Hello. In this video, we’re going to be looking at gathering and defining technical requirements for a wireless LAN project. Now, technical requirements are born out of the business requirements discussed in our last video. And sometimes these requirements will be defined by the wireless LAN design engineer to meet the gathered business requirements. But sometimes some of these technical requirements will be defined by the business themselves.
Now, it is important we don’t just see technical requirements as a big checklist, like Stanley here. So it’s not just, “Do we need voice? Yeah, check.” No, we need a much more collaborative process. So yes, you need voice, but which phones will you be using? Which VoIP applications? And what are the application and codec requirements? What are the jitter packet loss delay requirements? What are the roaming expectations for a voice call? Do we need to roam between floors, or between buildings, et cetera?
So this process may involve engaging with end users, just as Misty is doing here. We want to know, “How does the user use the Wi-Fi currently? What applications that they using? And what would they like to do in the future, which they currently can do?” Now, from these interactions, a set of technical requirements can be created, which will provide users with their desired level of experience.
So let’s look at some of the things we need to consider when gathering and defining technical requirements. First, we need to look at defining a set of requirement areas. Now, this is done by taking the areas which the business requires wireless LAN services to be available in, we might call them service areas, and splitting them up into a set of areas with common requirements.
For example, let’s consider a university building and a lecture theatre, which might be considered a high density area with seating for 800 students in a relatively small area. However, this area might also be considered a medium to low throughput area, as students would be mainly listening to the lecturer and not using devices. Contrast that with a student study room, which may have space for 40 students, but each student may be streaming a different lecture, or accessing different online resources, checking their emails, while also connected to some music streaming service on their phone. You see, here, we have a lower user density area, but higher throughput requirements.
You see, we have to define the requirements area by area. And then for each of these areas, we need to know, “What is the maximum number of users in each area?” Now, notice I said users, not devices. And a user may have multiple devices, a phone, a laptop, a smartwatch, for example. So do we not need to take account of the number of devices in our capacity design? Well, for very high density designs, such as stadiums, the number of devices is important. We need to make sure that we have enough association capacity to allow all the devices to connect. Now, a Mist AP can support a maximum of 256 associations per radio. So in a high density environment, yes, we have to take into account a maximum number of devices into our capacity calculations. And I will talk more about it in our high density design video in our design section of this site.
However, when looking at the throughput requirements, I’m more concerned with the amount of throughput each user requires, and not how many devices they will be split in that across. For example, let’s go back to our student’s study room. Now, we may have one student using headphones listening to a streamed music provider via their mobile phone, and also accessing study material on their laptop. But another student might be doing the same two activities, but both via their laptop. Now, both students are consuming the same bandwidth, just on a different number of devices. So from a capacity point of view, I’m more interested in the amount of bandwidth each user requires.
Now, I do also need to consider what device types are going to be used. You see, different devices have different capabilities. For example, which 802.11 physical layers will support the number of supported spatial streams? These credentials will dictate the maximum data rate a client is able to support. Now, given the wide-ranging mix of devices which connect to modern day networks, calculating the available capacity of an access point can be an impossible task with just too many variables to consider. Not only the capabilities of each client needs to be considered, but how much data will there be transmitted? How often? What are the chances clients will be transmitted at the same time?
Now, one of the aims of a wireless LAN design is to calculate the maximum number of clients which can be associated to an access point radio while maintaining the client’s requirements. Now, calculations here should allow room for changes in the environment as well.
Now, luckily for Juniper Mist users, Mist AI has an AI driven service level expectation, what we call an S elite, for capacity, and it keeps track of all the device capabilities and usage, and will inform a network administrator if there’s not enough capacity. So as capacity requirements change and increase on our networks, network administrators can proactively deploy additional APs.
Now, from a design perspective, we still have to design a network to meet a given set of requirements. And because different devices have different area of circuitry, the number of access points needed in their location may be different to meet the requirements for, say, an iPhone compared to a laptop. So when we design, we need to pick a device, or maybe a group of devices, which we are going to create a design for. We call this device the LCMI, or Least Capable, Most Important. Now, for more information on the importance of picking the correct LCMI and how we select it, click on the LCMI, Least Capable, Most Important link below.
