NIC. How does an ethernet work– NIC work? Well, first off, it uses carrier sense multiple access collision detection. We can actually detect collisions on an ethernet network.
When an ethernet device is transmitting, it will periodically stop transmitting and listen to see if anything else is transmitting. And if it hears something else, it will transmit a jam signal and go, oh, we had a collision on the network.
In fact, I have an old dinosaur relic here. This is an old Lynx this workgroup hub. And I keep this around for doing packet captures just in case I need to connect thin devices to my twisted pair network. You never know, right?
How else am I going to get my Macintosh up there on the corner online, right? You’ve got to have one of these around. But this actually has a collision– there’s actually a collision LED right there. And so it will actually show when there are collisions on the network.
Now, on a modern network, we only– really, every ethernet NIC out there only receives the frames that are destined for it. We have a switch. The switch sends the packets down the right port down to the right device. And we only have to receive the frames that are destined for it. We also have a consistent data rate. If you have a 1 gigabit per second link, you’re probably going to see throughput very, very close to 1 gigabit per second.
And so as you’re thinking about this, as you’re thinking about ethernet NICs and how they work, one thing I want you to noodle on for a second, just think about this, how did we make ethernet faster? I’ll come back to that in just a second. How did ethernet get faster over time? Did the copper get better? Do they use higher quality copper? How did that work? Just think about that.
If we look at a wireless NIC though, a wireless Network Interface Card, things are very different over here in the wireless NIC world. We don’t use carrier sense multiple access collision detection. We can’t detect collisions.
Because when a wireless Nick is talking, it can’t listen. It can’t do both at the same time. It can only talk or it can only listen. It can’t do both. So we have to use carrier sense multiple access collision avoidance.
And so it’s very, very different. It’s all about avoiding collisions. But we cannot detect them. If it happens, we don’t know the only thing. That we might know is that we didn’t receive an acknowledgment back. That’s the only indicator that we have that there might have been some kind of collision.
A wireless NIC also has to demodulate as to decode and understand every frame that it hears. As to demodulate, decode the frame and go, OK, was that for me, as to listen in on the whole conversation before it understands whether it was for it or not? It’s very, very different.
The data rate also varies wildly based on signal strength, signal to noise ratio. If you haven’t heard that before, don’t worry. We’re going to get into that. How busy the channel is you. Remember, we talked about half duplex Wi-Fi. It just varies. It’s all over the place.
There’s huge variance in what our actual data rate is going to be, which is why when you go to Best Buy and there’s a– I don’t know– like TP link or something like that there and they claim 1.3 gigabits per second, you don’t know. You’re never going to hit that. And we don’t like to say that it missed either. We don’t like to throw around terms like that because it’s just not what’s going to happen.
Some people say that the actual throughput is roughly half of the data rate. That’s very rough. It can depend. It can be 60%. It can be 40%. It can be all over the place. But you’re really only going to get about half of the– half the data rate ends up becoming your actual throughput. And then when you add more clients to the coverage cell, that changes. So it’s all over the place. Throughput decreases as the number of clients increases.
So let’s go back and think about, how did we make ethernet faster? Did we get better copper? No. All the copper is the same. What we did– John says we charged customers more for it. [LAUGHS] Come on, John. Help me out here, buddy.
But we did it– what we did to make ethernet faster is we decreased interference. We changed things like the twist rate, how those pairs are twisted. We added shielding to the cable. And the twist rate specifically was to decrease crosstalk so we could talk faster on the cable and not interfere with ourselves, right.
Well, that’s what we need to do in Wi-Fi as well. We need to achieve better signal strength and thus, we need to achieve a better signal to noise ratio. And that way, we can use more complex modulation, more complex ways of talking, and then we can hit faster data rates for faster Wi-Fi. That’s how we’ve made Wi-Fi faster over time. That’s how we’ve made ethernet faster over time is to decrease interference in one way or another.
So signal to noise ratio defines how much signal there is above the background noise. So let’s say that our noise floor is down here. This is where all the background noise is. Maybe it’s hanging out at negative– maybe that’s like negative 95 dBm. That’s kind of what that looks like to me.
And then let’s just say for simplicity that we’re looking at negative 60 dB of RSSI. Well, basically, I have to do is subtract negative 95 from negative 60 and boom, there’s your SNR. Most people say that you want an SNR of 25 dB. That’s how much signal that we want to have above the background noise for things to talk fast, for conversations to be efficient, to limit re-tries. All that stuff. We want to shoot for about 25 dB.
It’s really funny we talk about signal strength all the time. We always say, oh, yeah, negative 67 dBm. That’s the signal strength that we want and blah, blah, blah. That’s not the important one. The important one is actually signal to noise ratio. It’s just a little bit harder to measure. And so we end up using signal strength. We end up using signal strength a lot more.
The reason why signal to noise ratio is more important than signal strength is because the background noise changes. This number right here changes depending on the environment that you’re in. Here at my house, it’s pretty low. It’s like negative 90, negative 95.
You get into a manufacturing environment where there’s lots of electric motors and things like that big AC, brushed motors, things like that, you can see noise floors come way up. You could see a noise floor all the way up at neg 80. You could see a noise floor all the way up, up to neg 75. It depends on the environment. I’ve even seen bad Romax in buildings cause the noise floor to come up.
