In doing some research for April?s feature story on wireless deployments, I came across an interesting difference of opinion when it comes to doing site surveys. Some of my sources are all for them, others are dead set against them. Why the split opinion?
Site surveys refer to the radio spectrum of your site, and looking at what signals presently exist that might interfere with providing Wi-Fi coverage around your office building and campus. They also examine how many wireless access points will be needed to provide enough coverage so that all your users can receive strong enough signals.
The pro surveying camp goes like this: you want to estimate the number of access points you?ll need before you put together any request for proposals and send them out for cost estimates. Every radio environment is different, these folks say, and you want to get a handle on what is happening in your office before you deploy. Some of the vendors make it easy to import floor plans from various CAD programs into their products and illustrate where you intend to place your access points and how far their signals will actually reach.
The con surveying camp says, not so fast. Yes, every environment is different, and what is more difficult to do any actual estimation. It is far better to start with a trial deployment of the actual product that you intend to use and then see what you really need based on real access points talking to real laptop clients. No survey can capture this reality, and thus you are better off skipping them and going straight to the field testing stage.
Field testing will also tell you what issues you will have with wiring up your access points in the selected locations, and whether that will be an issue too. The other thing field testing tells you is where you have radio signal beyond your office?s boundaries, which is important for security reasons when you examine who can sit in your parking lot or in an adjoining office and receive your radio signals just fine.
I used to be firmly in the site surveying camp, but after hearing these latest arguments I understand why having the actual product makes sense. Site surveys are fine for simple layouts and limited configurations, but especially when dealing with large deployments they have limited use.
For a big rollout like a college campus, a survey could take several days and lots of trial-and-error placements of access points, and for that you are far better off using the actual product than any predictive tool.
?If you are off just ten or 15 percent in how many access points or users per access point you need, the ramifications could be expensive," says Kurt Sauter, the director of product marketing for Xirrus.
But before you deploy, spend some time calculating how many wireless users you intend to serve and what kind of bandwidth they will need when they are on Wi-Fi. This information will come in handy once you start setting up your network.
The good news is that the wireless vendors are making better tools to estimate coverage and make recommendations about access point placement and bandwidth usage. The bad news is that you have to purchase them (in most cases) first, so it is harder to make a determination about which vendor?s architectural claims are going to resonate with your own needs.
Aruba is one vendor that offers a license-free version of its management tool, so you can get familiar with it before you make a purchase decision, a notable differentiator. Because sometimes you have to start operating a wireless network in the real world before you can figure out your coverage gaps.