you are designing a wireless network for a client This is a topic that many people are looking for. star-trek-voyager.net is a channel providing useful information about learning, life, digital marketing and online courses …. it will help you have an overview and solid multi-faceted knowledge . Today, star-trek-voyager.net would like to introduce to you High-Density Wi-Fi Design Part 1 – Forecasting AP Capacity. Following along are instructions in the video below:
“I m andrew donaghy with arab networks and today i d like to talk to to you about high density wi fi design and one of the first steps in a high density wi fi network is requirements gathering high density wi fi networks are becoming much more prevalent. Today than they were in the past most people think of high density wi fi networks as only being in stadiums. Large lecture halls and universities and things of that nature but actually high density wi fi networks are becoming much more prevalent in common organizations for conference rooms meeting rooms cubicle space and even common areas and in education within classrooms. As well the reason for this is because white networks in the past were designed around coverage as the major design variable.
Did clients have an adequate signal to noise ratio and rssi at a given distance from the ap and to make sure we had coverage throughout the entire facility as we have more clients coming into common wi fi networks we find ourselves being limited by capacity and so capacity is becoming just as equally important in designing a wi fi network as coverage has been in the past so anywhere where a single access point can provide coverage but not enough capacity we need to start thinking about designing wi fi networks differently we need to start taking into account client capabilities and application throughput capabilities in order to make sure we have enough capacity for the wi fi network. And this is because wi fi uses or the air transmissions. Or rf. And in this.
Case. Airtime. Is the limited resource that multiple clients are contending for with limited airtime each client. Only gets a piece of the pie or a piece of that rf spectrum to use at a given time therefore we need to design with all the client capabilities in mind in the mix of those clients and environment to understand how much airtime each client is going to be using in order to properly design our network.
The adequate amount of capacity so we can highlight this with a simple drawing of an access point if we have one client. Connect let s take a simple example of a laptop. Now if we take this laptops capabilities into account. Let s say.
It s a three by three. My mo and it s capable of four hundred and fifty megabits per second it s actual tcp ip throughput will be anywhere between 40 to 60 percent of the raw data rate and this is because of overhead on the wi fi network for this example we ll make it very simple let s use 50 of the actual raw data rate is actual tcp ip throughput. So in this case. We ll say 225 make it it s per second is the actual tcp ip throughput.
Whereas this one to the wrong. If he s the only device on the access point this means that he has no intention for the uptime with any other clients and he can probably get all 225 million bits per second through and this usually results in about 80 to 90 of channel utilization or the error on that frequency or on that channel. But to highlight how different device is interacting on the same frequency than same channel impact the capacity of an access point and of the network. Let s add in a second device to this example.
So say we have a second device. And it s a tablet device this tablet is a one by one my mo capable of 65 megabits per second draw. And that translates into let s say. 30 megabits per second of actual tcp ip throughput.
Now in a equal scenario. Where both devices have equal airtime. This will result in both getting..
Approximately. Let s say 50 just to make it simple. Here time on that ap radio. This means that if the laptop previously got.
225 megabits per second. When it was the only device. Now. It s only getting maybe.
110. Megabits per second under this new scenario sharing their time. With this tableau this tablet. Because it s only capable of 30 megabits of well tcp ip throughput.
If it s usually half of your time. Let s say it gets maybe. 15. Megabits per second.
We add those two together and we ve got 125 megabits per second of actual throughput. Now when the airtime is shared between those two devices. And it s a split between the two devices reflecting their individual capabilities. So we can see right away that the capacity of this single channel has gone down from 225 megabits to 125 megabits just by introducing the interactions between multiple stations on the same channel.
That s very easy way to see on a very simple scenario or there very simple example how the interactions between devices affect one another on the wireless network. Because that s a shared airtime and shared rf medium so it s important when we re designing high density wi fi networks that the capacity that we take into account is not just based on what access point would buy it could be a low end ap. It could be a high end ap. It s going to be the interaction between the ap and its capabilities.
Along with all of the client devices on the channel and on the same frequency and all of their unique individual capabilities and the interactions between those that govern the actual capacity of the network. So when we re thinking about doing requirements planning for a high benefit wi fi network. We need to go through a process to plan out exactly and map out exactly what those client and ap capabilities are we start by taking a client inventory. So we wanted to understand what client devices.
And what types of devices are coming into the network as well as plan for multiple devices per user. Today users are carrying anywhere between two and three devices. Usually a some form of a laptop tablet the smartphone..
Two of those three may be using multiple even at once we need to understand what s being brought into the environment and having a tool in your infrastructure. That can report on client device type and operating system is very useful second we need to understand the client and the access point to radio capabilities. Because the different mixture of these unique variables is what determines their inter then they ultimate capacity on the wireless network. That you can plan for so we need to understand what the radio capabilities are such as multiple input multiple output and spatial division multiplexing how many spatial streams can client support what channel width can each of those clients support and what are we designing our network for in terms of channel width on the access points.
We re doing 20 megahertz or 40 megahertz channels. And that will impact how much airtime is required for each client to get through the amount of data for the application that the user is using we also need to understand channel support in order to understand and to plan our network during a site surveying process. What channels can and can t be used so. We don t end up with any coverage holes for for our certain clients that maybe don t support.