USB-C. My USB-C dock is out of reach. But if you take a spectrum analyzer, which will show you that noise floor and put it next to a USB 3.0 device of some kind, you’ll see the noise floor come up. And so signal to noise ratio is what it’s all about in wireless.
Now, ultimately, signal to noise ratio determines what data rates will be able to hit. There’s a lot of different data rates out. There are a ton of different data rates to choose from.
Now, back in the 802.11 days– you heard me say this earlier– we had 1, 2, 5.5, 11 megabits per second. 802.11a and g we had six nine 12, 18, 24 all the way up to 54 megabits per second. If signal to noise ratio was bad, we would fall down to one of these data rates. If signal to noise ratio was good, we could talk faster, and so we could hit faster and faster data rates.
So if we move on to 802.1n and 802.11ac, then things change quite a bit. Because now, instead of just having data rates to work with, we have this. The MCS index. MC– I should have written this down on the slide, but we’ll just write it down here– stands for Modulation and Coding Scheme.
So basically, there are a lot of different data rates that were introduced with 802.11n. And there were a whole bunch of data rates that were introduced with 802.11ac. And then for 802.11ax, it’s about to get nutty. It’s like thousands of potential data rates that we can use.
But what we’re trying to do, as wireless network engineers, is we’re trying to help client devices choose data rates that go this way. We want to help them get to faster data rates. But if the environment is noisy, if there’s a lot of channel interference or adjacent channel interference, if things are really nasty, then data rates are going to naturally trend back this way.
So what happens is– by the way this is completely dependent from the– on the app and the client device. They don’t rely on each other to determine data rates or anything like that at all. Right now, my MacBook Pro is talking to my access point downstairs. Yes, I’m on Wi-Fi. This is not a good idea. I just couldn’t get ethernet working in time for this session today.
My MacBook is transmitting data frames down to my access point. And sometimes, they fail and sometimes, they work. And if we have a lot of failures, then the MacBook will go, ooh, that’s not good. And it will step back to a slower data rate to try to help the access point understand it to minimize re-tries .
On the other end of the equation, the access points doing the exact same thing. It’s transmitting frames back up here to the MacBook. And depending on how many of those transmissions are successful, how many acknowledgments it gets back, it will ratchet up to a faster or slower data rate accordingly.
Every device handles this differently. As far as I know, every device has a unique algorithm that they use to determine which data rate they want to use. I’ve heard of some devices. Like for example, I heard of one device that it would try at 144. It try it once. It try it twice.
It try it 3 times and it would go, well, that didn’t work, and it will fall all the way down to 13 megabits per– I’m sorry– it’ll fall all the way down to 14 megabits per second and stay there for forever. It will never try to creep back up and find a data rate that will work. It won’t slowly and gradually move down to slower data rates.
It’ll just go as fast as it can, well, that didn’t work, fast as it can, well, that didn’t work, and go as slow as it can from there on out. And so client devices all handle this differently. And so we have to be cognizant of that and think about that. But our job as a wireless network engineer is to try to give devices, good signal, good signal to noise ratio to help us hit as fast of data rates as possible.
Now, if you want to look at this chart– and I look at it all the time, constantly look at it– go to mcsindex.com or this version right here is really nicely formatted. This is on wlanpros.com. Nice website maintained by Keith Parsons. They’re both really nice.
So go check those out. mcsindex.com. That’s one that I keep bookmarked. That’s why I refer back to all of the time.
Basics 1.14- So signal to noise ratio defines how much signal there is above the background noise. So let’s say that our noise floor is down here. This is where all the background noise is. Maybe it’s hanging out at negative– I think that’s negative 95 dBm. That’s what that looks like to me. And then let’s just say for simplicity that we’re looking at negative 60 dB of RSSI.
Well basically, all I have to do is subtract negative 95 from negative 60, and boom, there’s your SNR. Most people say that you want an SNR of 25 dB. That’s how much signal that we want to have above the background noise for things to talk fast, for conversations to be efficient, to limit retries, all that stuff. We want to shoot for about 25 dB.
It’s really funny, we talk about signal strength all the time. We always say, oh yeah, negative 67 dBm. That’s the signal strength that we want, and blah, blah, blah. That’s not the important one. The important one is actually signal to noise ratio. It’s just a little bit harder to measure, and so we end up using signal strength. We end up using signal strength a lot more.
The reason why signal to noise ratio is more important than signal strength is because the background noise changes. This number right here changes depending on the environment that you’re in. Here in my house, it’s pretty low. It’s like negative– it’s like negative 90, negative 95.
You get into a manufacturing environment, where there’s lots of electric motors, and things like that, big AC, brushed motors, things like that. You can see noise floors come way up. You could see a noise floor all the way up at neg 80. You could see a noise floor all the way up up to neg 75. It depends on the environment.
I’ve even seen bad Romex in buildings cause the noise floor to come up. USB-C, my USB-C dock is out of reach. But if you take a spectrum analyzer which will show you that noise floor and put it next to a USB device of some kind, you’ll see the noise floor come up. And so signal to noise ratio is what it’s all about in wireless.
Now ultimately, signal to noise ratio determines what data rates will be able to hit. There’s a lot of different data rates out there. There are a ton of different data rates to choose from.
