WiFi is a Utility, and Needs Capacity Planning
When is the last time you said: “Wow, this WiFi is great!”? You don’t really notice it when it works. You are more likely to say: “This WiFi is crap” when it doesn’t meet your expectations. WiFi is no longer a convenience, it’s an essential utility like electricity. You would like it work every time and without hesitation, like turning on a light.
Like the power grid, one of the biggest challenges in designing a wireless network is capacity planning. The goal of capacity planning is to determine how many access points are needed to provide a good user experience. Deploying too many APs is a waste of money and can make performance worse, but deploying too few will cause user experience problems (the equivalent of brownouts) when an AP becomes oversubscribed.
There are a number of WiFi capacity planning tools that list the variables and help do the math, based on your assumptions. The problem with most of these tools is that they assume similar APs have similar abilities. As we have shown in our competitive benchmark tests, not all APs are the same. They have different capabilities and limitations, which must be taken into consideration when designing a wireless network for the best user experience.
To illustrate, we recently compared our C-120 to Aerohive’s flagship access point, the AP250. It may not seem fair to compare a 4x4:4 (C-120) unit to one that’s only 3:3:3, but the AP250 is the best AP Aerohive offers.
One metric for an AP’s capability is the number of clients it can support performing a given function. In this test, 1080p video was streamed to clients. The AP’s capacity was determined by adding clients, one by one, until the user experience was impacted. There was no other load on the AP.
What this test shows is that as few as 17 students streaming video can fully occupy the Aerohive AP. Any additional load on the AP250 caused pixelation and video screen freezes: a poor user experience. Mojo was able to support three more high quality video streams. One could argue that this could be the difference between in a 4x4:4 and a 3x3:3 AP. Fair enough. But let’s look at another test.
Mixed Traffic Test
Network traffic, especially WLAN traffic, is unpredictable and so is extremely difficult to model. The applications and their requirements change from year to year, and are hard to plan for. Currently, one computer per student has become the norm, BYOD is expected, and 1080p is good-enough quality. (4K content is available and 5K is on the way.)
Another important measurement of how well an access point performs is how well it handles a mixture of applications under heavy load. This would be similar to having two classrooms sharing an AP, but supporting different applications.
In this test, when the Aerohive AP250 was oversubscribed, it favored data clients over real-time applications like video. This situation might happen when one classroom is watching streaming video, and the other class is using the AP for data. The Aerohive AP250 doesn’t have the capacity to serve both, and the video users suffer. This is unacceptable in K-12 where video is now key to the learning experience.
All Wireless LANs need to be designed properly to meet the organization’s requirements. Mojo believes a primary requirement is excellent user experience. (Tied with security, where we rock!) And because not all APs are created equal, we also believe that an AP’s capabilities are an important design consideration. If you use underpowered APs, one AP per classroom may make sense. With more powerful, state-of-the-art access points, one for every two classrooms might well be all you need.
When spending your E-Rate dollars, design your WLAN for excellent user experience by taking application type and AP capacity into account. Not only is the C-120 a very high performing AP that balances data and real-time applications, it’s the most cost-effective enterprise class AP on the market, with excellent price and excellent performance.