Now, having defined our areas and the devices which will be operating in those areas, we next need to think about the requirements for each of the applications the devices will be using. For each application, what are the bandwidth requirements? For some applications, this might be publicly available. For example, Zoom bandwidth requirements are published on their webpage. For other applications, perhaps a custom warehousing application, for example, we may need to contact the application developer, or even perform some protocol captures to get this level of information.
Another technical requirement we need to consider is, “Does the network need to support VoIP or some other latency sensitive application?” Mobile voice applications require a wireless LAN, which supports seamless roaming. And therefore, secondary coverage is required. Now, secondary coverage is the minimum signal strength we require from the second-strongest access pond in each location. And it is used to ensure a wireless client can always see another access point to which it can roam.
Now, we will be discussing more about secondary coverage in another video in the design section of this website. But it should be said that it’s very unusual today not to design for secondary coverage. Voice and video have become expected applications on nearly all modern day Wi-Fi networks.
Another consideration is, “Does the system need to support realtime location services? Will this be Wi-Fi location, or BLE location?” This requirement will not only affect the choice of APs, but also the location of the access points.
For example, let’s consider this floor plan. So when we consider a room, such as this, if we were going to design a wireless network for this room, we might consider placing three access points along the center of the room. With integrated omni direction antennas, these access points being located in the center provide the best coverage for a room like this.
But let’s consider we also wanted to do location-based services. And let’s pick a location, as we can see marked here by a red X. If we wanted to locate a device in this area, and then we want to consider, “Well, how do all the access points see the signals from a device located with that red X?” Well, obviously, the APs that are closer to the device are going to see it with a stronger signal strength than the one which is further away. Let’s consider a second location by the red X down here. Now, if we consider the same thing, how do our access points see this device? We’ll see that it’s exactly the same. Both of these locations are indistinguishable from a location point of view when we have access points all in a line.
So you see, when we do a location-based design, what we want to do is we want to try and place our access points around the perimeter of the area we want location to work in. So let’s consider our location design. Here, I’ve done a design based on location, and you can see that the access points have now been placed around the perimeter of this room. Let’s consider the same two locations again. So location one, we can see the sort of distance this location is from each AP. And now, when we considered the second location, we can see that it’s at very different distances at the second location. The second location is much closer to our AP down here at the bottom than our first location. And this allows us to be able to do more accurate location-based services.
But notice something of this design. Not only has the position of my access points changed, but the number of access points required. I now have six access points instead of three. So when designing for location, we need to realize that not only will the number of APs required increase, but also the location where they go will also be different.
Other technical requirements we need to consider are, “How many SSIDs are required to meet a gathered business requirements? And what security will be applied to each one?” We also need to consider, “Is AP redundancy required? And if so, in what areas?” And finally, one of the important technical requirements will be the choice of access point.
Now, it’s essential to choose the access point model before the design phase. I have had customers say to me, “Can you do the design, and we’ll choose the AP later?” Well, my answer is always no. Because different access points have different antennas, and different antennas have different coverage patterns. Let me show you this using the Ekahau design software. In this example, we’re going to look at the coverage of a Mist AP if it was placed in the center of an American football field.
So at the moment, I’ve placed a Mist access point in the center of this football field, and we’re just going to visualize the signal strength, the coverage we get from that AP. Now, at the moment, I’m looking at a Mist AP43 placed right on the center point. If I was to compare that to what a Mist AP32 would look like, we can see how the coverage changes, how the coverage pattern looks different.
If I compare that to a Mist AP12, the wall plate AP, then we get a very different coverage pattern, because we get… It’s obviously meant to be wall mounted, and the signals pointing in one particular direction. We could, obviously, change the direction that that AP was pointing in, and we would see, yet again, a different coverage pattern. Now, you might say, “Well, comparing a ceiling mount AP with a wall mount AP isn’t fair.” Well, I could take the Mist AP12 and say, “Well, what if we placed it on the ceiling?” Well, again, we get a very different coverage pattern. Not an ideal coverage pattern, because not the way the AP is designed to be mounted. But when we compare these three APs all mounted on the signal, on the ceiling, sorry, we’ll see very different coverage patterns.
So it is this very reason why you can just do a rip and replace. If you change your AP model, the number of APs required and their positions of those access points will also change.
Now, to learn how to choose the right Mist AI access point for your design, click on the link below to look at the video Selecting a Mist AI Access Point. So thank you for watching this video on defining and gathering technical requirements. And goodbye for now.