Now back in the 802.11b days, you heard me say this earlier, we had 1, 2, 5.5, 11 megabits per second. 802.11a and g, we had 6, 9, 12, 18, 24, all the way up to 54 megabits per second. If signal to noise ratio was bad, we would fall down to one of these data rates. If signal to noise ratio was good, we could talk faster, and so we could hit faster and faster data rates.
So if we move on to 802.11n and 802.11ac, then things change quite a bit. Because now instead of just having data rates to work with, we have this, the MCS Index. The MCS, I should have written this down on the slide, but we’ll just write it down here, stands for Modulation and Coding Scheme.
So basically, there are a lot of different data rates that were introduced with 802.11n, and there were a whole bunch of data rates that were introduced with 802.11ac, and then for 802.11ax, it’s about to get nutty. It’s like thousands of potential data rates that we can use.
But what we’re trying to do, as wireless network engineers, is we’re trying to help client devices choose data rates that go this way. We want to help them get to faster data rates. But if the environment is noisy, if there’s a lot of channel interference or adjacent channel interference, if things are really nasty, then data rates are going to naturally trend back this way.
So what happens is– by the way, this is completely dependent on the AP and the client device. They don’t rely on each other to determine data rates or anything like that at all. Right now, my MacBook Pro is talking to my access point downstairs. Yes, I’m on Wi-Fi. This is not a good idea. I just couldn’t get ethernet working in time for this session today.
My MacBook is transmitting data frames down to my access point, and sometimes they fail, and sometimes they work. And if we have a lot of failures, then the MacBook will go, ooh, that’s not good. And it will step back to a slower data rate to try to help the access point understand it to minimize retries.
On the other end of the equation, the access point’s doing the exact same thing. It’s transmitting frames back up here to the MacBook. And depending on how many of those transmissions are successful, how many acknowledgments it gets back, it will ratchet up to a faster or slower data rate accordingly.
Every device handles this differently. As far as I know, every device has a unique algorithm that they use to determine which data rate they want to use. I’ve heard of some devices. Like for example, I heard of one device that it would try at 144, it tried it once, it tried it twice, it tried it three times, and it would go, well, that didn’t work.
And it will fall all the way down to 13 megabits per– I’m sorry– it’ll fall all the way down to 14 megabits per second and stay there for forever. It will never try to creep back up and find a data rate that will work. It won’t slowly and gradually move down to slower data. It’ll just go fast as it can, well, that didn’t work, fast as it can, well, that didn’t work, and go as slow as it can from there on out.
And so client devices all handle this differently. And so we have to be cognizant of that and think about that. But our job, as a wireless network engineer, is to try to give devices good signal, good signal to noise ratio to help us hit as fast of data rates as possible.
Now if you want to look at this chart, and I look at it all the time, constantly look at it, go to mcsindex.com or this version right here is really nicely formatted. This is on wlanpros.com. Nice website maintained by Keith Parsons. They’re both really nice. So go check those out. Mcsindex.com, that’s one that I keep bookmarked. That’s when I refer back to all of the time.
– Another thing I want to do really quickly, very quickly, is give you a very quick introduction to the different types of frames that we transmit back and forth between access points and client devices. Now a lot of times, we accidentally call these packets. We’ll say, Oh, Yeah. Well, I fired up a packet analyser. I did a packet capture to capture a conversation between an access point and a client device.
When we say packets, we really mean 802.11 frames. Every now and then, somebody, if I say packets, they’re like, they’re actually frames. I’m like, I get it. I know, but we call it packet analysis. We don’t call it frame analysis, right?
But there are several different types of frames in Wi-Fi that we just need to be aware of. There’s management frames, control frames, and data frames, and they all have different jobs. Management frames help stations join and leave wireless networks. So when you get home and you want to connect to your network, management frames are what makes that happen.
So there are several different types of management frames. The first is a beacon frame. A beacon is simply an access point announcing that it exists.
It says, hey, I’m Joel’s network. I’m here. I support these data rates. I support this type of encryption, you can connect to me. And it does that 10 times a second.
Hey, I’m Joel’s network. I’m here. I support these data rates. And it just does that over and over, and over and over again.
The next one is kind of an opposite of a beacon. It’s called a probe request. That is where my phone will probe for familiar networks that it has connected to before. So if it’s not associated to a network, it’ll periodically go, hey, Joel’s network. Are you there?
Hey, Hilton Hotels. Are you there? Hey, SFO Wi-Fi. Are you there?
And then, if SFO Wi-Fi is there, it will reply with a probe response. It’ll say, Yes. I’m SFO Wi-Fi. I’m here.
And the probe response and the beacon are almost exactly the same thing. They’re just ever so slightly different so we differentiate between the two. But they’re basically the same thing.
So then, there’s some other types of frames, like an authentication request and an authentication response. That’s some exchanges that happen between the access point and the client device to get authenticated to the access point. And then the association request and association response, we are going to talk about these a little bit more in just a couple of minutes.
Next are control frames. Control frames, they control the RF medium and ensure that other types of frames get delivered. You probably recognize, switch pen colors, that’s just brutal. There we go.