Let s say. The dynamic frequency selection bands and finally birds third. We need to understand application throughput. So once we have let s say the raw specifications for client device.
Like three by three mile and a raw data rate of 450 megabits per second with an actual data rate or actual throughput of 225 megabits per second. We need to understand what application is these are going to be using in this environment on that device. Because if a user is using a high bandwidth application. They re going to be using more airtime than if they re using a low bandwidth application and the same application being used on different device types can take different amounts of airtime or different capacity on the network in order to get that data through so for instance both of these devices a laptop and tablet.
If they are both doing the same application the tablet would take significantly longer or more airtime in order to get that same data through as the laptop would because it s a slower speed client once we understand the application throughput for every user or every device type then we can actually start forecasting ap capacity as an initial starting point into our rf design process to forecast the initial design or to forecast the initial access point capacity. We need to take the full interaction of all the devices that are anticipated on the network. So to take an example here. We re gonna do two examples.
The first one we ll take an example of 30 tablet device is in a classroom environment. And and the class is the lecture. Is based on let s say a 2 megabits per second streaming video. So here.
We have an access point a classroom environment or perhaps. Multiple access points. That s what we re determining we ve got 30 tablet devices. So 30 of these guys they re all going to be doing.
2 megabits per second streaming video that s our application throughput that each requires so these tablets. We re gonna sell one by one my mom capable of 65 megabits per second and an actual tcp ip throughput. Then of 30 megabits per second..
So to understand how many access points radios and channels. We need to support all three of these tablets in a classroom. We need to understand how much airtime each one is going to use in order to get through that 2 megabits per second video. So here we take 2 megabits per second divide.
It by the actual capability of the client device. Which is 30 megabits per second of actual tcp ip throughput and it gives us approximately 7. Airtime that each client device is required in order to get the desired application through player. If we have 30 devices times 7.
Percent. We have two hundred and ten percent required air time in order to serve all 30 tablet devices. Concurrently and what this means is if each ap radio or each channel is 100 percent. We would need at least three ap radios in order to serve all three all 30 tablet devices at once that s gonna translate typically into two access.
Points that have dual radios one on 24. And one on 5 and we ll have some headroom for growth or perhaps to support a higher band a little bit higher bandwidth application for every tablet device. So that s a very simple example using a homogeneous environment with all tablets for a second example. Let s take one that s a little bit more complex.
So for a second example. Let s take more of a heterogeneous environment. Where we have a mixture applying device types so let s say we re planning a lecture hall a large university. We re going to plan for 200 users.
These 200 users are going to bring in two devices each maybe a laptop and a tablet. We re gonna have 400 total devices this accounts for 200 laptops and 200 tablets. Now. The laptops are let s say.
Let s plan for our two spatial stream. Because maybe we have some lower end laptops. We don t know exactly what these students are gonna bring in we re going to plan a little bit worst case scenario rather than assuming they have the latest and greatest maybe they re on a low budget or something so they have two spatial stream capable of 300 megabits per second or off or 150 megabits per second of actual tcp ip per client and the tablets are one spatial stream devices. Therefore they re capable of 65 megabits per second or 30.
Megabits per second of actual tcp ip throughput. Now for an application or in assume that each client is doing some casual web browsing email and taking notes during the lecture let s give them just and assume the value of one megabits per second that is the minimum threshold that we want to be able to provide to every client. So our application throughput in this case is one thing this per second whether they re on a laptop or a tablet..
Now for the laptop device. We need to figure out how about how much their time each device will take in order to get through that one megabits per second of data. So for the laptop. We take one megahertz per second divided by what each one is capable of which is 150 megabits per.
Second and that gives us an airtime utilization for each device of 066. We have 200 of these devices in the environment. So multiplied by 200 to account for all. The devices.
In the lecture hall and that means we need 133 percent air time total in order to serve all 200 laptops. So obviously one access point radio on one channel is not enough will need at least two radios or dual radio access point in order to serve all the laptops for the tablets. They re needing the 1 megabits per second. But they re only capable of 30 megabits per second of actual throughput.
So the airtime utilized by every tablet device is going to be higher. It s going to be three point three three percent. So already you can see the disparity for the same application throughput that the tablet requires much more airtime than the laptop. We have 200 tablet devices.
So i ll go ahead and multiply that out and we ll end up with a total airtime utilization for all 200 tablets of 666 percent. We add the two together because we need to support all of these in the same environment and we end up with 800 percent air time utilization now if every access point radio on a single channel. Can serve 100 air time that would equal 8 access point radios. But in the real world not every access point radio on a single channel can max out the airtime utilization to 100.
We typically see airtime saturate around 80. So if we divide 800 percent by 80 percent. Which is a realistic maximum airtime utilization on a channel. We end up with 10.
So we need 10 access point radios in this lecture hall to simultaneously serve 200 laptops and 200 tablets. This can be 5 dual radio access points or any combination in between so if you d like more information on designing high density wi fi networks you can check the arrowhead website and on this blog we ll link to the arrowhigh of high density wi fi design principles white paper as well. As the era of high density life i design and configuration guide. Thank you ” .
Thank you for watching all the articles on the topic High-Density Wi-Fi Design Part 1 – Forecasting AP Capacity. All shares of star-trek-voyager.net are very good. We hope you are satisfied with the article. For any questions, please leave a comment below. Hopefully you guys support our website even more.