A site survey is a really important task to perform. The evidence of this may not be quickly felt. Down the road when you encounter a problem you have the survey to fall back on. Sure there are several reasons for the wireless not to work. I have personally asked a customer about the coverage and whether or not a survey was performed. Most likely then not the answer was no. As luck would have it the problem turned out to be a gap in coverage. This was a shiny new network that added an AP every other classroom and resulted in unforeseen gaps. Had they had a site survey performed they could have remedied this to start with.
We all know that the biggest reasons to forego the survey can be money reasons or time constraints. Whatever the reason, sometimes it just doesn't get accomplished. I believe network planning is a must and part of planning is the survey. Some networks may operate fine, with very little planning based on the premise of scattering out random AP at intervals but most will not. I think the best foundation for a network is a complete survey suite consisting of predictive , active, and followed up with a validation (passive) survey. If you had to pick 1 and only 1 form of survey I would say the validation would be the next best alternative to the full survey. The validation will locate the trouble areas before the day of the rollout and allow you to remediate the gaps in coverage.
When you get ready to perform a survey you should take everything you might possibly need. (This was beaten in my head from the start.) You should plan for a worst case scenario for your equipment. You could have batteries die, bricked laptop, you may need to capture some packets or do some spectrum analysis. Unless the survey is next door to your home or office it will probably be inconvenient to make another trip to see what the microwave will do to the network.
Another important part of the survey will be spectrum analysis. Spectrum analysis is not a site survey, but it should be used in tandem with the active survey to determine problematic areas of interference. One large source of interference can be the break room. Most likely the break room will have a microwave oven in it which plays havoc with 802.11 transmissions as the microwave is not polite to the rf spectrum. All sources of interference should be documented with the findings from the spectrum analysis.
Networks are used for many different applications, and your survey should reflect the needs of the network. If a network is destined to be used for VO-WiFi then you should incorporate an active call into your survey and record information on the call quality.
What process do you use in site surveys? What gear do you take? What are your favorite tools?