You probably remember acknowledgments. Remember, we send data, we receive acknowledgment. Data, ACK, data, ACK, data, ACK. It’s a back and forth process. The acknowledgment says, yeah, check. I understood your data. I receive that data. Then there’s a block acknowledgment. What a block acknowledgment allows us to do is instead of having to go data, ACK, data, ACK, data, ACK, we can go data, data, data, data, data, data, acknowledgment.
Data, data, data, data, data, data, acknowledgment. And so we can acknowledge multiple data frames at the same time. The risk there is that if one of those data frames gets corrupted, we have to re-transmit the whole thing again. But for something like streaming Netflix or something like that, that can really make things a lot more efficient.
Next is request to send and clear to send. Probably, I can’t get into the specifics of that today. I just wanted you to know that RTS/CTS is another type of control frame that is out there. Maybe we can do something dedicated to packet analysis later on. Not sure. Maybe? But we can talk about that a little bit more in depth.
The third type of frame this is the whole reason why we have Wi-Fi. The whole reason that we have the network at all is for data frames. These are the actual frames that carry the upper layer protocol data. We have data frames. We have quality of service data frames, which have a quality of service aspect attached to them, so we could prioritize them in transmissions.
And then we have null data frames which are things like power, save, poles, and basically stuff that I think they didn’t know what category to put them in, so they just threw them in there in data frames. So these are the three main types of frames in wireless that we need to be aware of that we need to watch out for if we’re doing any kind of troubleshooting or packet analysis or anything like that.
So the next thing I want to talk about is I want to talk about the airtime arbitration process. Basically how devices decide, who is going to talk next. So we’re going to take our knowledge that we just got looking at different frame types to understand how this works.
What I’ve basically got here is I’ve got six stations. Now a station can be either an access point or a client device. It doesn’t matter, but they’re just an access point or a client device of some kind. And then we’ve got time going in this direction.
So pay attention closely here because this is a lot of steps to understand exactly how this works. If we have station 6 here. Let’s say that station 6 has just finished sending some data, and then we have a little break in between when it sends us data and when the next device starts talking, it’s called a SIFS or a short interframe space.
Don’t worry about the name don’t worry about remembering what it’s called or anything like that. Just know it’s a little, it’s just a little break. That’s all it is. So then we have station 5. Station 5 then acknowledges the data. So you see that exchange that we just had there? We had our data. We had our acknowledgment. Cool.
Nice little conversation and their little conversation is now done. Now there’s a little gap in space. It’s 50 microseconds long, so it’s a very short. But it’s a little gap. And when all of these devices that need to transmit on the channel like station 2 and station 4, when they hear that little 50 microsecond gap, they go, ooh, there’s an opportunity to talk.
And so now they know that they might have an opportunity to get access to the medium, and they might be able to talk. So they get excited, and they both basically roll the dice. In their own head, they roll the dice. Station 2 rolls a seven. Station 4 rolls a nine.
So they both roll the dice in their own head and then station 2 says, OK, I’m going to count to 7. And if no one starts talking before me, I’m going to start transmitting. And station 4 also does the same thing. It says, OK, I rolled a nine. I’m going to count to 9.
And if no one starts talking before me, I’m going to talk. So now they both start counting in their own heads in unison. Station 2 goes 1 2 3 4 5 6 7, cool, nothing started talking, so it’s my turn to talk. And then it transmits a request to send to its access point.
Meanwhile station 4 goes 1 2 3 4 5 6 7, oh, because it just heard that request to send from station 2, and so it just went dang it. I just lost access to the medium. I didn’t win the game this time. And so now when that request to send is transmitted and then the access point replies back with a clear descend, that alerts everyone on the channel it says, hey, everyone, set your timers.
Everyone shut up for this amount of time because we’re having a conversation over here. And so everyone, they all synchronize watches, and they start counting. And they all shut up and be quiet, and wait for that conversation to finish. They all start a timer, and they wait for the timer to expire.
Station 2 transmits its data, station 1 replies that acknowledgment, and boom. They’re done, and their timers all expire. And then we have another DIFS, another– what does the D stand for? I can’t remember. It doesn’t matter. But we now have another DIFS which tells everybody distributed interframe space.
I can’t remember, that’s going to bug me. But that’s that signal that tells everyone hey, the medium is available once again. Remember that station 4 got interrupted. It rolled a nine. It counted up to 7 and got interrupted. Well it’s going to hang on to that number. It’s going to go, OK, I only have two left.
And so now it’s only going to count to 2 before it starts talking whereas station 2 is going to roll a 10 this time. And so it’s not going to win contention. It’s not going to win access to the medium this time. And now we’re going to send our data frame that alerts everyone else on the channel.
And then we get our acknowledgment, and then other stuff happens. This is happening if you have your smartphone on, if you have Wi-Fi on your smartphone right now, if it’s in your pocket, this is happening millions of times a second. Your smartphone is going through this process millions of times a second.
Every time your access point beacons says, hey, I’m here. It has to do this. Every time any device transmits, it has to do this. And so after I learned how this worked, it just blew my mind, that Wi-Fi, the 802.11 works at all. It’s amazing that this even works at all, but this is what is happening to every wireless device millions of times a second.
I think it’s absolutely fascinating that this works. Cool. A couple other things. Let’s see. A question from John. The RTS includes the bytes to be transmitted by which other stations and further airtime. Skip, I think you can answer that. Looks like you’re on it there, so I’ll let you handle that.
