We all know performance testing is as important a task as any in the WiFi business. We need to know the limits of our access points after all, and it certainly helps you to have all the information at hand to make the most informed decision you can. But performance testing focused soley on speed? Well, while it's important, it doesn't give you all the information you need. Because you, dear reader, are not just upgrading your WiFi for speed. You want all the bells and whistles that come standard with enterprise WiFi platforms today. And you want to turn them all on. And most importantly, you want to know if and how they affect speed.
Our Next Step Into Real World Testing - Performance + Security
Building from our earlier benchmark tests assessing raw performance amongst the top 802.11ac Wave 2 access points, we've developed a new series of tests that place these same devices (and some new ones which have recently come to market) into situations that test just how well they handle multiple features running at the same time. We’ve said often that we strive to test according to “real-world scenarios”, and nothing is more representative of a real-world scenario than an access point operating with all the bells and whistles flipped on. So to begin this series, we conducted a benchmark test that compared enterprise class access points to evaluate how well they handled concurrent mixed application performance along with active wireless rogue detection and prevention.
Regular readers will not be surprised to find out that access points with a third radio outperformed those with only two radios, since having that third radio dedicated to scanning allows the two access radios to focus on the task at hand. Two-radio access points must split their attention between serving clients and performing other important functions like dealing with rogue access points, monitoring channels for radio management, and troubleshooting active issues, making them less effective overall. Our mission was to see just how less effective they are.
For the performance part of this test, we kept both the 2.4 and 5GHz radios busy with a mixture of data, voice, and video traffic, sourced by IxChariot. The results were evaluated against the following services levels:
|Data||>= 1 Mbps|
|Voice||>= 4.2 MOS|
|Video - Delay Factor||<= 50 ms|
|Video - Media Loss Rate||<= .004|
In order to receive a passing grade, a client needed to meet or exceed its respective service level(s) based on the application it ran. The concurrent rogue detection part of this test introduced a rogue access point on the wired and wireless network to see how long it took the access point under test to detect and shut down the rogue.
We will continue to develop tests that put the real-world into our lab to ensure you know exactly what to expect in any given scenario. We welcome ideas and scenarios from all sources, so don't hesitate to contact us if anything comes to mind.