Q: In planning for a new implementation, is a site survey a waste of time and energy or not?
- Patrick B.
Site surveys were originally intended to make the most of scarce radio access. Now access points are cheap and smart, some vendors have suggested that there is no need to do a site survey: just put up the access points and let RF management sort it out - and if there are any problems, simply add more APs till the network works. Aruba's Grid Point architecture is one extreme example (and Aruba's CEO points out that site surveys measure a moving target anyway). We have covered the debate in depth, along with the related issue of whether we still need to stick Wi-Fi APs in the ceiling to maximise access.
Now, with the world holding its breath, awaiting the final answer, we ask the Wizards to gaze into their crystal ball and respond:
Seth Goldhammer, Roving Planet
In my experience, I do not believe a site survey is a waste of time and energy. Many wireless networks started by putting up access points in perceived central locations. If this network is only used for convenience, this still may be suitable. But with this approach, you would be surprised with the amount of dead spots, channel conflicts, and areas of low power (therefore low bandwidth connections) you will find.
Often, this is emblematic in calls to the help desk, as users experience inconsistent behaviour on the wireless network. Troubleshooting becomes tricky - was the user's experience due to the end device, the access point, or the application server?
Many hours can be spent troubleshooting by both looking and attempting to correlate logs, and going out in the field to inspect the environment with a wireless sniffer. The time spent on a good initial site survey to really know your RF environment can prevent many headaches down the road.
Dan Simone, Trapeze Networks
With the planning and deployment tools available today from application and wireless switch vendors, the traditional site survey consisting of climbing ladders in prospective access point locations and walking around to test coverage may not be the best use of your time and energy.
What remains useful (always) is to plan a deployment for both coverage and bandwidth capacity, so you know how many access points to deploy and where to put them. The better planning tools will ask you to depict the types of wall construction you have in order to place the access points and eliminate guess-work. Familiarising yourself with the types of wall construction in your environment is worthwhile.
Some level of site survey is going to be needed, but the amount of surveying will vary depending on the WLAN system you choose. Site surveys are mainly used to determine access point placements for optimum coverage in a multi-access point deployment, based on minimising co-channel interference among access points while maximising each access point's range. The real complication is the RF planning element of the deployment, typically requiring an RF Planning solution to determine signal levels at various points to optimise coverage and minimize interference.
Joel Vincent, Meru Networks
Traditional WLAN systems support manual optimisation of off-the-shelf access points by allowing users to adjust access point power levels and RF channels via the WLAN switch or controller. Using these tools simplifies the site survey and RF plan by using complex mathematical models to predict where access points should be placed.
However, there is still some level of “walk-around” site survey (i.e. physically walking with a radio meter throughout the building) that is recommended - to make sure there are no dead spots in the deployment. These tools must be continually re-used as the network or environment changes, driving up ongoing management costs.
Other WLAN systems use site survey tools for the initial deployment, and then do some level of automated re-adjustment to help with ongoing management tasks. However, this method still requires the prerequisite walk-around to double-check for dead spots. The simplest site surveys and ongoing management come from WLAN systems that can coordinate RF transmissions across access points to eliminate co-channel interference completely. With these systems, access points can be placed wherever coverage is needed without fear of channel interference, thereby providing a seamless blanket of coverage.
Who's right? Would you be happy never to do another site survey, or is the technology not yet good enough to do without it? Join the discussion in our Forum.