David has a question though. That’s very pertinent this conversation. He asks, does the Wi-Fi client stop sending probe requests once it connects to an SSID? David, your question could not be timed any better.
– Because now I want to talk about roaming and why roaming is so tricky and what is happening, the basics of what is happening in the roaming process. We don’t have time to get into OKC or NXWEB and R or any stuff like that. That could maybe a future episode, something like that. But let’s talk about the basics of it.
The thing that I want you to keep in mind with roaming is that when a client device has to roam from one access point to the next, the client device is always going to decide when to roam, and the client device is always going to decide who to roam to. My friend Keith Parsons at Wireless LAN Professionals, he kind of pioneered this nice way of explaining.
This is totally his idea. This is his thing that he calls the green diamond. And the green diamond is a decision matrix that every client device has in their head, every chipset has a green diamond. This is a totally made up concept that Keith just made up as a great way to explain this and what is going on here.
So the client device decides when to roam, and the client device decides who to roam to. And the methods that every client device uses is wildly different from each other. For example, my Nokia phone here and my MacBook Pro, they both have completely different green diamonds that they use to decide when to roam and who to roam to.
There’s a lot of different things that it can be based on. I think one of the big ones is, well, first off, signal-to-noise ratio. How much signal do we have above the background noise? If our signal to noise ratio is really good, we’re not going to roam to another access point. Life is good, things are great. We don’t need to roam. If signal-to-noise ratio is really bad, well, then it might be time to start looking for another AP to roam to.
Of course, signal strength falls in there as well. We might use signal strength. Let’s see. Retry rate. Retry rate is a big one. There’s all kinds of different things. We can make a list of about 30 different items that a client device can use to decide when to roam and who to roam to.
Most often, though, I believe it’s going to be signal-to-noise ratio, signal strength, and retry rate. If retry rates are really low, why would we roam to another AP? Things are good. If retry rates are really bad, that’s when your client device is going to start to get grouchy and go, time to find a new access point to roam to.
OK. So let’s take a look at that process very quickly, just a high level overview of what the client association process looks. Like let’s say that we have a MacBook Pro, and it was associated to an AP, but then the person picked it up and carried it over to the conference room. And so our old AP is like way down on signal strength. It’s down to like neg 75 dBm, retry rates are through the roof.
The MacBook Pro starts to go, this is lame. And it’s going to start probing for a new access point to connect to. So in this example, here’s our MacBook running it’s green diamond. And it’s going to send a probe request basically anyone on the channel with the same SSID. It’s going to say, hey, any access point with aperture networks, are you there?
And then any access point with aperture networks as their SSID is going to reply with a probe response. You remember that was one of our frame types. And so now, the green diamond is going to run, and it’s going to decide which access point to associate to. And so then it’ll choose an access point. It says, I choose you.
And it’s going to associate– it’s going to send an authentication request to this access point. We’re going to get an authentication response back. Then we send an association request, hey, can I connect to you? Yes, you meet all the criteria, you support the right data rates, you do all the right things. Yes, you can connect to me, and then we conduct our four-way handshake to set up encryption for WPA and things like that.
Then we still have to do a bunch of stuff. We might have to– DHCP might have to happen again. We might have to get a new address if we’re on a different VLAN or something like. That could be bad. But there’s a lot of different steps that have to happen for a client device to roam from one access point to the next.
Here’s the thing that I want you to keep in mind. The thing that I want you to know is that every client makes a decision. It’s not up to the access point. It’s up to the client, and every client has a different green diamond. Every client device makes these decisions differently. You take an iPhone 11 Pro, you take this super cheap Nokia 6.1, they will have different roaming characteristics.
You take a MacBook Pro, and you take an iPad, they’re going to have different roaming characteristics because the iPad is designed to be a mobile device. The MacBook Pro is designed to be a portable device. They’re going to have different roaming characteristics. So Roku. Rokus are not going to roam aggressively because they don’t move. They get plugged into a specific TV and they stay there.
John asks, does it have to hop channels to do the probe? Yeah, absolutely. It’s kind of fun if you look on a spectrum analyzer. It will always probe at 1 megabit per second, the lowest, oldest data rate down in 2.4 gigahertz. If you look at the whole spectrum on a spectrum analyzer, you’ll see a curve 802.11b shape, go [MIMICKING COMPUTER BLOOPS], crest all the channels as a device probes for a network.
So it’ll ask, is anyone there? And wait a second, move on to the next one, ask, is anyone there? And wait a second, and move on and go through all the channels. It’s really fun to watch. Every now and then if you have a good spectrum analyzer, you can catch a device probing for networks. And you can also run a packet analyzer and look and see if anything shows up there as well. So really, really fun stuff. OK. Let’s see.
– The last thing that I want to talk about before we get into a couple of live demos things, I want to just give you a very, very quick look at the different security methods that are out there for wireless. Again, this is super duper overview stuff here. We’re not going to get in depth, here this could be an entire session of its own. And maybe we can do that at some point, if you’re interested in doing that.
There are several different ways that we can secure a wireless network. Remember that with ethernet, ethernet is a bounded medium. It is inherently secure and that it has physical security that you’ve got to basically get into the cable to see the traffic going by. Wi-Fi though, this is completely different. This phone broadcasts in all directions evenly so that it doesn’t matter how I hold it in relation to my access point downstairs.
