A WiFi site survey is a set of measurements and observations that shows how wireless signals behave in a specific building and whether the network can support the people and devices that depend on it. It is the basis for designing a new network, fixing a poor one, or proving that an installation does what it was meant to do.
Radio signals do not respect floor plans. Concrete, brick, metal studs, elevator shafts, mirrored glass, stocked shelving and even people absorb or reflect WiFi signals in ways that are hard to predict from a drawing. Two buildings with identical layouts can behave very differently because of what the walls are made of or what sits on the racks.
A survey replaces assumptions with evidence. Instead of adding access points wherever people complain, you find out whether the problem is weak signal, interference, too many users on one access point, a poor channel plan or a configuration fault. Each of those has a different fix.
What a survey measures
A survey collects readings as a surveyor walks the space with a laptop or handheld running survey software. At ehZee we use AirMagnet Survey Pro. Typical measurements include:
- Signal strength from each access point, usually expressed in dBm. See RSSI, SNR and signal strength for what the numbers mean.
- Signal-to-noise ratio (SNR), the gap between the WiFi signal and the background noise. A strong signal in a noisy environment can still perform badly.
- Channel use and overlap, showing where several access points on the same channel compete with each other.
- Neighbouring networks from tenants and adjacent buildings that share the same airspace.
- Data rates, throughput and connection quality, when the survey connects to the network as a client.
The software turns those readings into colour-coded heat maps laid over the floor plan, which make problem areas easy to see. WiFi heat maps explained covers how to read them.
Types of WiFi survey
Predictive survey
A predictive survey is a computer model built from floor plans and assumed wall materials. It is useful for budgeting and early design, especially before a building exists, but it is only as accurate as its assumptions. Predictive vs on-site WiFi surveys compares the two approaches.
Passive survey
The survey device listens to every access point it can hear without connecting. This shows coverage, signal strength, SNR, channel overlap and neighbouring networks.
Active survey
The survey device connects to the network and measures how a real client performs: data rates, throughput, packet loss and how it roams between access points.
Access point on a stick (AP-on-a-stick)
For new designs or major changes, a surveyor temporarily mounts an access point at a proposed location and measures how far its signal reaches through the actual building. This validates or corrects a predictive design.
Spectrum analysis
A spectrum analyzer looks for non-WiFi sources of interference, such as microwave ovens, wireless cameras or industrial equipment, that ordinary WiFi tools cannot identify.
What else a good survey covers
Radio readings are only part of the picture. A survey that leads to a working network also records the physical and business details that shape the design. Our checklist covers:
- Objectives: number of users, density, applications and bandwidth needs
- Installation: cabling, ceiling type, mounting options and IDF/MDF space
- Security: authentication through RADIUS and Active Directory, and physical security of equipment
- Extras: guest WiFi, RFID and asset tracking
- Radio environment: metal structures, neighbouring networks and SNR
When you need a survey
- Before designing a new network or replacing an old one.
- When users report dead spots, dropped calls, slow connections or devices that lose their connection in certain areas.
- When the building changes: new walls, racking, a change of tenant or a big increase in users.
- When adding voice over WiFi, location tracking or other services with stricter requirements.
- After an installation, to confirm that the network meets the design targets.
Start with requirements. A survey is only meaningful when it is measured against something. Agree on the coverage and signal targets for each area, based on the applications people use, before the survey starts.
How a survey is carried out
The steps are similar whether the survey supports a new design or troubleshoots an existing network. What gets recorded at each step is what makes the results usable later.
- Agree the purpose and targets. Record why the survey is being done, the areas included, the devices that matter and the signal and SNR targets for each area.
- Prepare the floor plans. Load scaled plans into the survey software and calibrate them, so every reading lands in the right place. Note the date of the plans and any known differences from the building as it stands.
- Plan the routes. Choose walking routes that cover every area at a consistent pace, including corners, stairwells and rooms people actually use. Record areas that cannot be accessed.
- Collect the readings. Walk the routes with the survey device, separately for each type of survey being done. Note construction details, ceiling types, mounting conditions and anything unusual, such as a fully stocked warehouse aisle or a room full of people.
- Check the data. Review the readings on site for gaps or obvious errors and re-walk any section that needs it before leaving.
- Analyze and report. Compare the results with the targets, identify causes rather than symptoms, and write recommendations in order of impact.
Common survey mistakes
- Surveying with the wrong device. A laptop with a good antenna hears the network better than a handheld scanner or phone. Targets and active tests should reflect the weakest device that matters.
- Surveying an empty building. Stock on racks, furniture and people change how signals travel. Note the conditions and, where possible, survey when the space is in normal use.
- Covering only the problem area. Interference and channel conflicts often come from access points elsewhere, including neighbouring floors.
- Looking at signal strength alone. Strong signal with poor SNR or heavy channel overlap still performs badly.
What the report should give you
A useful survey report goes beyond pictures. It should state the targets used, show heat maps for each floor and band, explain what the maps mean in plain language, and list specific recommendations: which access points to add, move or adjust, and what to change in the configuration. It should also be clear about what was not covered, such as areas without access.
How ehZee can help
We carry out passive, active and validation surveys with AirMagnet Survey Pro, starting with a conversation about why you need the network and what it must support. Our WiFi surveys service gives you heat maps, plain-language findings and a ranked list of fixes, and can lead into design and build if you want us to carry the work through.
Common questions
How is a survey different from a speed test?
A speed test measures one device at one spot at one moment, mostly reflecting the internet connection. A survey measures the wireless network across the whole space and shows why performance changes from place to place.
Do we need to provide floor plans?
Yes, scaled floor plans make the heat maps accurate and easy to read. If you do not have current plans, let us know before the survey so we can plan around it.
How often should a network be resurveyed?
There is no fixed interval. Resurvey when the building or its use changes, when new types of devices or services are added, after major equipment changes, and whenever complaints suggest the network no longer matches how the space is used.
Can a survey be done while people are working?
Yes. Surveying during normal use is often better, because it shows the network under real conditions and with real interference.