In the Wi-Fi industry we have been throwing around this word, ‘Survey’ for a long time. Many use it incorrectly and it has taken on many meanings over the years. This is the fault of those inside the industry as well as those just trying to talk like they know what they are doing.

The first usage is normally referring to something that is really defined as ‘Wireless LAN Design’ – but many have taken a single approach, putting an access point on a tripod (thus the AP-on-a-Stick moniker) and moved it around over and over in the design space trying to see where the best placement of access points would be for the best coverage. Since this was a ‘site survey’ many confused it with the actual process of designing a wireless network.

There are many more facets to WLAN design than merely thinking about coverage. I’ve been teaching advanced wireless design classes for well over a decade and use the following term repeatedly until it sinks in the students’ heads… “Coverage is Easy” – repeat three times to yourself. “Coverage is Easy, Coverage is Easy, Coverage is Easy”.

Perhaps 10-15 years ago the idea was to design for coverage. But we have been far past that for the last 10 years. We now know we need to design for more than mere coverage. We need to design to meet ALL the network design requirements. Here are just a few we normally define metrics and requirements so we’ll know when the WLAN is ready to work without issues.

• Primary Coverage – quoted in RSSI – perhaps something like -67dBm
• Secondary Coverage – also quoted in RSSI – this is the ‘overlap’ but from the point of view of the Wi-Fi clients. They need a ‘backup’ AP
• Signal to Noise Ratio – this is the difference between RSSI and the ambient RF noise – the higher the better – it allows for faster data rates
• Co-Channel Interference – VERY important to make sure we are accurately getting frequency reuse. If two or more radios are working in the same frequency they ‘share’ the capacity, NOT double capacity
• Devices per Radio – a ratio to see how many devices will be sharing the bandwidth on a certain frequency in a given coverage area

And those are just the start, we have Jitter, Latency, SINR, throughput rates, CRC Rates, Retry Rates, and many more.

Well then – how can we validate if our Wireless LAN meets or exceeds the design requirements?

To answer that simple question, let me refer back to how we deal with cable plant in our facitilites. Lets say you wanted gigabit to the desktop, and determined you need Category 5e or better Ethernet cabling to support your design requirement. You might purchase Cat5e-spec’d cable, but you won’t know for sure if it will carry gigabit traffic until you validate it and prove it meets the specific requirements for near-side cross-talk, far-side cross-talk, pin-outs, twist ratios, etc. Only after it has been validated will you know for sure that it meets your specs.

In the case of a wireless network, you may want to survey the site before you install, like an AP-on-a-Stick process… but you MUST do a post install validation survey to know for sure if the installed WLAN is meeting your specs.

Tools like Ekahau’s Site Survey software allow one to perform this post validation survey and ‘prove’ your installed systems meet or exceed your WLAN requirements. Without doing a validation survey, how can you know if your wireless network is working as designed? You can’t.

Don’t simply trust the install worked as you’d planned – test your network before you get called back to fix it later. Validate your Wireless LAN – each and every time!

Keith R. Parsons CWNE #3

Twitter: @KeithRParsons
Email: keith@wlanpros.com
Website: www.wlanpros.com