So that means that if there’s somebody sitting out on the street in front of my house, they can put a wireless adapter in monitor mode, and they can capture all the packets that are flying back and forth between my phone and my access point and my MacBook and my access point. They can easily do that. It’s very easy to eavesdrop. And so because of that unbounded nature of wireless, we have to apply security to it.
So there’s several ways we can do that. Well, first off is open. That is no encryption, no security at all. You go to the airport, that’s what you’re going to use is an open wireless network. There is no security. Luckily, you have security at different layers when you do web browsing, things like that. You have security there, so it’s not that big of a deal.
The next one is called WEP or Wired Equivalent Privacy. This was the original way of securing wireless networks, and it is hilariously broken today. Wired Equivalent Privacy is super broken. I think that just naming it Wired Equivalent was their first mistake. That just guaranteed that it was going to get broken at some point.
You have to use a hexadecimal key, so it has to be all in a hex, which is really annoying to remember. I remember a few years ago, I visited my grandma, and she had WEP on her network. And it was easier for me to crack the network. It was faster for me to crack the network than it was for her to go find the sticky note that had the passphrase written down on it.
So next is WPA. This was introduced as a stopgap. When WEP was completely broken, WPA or Wi-Fi Protected Access was the stopgap until WPA2 new hardware could be rolled out. So you could basically implement this on WEP hardware with a firmware or software upgrade. But because it’s a stopgap, the IEEE limits you to 54 megabits per second. They say, hey, 54 megabits per second, no faster.
That’s as fast as you can go because ultimately, they want you to transmit– or transmit– or transmit– they want you to transition over to WPA2. WPA2 uses a brand new cipher suite that requires new hardware, but it’s nice and secure. Compared to WPA, it’s nice and secure. There are two variants of both WPA and WPA2. There’s WPA2 Personal, which is just a pre-shared passkey. This is probably what you’re using at your house.
And then there’s WPA2 Enterprise, where we can actually at the access point we can receive credentials from the user that’s trying to get on the network, and we can check in with our RADIUS server. And the RADIUS server gives a thumbs up or thumbs down. If it gives a thumbs up, then we unblock that port and we let the device start passing traffic and set up encryption and all that stuff.
So the new one that’s on the horizon, and this is what I’m not super knowledgeable on yet. I need to spend some time studying this we probably all should because this is coming up quickly. And we’re actually introducing this at MIST I believe next month. Don’t quote me on that, but very, very soon. WPA3 is Wi-Fi Protected Access 3.
One of the things I’m most excited about it is it has what’s called enhanced open. If there is an open wireless network that does not require a passphrase to join, we can still get encryption on it with Opportunistic Wireless encryption. Very, very cool. I’m really excited about that. And also beefs things up with WPA3 Personal and WPA3 Enterprise as well. We might be able to get into that more in a later session but not today. Cool.
– So what I want to do now is I want to spend just a few minutes talking through what it looks like to design a wireless network. I just want to show you the design process from just the overall design process from the beginning, and so you can kind of see what that looks like. It’s not going to be an in-depth thing.
I’m not going to try to help– I’m not going to try to get to you to a point where you can do it. I just want you to be familiar with the concepts, so that you understand the different steps involved. The very first thing that I like to do when I’m designing a wireless network is I want to determine the network requirements.
There was a guy named Zig Ziglar, and he’s one thing that he’s famous for saying is if you aim at nothing, you’ll hit it every time. And what that means is that you need to plan. You’ve also probably heard the failure to plan and planning to fail, stuff like that, right?
What I always like to do is I like to meet with the customer to determine network requirements. I want to find out what do you need the network to do, and then I want to design the network to meet those requirements. For example, I’m going to ask where do you need Wi-Fi? Where do you not need Wi-Fi?
I mean that’s a real serious question. You could save a customer thousands of dollars by eliminating a few APs in places they just straight up don’t need wireless, and so that’s a good question to ask. I also like to ask what devices you’ll be using? Are they going to be using laptops, tablets, smart phones, barcode scanners?
What are they going to be using, so I can design the network to meet those specific requirements. I’ll also ask how many devices or users will there be. Are we designing a network for 100 devices or 500 devices or 20,000 devices? We need to know, we need to iron this stuff out ahead of time and design a network that meets those requirements.
We might ask what applications will they use. Are they going to be using web and email? Are they going to be doing voice over IP? That’s going to change things like what kind of signal strength we want to offer on the wireless network. I’m also going to ask them what clients they have, so I can make sure that I use channels that the client supports.
Back behind me on the shelf, I don’t think you can see it not microphones in the way. But right back behind me on the shelf, there’s my Nintendo Switch. The Nintendo Switch does not support channels 130, 132, 136, 140, or 144. No support for those channels.
So if I was designing a wireless network for a college dormitory, I would not want to use those channels in my network plan, because then my client that the Nintendo Switch would have dead spots. There would be spots in the building where it could not find an access point to connect to because it does not support those channels.
If you want to look and see what channels your client device supports, then you can check out this website, clients.mikealbano.com. Mike is a wireless network engineer at Google, and he put together this fantastic resource that you can check out to see what channels your clients do and do not support. Really, really cool tool.
