The 802.11ac Shakedown

Share Post

In 2002, Aruba took on the Tasman Family by forming a company determined to deliver Wi-Fi as an alternative to the wire. Today, Aruba is a leading provider of Wi-Fi with a long list of customers that have successfully cut the cord.GF.png

The 12-year journey was no walk in the park.  The Tasman Family employed some pretty heavy-handed scare tactics to thwart Aruba's success. Most recently, Tasman's 'Don' made some bold claims about how smaller best-of-breed technology companies would vanish by 2018. A repeat of a similar claim made about a decade ago at one of the family's extravagant Sales Kickoff Meetings - compliments of their switching business.

Over the past 12 years Cisco acquired three different Wi-Fi companies to compete against Aruba, but they didn't have the same underlying determination of enabling the all-wireless-workplace.

Although Cisco makes some pretty audacious performance claims, you don't hear them talking about eliminating switches or desk phones. Incumbent's dilemma?

Instead, they'll push you to upgrade your wiring closets to support mobility or payShakedown.png

the Customer Loyalty TaxI guess it's better than waking up with a horse's head in your bed. 

Taking on the big guys is a relentless task and that's why Novarum's 802.11ac performance report is a breath of fresh air.  Free of strong-arm tactics, this report captures the performance differences in a real-world office environment between Aruba's AP-225 and Cisco's Aironet 3702i, 11ac Access Points.

Here are some of the highlights from Novarum's 802.11ac Performance Report.

High-Density Throughput Test: Aruba is 2x Faster than Cisco.

This test measured the TCP throughput performance of Cisco's Aironet 3702i and Aruba's AP-225 in a high-density environment consisting of 48 MacBook Airs (2x2:2) and 12 MacBook Pros (3x3:3).  The Aruba AP-225 averaged 302 Mbps in bidirectional throughput, which is more than double Cisco's 3702i.  



High-Definition Video Streaming: Aruba handles 27% More Video than Cisco.

This test looked at how well Aruba's AP-225 and Cisco's 3702i can meet the demands of high-bandwidth, lantency sensitive video.

Aruba's AP-225 delivered 27% more concurrent high-definition video streams than Cisco's 3702i. Part of the reason Aruba outperformed Cisco is the fact that the AP-225 has a dual-core 800-MHz CPU, in addition to CPUs on the radio chipsets. The Cisco 3702i 800MHz CPU can only operate as a single core and the CPU tapped-out at 94% utilization with only 56 clients.  Aruba's CPU averaged around 60% utilzation with 71 clients streaming video.



High-Density Client Roaming: Aruba ClientMatchTM delivered 39% More Throughput.

This test evaluated the impact of sticky clients on WLAN performance. When clients first associate to the network, they often hold-on or stick to the first AP they connect and as they roam away from the AP, data rates falloff.

The test simulated a common scenario where a group of people congregate outside of a classroom, conference room, or auditorium, and then spread out when they enter the room.

Before the roam, all the clients were in the Initial Client Location, as identified in Figure 3. As expected, the majority of the clients associated to AP4, which is represented by the blue in the pie charts, for both vendors when the test started as shown in Figures 4 and 5.


After the roam, Aruba successfully distributed the clients across all three APs to make the best use of available resources. This was a direct result of having Aruba ClientMatchTM enabled. See Figure 4. Cisco was not able to effectively redistribute the clients across the APs after the roam as only four clients changed their AP association. See Figure 5.



To quantify the performance benefit of fixing the sticky client problem, Novarum ran a series of Ixia Chariot Download TCP throughput tests after the clients roamed. Or in Cisco's case, the clients didn't roam and remained stuck on their original AP.

After the roam, Aruba delivered an aggregate throughput of 752 Mbps total for all of the clients. The Cisco infrastructure delivered 540 Mbps of aggregate throughput in total, as seen in Figure 6. Aruba's overall system throughput was 39% higher than Cisco.


Novarum attributed Aruba's performance advantage to ClientMatchTM, which continually monitors network and client performance and makes sure every client is connected to the best possible radio. Irrespective of what Cisco's marketing team claims, they don't have the intelligence in the network to match clients to the best possible radio as they roam.

As you can see from the test results from the Novarum report, Aruba's software innovation and AP design are enabling customers to cut the cord and avoid the Tasman shakedown.