Best practices for 802.11n Wi-Fi design, deployment

Real-world Wi-Fi results shape recommendations and test plans

As 802.11n takes hold as the enterprise Wi-Fi standard, wireless tester VeriWave has recommendations for designing and testing enterprise wireless LANs.

 Just in time for 802.11n's first birthday as an official IEEE standard, VeriWave is releasing a guide to best practices for enterprise Wi-Fi networks. The new guide includes test plans and evaluation criteria, all based on the test vendor's extensive real-world experience with Wi-Fi equipment, both access points and client radios, in operational networks.

802.11n Wi-Fi making huge impact 1 year after standard ratified

The new document, titled "WiFi Pre- and Post-deployment Verification Best Practices Guide", outlines specific steps, including tests, that network managers can take to design, deploy and validate an enterprise Wi-Fi network before it goes live.

The goal is to give network managers a realistic set of Wi-Fi expectations, which reflect their vertical market requirements, and take into account the vagaries created by different classes of client devices and different traffic loads, according to Eran Karoly, vice president of marketing for VeriWave.

The recommendations are broken down according to vertical market, such as healthcare, general office, hospitality, education, retail and so on. For each, VeriWave specifies a range of upstream and downstream throughput values for a targeted optimum service level and for the minimal acceptable service level.

Apart from the document's recommendations, the test scenarios can be run using VeriWave's WaveDeploy product, introduced last May. 

WaveDeploy includes a small WaveAgent, which runs on each client device and shares Wi-Fi data with the companion PC- or server-based test and reporting application. Enterprise network staff can see a wide range of metrics that together reveal the quality of the user experience -- two-way TCP/IP and UDP throughput, Mean Opinion Scores (MOS) for measuring VoIP quality, video quality using the Media Deliver Index, packet loss, and roaming. (More details are at the vendor's Web page.)  

Optionally, enterprises can deploy a separate box that runs a scaled-down version of VeriWave's Wi-Fi traffic generator/analyzer software, called WaveDeploy Wi-Fi. This system creates up to 64 virtual 802.11 clients and their applications, which mimic real clients like a laptop or smartphone, and creates different traffic characteristics, including load.

The best practices guide and a data sheet on WaveDeploy can be found online at VeriWave's site.  

The best practices guide can be a valuable corrective for theoretical 802.11n performance data. VeriWave drew from its extensive work with WLAN equipment vendors to identify "ideal" network configurations, then replicated these with internal test networks. These were then subjected to load testing using VeriWave's software and test equipment. Finally, the vendor polled its enterprise WLAN customers about their Wi-Fi experience.

"The result is what a WLAN administrator realistically could expect," Karoly says. "A 3x3 MIMO [multiple input multiple output] 802.11n network should ideally have downstream TCP/IP traffic of 300Mbps. But in a real-world network, 50Mbps is a reasonable, achievable target. We don't specify theoretical numbers, but ones that are based on testing and real-world experience."

One VeriWave customer, a hospital, used WaveDeploy to decide among four tablet computers (HP 2740p tablet, HP Mini, a Panasonic Toughbook model, and Apple's iPad) to be deployed on its 802.11n WLAN for medical staff. In this particular case, considering the application mix, the target service was 30Mbps, with 15Mbps as the minimum. In tests, the Toughbook performed best, averaging nearly 29Mbps for data-only traffic at various locations, except on one specific area of the hospital. The iPad performed worst, averaging just over 11Mbps.

Notably, all the clients suffered a big drop in throughput when the traffic was mixed: data, voice and video. The Toughbook's average was cut in half, for example, but was still the top performer, and that's the one the hospital eventually chose.

John Cox covers wireless networking and mobile computing for Network World.

Twitter: http://twitter.com/johnwcoxnw

Email: john_cox@nww.com

Blog RSS feed: http://www.networkworld.com/community/blog/2989/feed

Insider Shootout: Best security tools for small business
Editors' Picks
Join the discussion
Be the first to comment on this article. Our Commenting Policies