I really like that tool a lot. So once I’ve gathered the network requirements, then I’m actually going to turn these into design requirements. I’m going to turn these into some hard numbers that I’m going to use to design the network. For example, I’m going to decide what’s my coverage going to look like. What is my primary coverage going to look like?
From the primary access point, what should our minimum signal strength be? And I’m just going to throw something at the dart board here and say let’s say -67 dBm or better. For secondary coverage, I’m going to say, OK, we have our loudest access point, but what about our second loudest access point.
Should we have another access points providing some overlap? Secondary coverage is a great way to determine that. From the second loudest access point, what should our minimum signal strength be? Maybe we’ll say -75 dBm or better. That way, that gives us a little bit of overlap between coverage cells.
It gives devices a little bit of time to complete a roaming event, and it also gives us redundancy if an AP goes down. Next 75 is a good number, I would say -67 though is an even better number. That’s typically what I do as I do secondary who is also -67.
Next is our minimum data rate. How low do we want to allow our data rates to go? I usually say no less than 24 megabits per second. But remember, this is the signaling rate not throughput. That’s the signaling rate.
Next is a really important one, signal to noise ratio. How much signal do we want above the background noise? Maybe it’s 25 dB. And finally, this one’s really important is how much channel overlap will we allow. Basically, how many access points will be allowed to share a channel above a certain signal strength.
And I typically go for no more than one access point on a channel above -85 dBm. If we have two access points sharing the same channel and they’re both above -85 dBm, then we would fail that requirement in that area. So what I’m going to do now is I’m going to hop out of– I’m going to hop out of the software for just a second here. Let’s hop out of the PowerPoint deck for a second here, and show you what this looks like in software.
But one last thing before we do that, I want to tell you a little story. You probably saw in that slide that I said secondary coverage. Basically, secondary coverage is we want to make sure that we have at least two access points covering the same area. Why do we want to do that?
– So what I’m going to do now is I’m going to hop out of– I’m going to hop out of the software for just a second here– or I’m going hop out of the PowerPoint deck for a second here and show you what this looks like in software. But one last thing before we do that, I want to tell you a little story.
You probably saw in that slide that I said secondary coverage. Basically, secondary coverage is we want to make sure that we have at least two access points covering the same area. Why do we want to do that? Well, one reason why we might want to do that is for roaming performance.
So I’m going to show you a picture here. This is my dad’s airplane. Don’t worry, he wasn’t flying it when that happened. He actually had sold it to somebody else when this had happened. But this was a nice little Mooney M20c. It’s pretty notorious for being a difficult plane to fly.
It’s very fast, but you can get in trouble with it very, very quickly. He sold it to another guy who was an experienced pilot. And he was up in Virginia one day, and he was taking off from a small airport. And he was taxiing towards the runway. And he started taxi down towards the very end of the runway.
It’s going to clear it down to the end of the runway. And somebody got on UNICOM, just a local radio channel, and said, Mooney, you don’t need to go down to the other runway. Just start off in the middle. And he went, oh, OK. And he just randomly took this random guy’s advice.
So he started take off down the runway. And by the time he got to the end of the runway, he knew he didn’t have enough air speed. But he got off the ground, barely tried to clear the tree line at the end of the runway, and put the plane in a stall over a lake. Very, very bad situation.
By the way, he’s got his wife and his newborn daughter in the plane. Really, really bad scenario. And he knew this plane’s going down, I just have to decide where. And he identified on the other side of the lake, he identified this house. And he said, that’s where the plane is going down.
So he went between a couple of trees. It sheared off the wing on the far side, notice, there’s no wing over there, and put the plane into the house. And they all walked away with just scratches and bruises, including a newborn daughter, and nobody was home. It’s a miracle, right? Ultimately, he didn’t have enough runway.
He didn’t have enough runway to do what he needed to do. He didn’t have enough runway to get enough speed to get off the ground. So when we are designing a wireless network, we need to design for secondary coverage. We need to give overlap between coverage cells to give our client devices time to complete a roaming event from one access point to the next.
That is why we– that is why we need secondary coverage. So now that story time is over, I’m going to jump into a site survey and design tool. And we’re just going to do a really quick network design. I just want you to see– I just want you to see what the process looks like. We’re not going to get in really in-depth here.
And by the way, it was not air traffic control. It was somebody local on UNICOM, I believe someone at the FBO to [INAUDIBLE] there. So I’m going to open up Ekahau Pro here. And I’m just going to bring in a floor plan. We’re going to do a quick design on this floor plan. I just want you to see what the process looks like really quick.
So the first thing I’m going to do is I’m going to add a map. So we’ll go to please add map here. And I’m going to grab this CAD file. And then what I’m going to do is I’m going to basically go through and I’m going to tell the software what the building is made of. We’re going to tell it, hey, these are brick walls. Hey, this is a drywall.
We’re just going to tell the software what this building is made of. Now, I’m importing a CAD file, which means it draws it in for me automatically. They usually doesn’t work. This is a nice CAD file that it happens to work on. Usually, you draw these in manually, but it’s not a big deal. It doesn’t bother me at all.
So I’m going to say at the doors, I’m going to say that those are a six dB door. That means they knock down wireless by six dB as the wireless goes through it. There’s an elevator shaft here in the center. See that blue there? Let’s say that’s an elevator shaft. That’s 30 dBs of attenuation.
So as the RF goes through it, it gets knocked down by 30 dB. Pretty big deal. Then we have the exterior walls. I’m going to say that those are brick. And then we have stairwells here. I’m going to say that those are a– I’m going to say those are a concrete wall.
And then we have our interior walls. I’m going to say those are a three dB drywall. So an RF goes through it, knocks it down by three dB. So then we’ll hit the Import button, and Ekahau’s going to go through and draw in all of our numbers for us here and give us some very important messages about what’s it doing.
Here we go. So now, we’ve got all of our walls drawn in on our floor. The next thing I’m going to do is I’m going to define where we want Wi-Fi and where we don’t. So I’m going to just use the area tool here to say, hey, this is where we want Wi-Fi. And I’m going to go through and just do a very quick shape here because this is designed just to be quick. This isn’t going to be a real design.
If this is a real design, I’d be following all the contours of the building. We would make it look really nice. That’s what we’re trying to do today here. OK. So now, we’ve defined where it’s going to be. So I’m going to hide– I’m going to hide that area just to get it out of the way.
And then let’s go ahead and place some access points. So I’m going to go and grab a missed AP here. Let’s do a– let’s do a 43. So we’ll grab our missed AP 43. And we’ll zoom in. We’ll zoom in a little bit here. And then I’m going to go ahead and start placing access points.
Now, notice that right now, I’m looking at signal strength in both the 2.4 and five year gigahertz band. Wrong. We don’t want to do that. We just want to look at the five gigahertz band. So I’m going to select that, and we’re going to place an access point. And look at that.
Now, we have some signal strength. So I’m just going to go through, and we’re going to place some access points here. Notice that anywhere where there’s color, that is where we are achieving our signal strength requirement of -67 dBm. Anywhere that’s gray is where we are not achieving our signal strength requirement.
So let’s go through and place some APs here. One thing you will notice is I’m not putting them in the hallways because where are the clients? This is a college dormitory. They’re not going to be in the hallway. The clients are all going to be in the rooms. We want high signal to noise ratio for the clients, so we’re not going to put the APs in the hallways. Never, ever, ever want to do that.
OK, cool. So there we go. We have our primary signal strength everywhere. Let’s check on our secondary signal strength. This is going to show us anywhere that has two access points covering it above -75 dBm. It looks pretty good except there’s a little bit of a gap right here. So I’m just going to add another access point right there to fill in that little gap.
And then I think over here, you know, I’m not going to add one more right there. It’s just one tiny room, where we’re going to fail that requirement. I’m OK– I’m OK with that. OK. So then the next thing I’m going to do is I’m going to switch to network issues.
So what I like about this view is it just shows us whether we are passing or failing our requirements. If we’re failing requirements, there’s a color assigned here. If it’s white, that means we’re passing our requirements. So I have great news. The great news is that we’re passing our requirements right here. And we’re also passing our requirements right there.
That’s the only two places where we’re meeting our requirements. Something is wrong. If you look, we’re actually failing our channel interference requirement. That means that all that– we’ve got too many APs all sharing the same channel here. In fact, let’s go modify our requirement there really quick.
So if we go to Coverage requirements. And I’m going to switch that down to just one access point above -85 dBm. Now, we’re failing it everywhere. So how do we fix that? Well, one thing we can do is we can actually run a channel plan on this. So what I’m going to do is I’m going to say, let’s run the channel planner.
Let’s get rid of channels 132, 136, 140, and 144, and hit Create plan. And there we go. Now, we have run a channel plan on our network, and we are now passing requirements on 94% of our floor plan. Now, it used to be that before missed, it used to be that you would have to go into a controller, and you would have to program in all of these channels and say, OK, well, this one’s going to be on channel 108. This is going to be on channel 124. This is going to be– no, you don’t have to do that with Mist.
Mist has RRM, Radio Resource Management. And it will automatically perform the channel plan for you. But the reason why I like to run the channel planner in a tool like this is I want to make sure that this channel plan is viable. So let’s bump it up to 40 megahertz channels. So we’ve got 20 megahertz channels right now.
If we bump that up to 40 megahertz channels, notice that our available channels gets cut in half. So let’s go ahead and create a plan with that and see if that works. OK. With 40 megahertz wide channels, now, we are running into some areas with channel interference.
And so we probably can’t quite make 40 megahertz wide channels work in this environment. Just for fun, let’s just go up to 80s. It’s not going to work but just for fun. And fail all over the place. It’s actually better than I thought it would be. And just for fun, just to push the limits, and just see how ridiculous we can do things.
160 megahertz wide channels. They just– yeah, this is going to fail. Yep, no surprise there. It completely fails. So that is what I like to do. I like to use a site survey and design tool. It doesn’t have to be Ekahau. There’s lots of different options out there, like IB Wave, or TamoGraph Site Survey, or VisiWave as well.
But I like to use these to make sure, do I have enough frequency space to do what I need to do? And then I can kick it over to Mist and let RRM take care of it from there. And we have validated that, yes, there should be enough frequency space, and Mits RRM will do our channel plan, and it will set our transmit power for us and get all that stuff ironed out.