Why a survey beats guessing
Wi-Fi that frustrates users is rarely a hardware problem — it is almost always a design problem. Too few access points, the wrong placement, overlapping channels or radios shouting over each other will undermine even the best equipment. A wireless site survey is the discipline that replaces guesswork with a plan: how many access points, placed where, on which channels and at what power, validated against the way the space will actually be used.
The payoff is a network that performs on day one and scales predictably. This guide walks the full lifecycle — requirements, predictive design, on-site validation and post-install tuning — and is a natural companion to our deep dive on designing high-density Wi-Fi 6 for airports and venues.
Start with requirements, not access points
Before a single AP is placed, write down what good looks like. How many users and devices per area? Which applications must work — voice, video, barcode scanners, payment terminals? What is the worst-case concurrency, such as a shift change or an event? These answers separate a coverage problem (make sure there is signal everywhere) from a capacity problem (make sure there is enough throughput where people cluster).
Getting this wrong at the start cascades through the whole design. A warehouse that scans barcodes needs reliable coverage in aisles and at loading docks; a lecture hall needs raw capacity for hundreds of simultaneous devices. The hardware might be the same Wi-Fi 6 access points, but the placement and density differ completely.
- How many users and devices per area, at peak?
- Which applications must work — voice, video, scanners, payments?
- Is this a coverage problem or a capacity problem?
- What is the worst-case concurrency, such as a shift change or event?
Predictive design: planning on the map
With requirements in hand, the next step is a predictive design: modelling the space on an accurate floor plan, assigning wall and material types, and letting design software estimate coverage and capacity as you place virtual access points. This is where the AP count and rough placement are decided, and it is efficient — for many standard offices an accurate predictive design is all you need.
The quality of a predictive design depends entirely on the accuracy of its inputs. A floor plan that ignores a concrete lift core or a glass atrium will produce a plan that looks fine on screen and fails in reality. Mark the materials honestly, and the model rewards you with placements you can trust.
On-site validation and AP-on-a-stick
For RF-hostile environments — warehouses with metal racking, hospitals with lead-lined rooms, heritage buildings with thick stone — a predictive model needs a reality check. An on-site survey measures actual signal, while an AP-on-a-stick survey mounts a real access point temporarily and walks the space to confirm how the building treats the signal.
This step catches the surprises that floor plans hide: a mezzanine that reflects signal, a cold-storage room that swallows it, a courtyard that needs an outdoor AP. Spending a day validating before you cable and mount dozens of access points is far cheaper than re-doing the install afterward.
Channel and power planning
Coverage is only half the job; how the radios share the air is the other half. In the crowded 2.4 GHz band there are only three non-overlapping channels, so adjacent access points must be planned carefully to avoid co-channel interference. The wider 5 GHz and 6 GHz bands give far more room, which is one of the big advantages of modern Wi-Fi — a theme we cover in Wi-Fi 6 vs Wi-Fi 5 vs Wi-Fi 6E.
Counter-intuitively, the answer in dense spaces is usually more access points at lower power, not fewer at high power. Turning every radio to maximum creates a shouting match where clients cling to distant APs and everyone’s throughput suffers. A good plan tunes power so each AP serves a clean cell, and lets a controller adjust channels dynamically as the RF environment changes.
Backhaul, power and the wired side
Every access point needs power and a path back to the core, and a wireless survey that ignores the wired side is only half a plan. Each AP consumes a PoE or PoE+ port, so your access switches must have both the ports and the total power budget to feed them. Busy floors of Wi-Fi 6 APs can aggregate enough traffic to justify 10G uplinks from the wiring closet.
Plan the switching and the wireless together. The number and placement of access points drives the port count, PoE budget and uplink capacity of the switching layer beneath them — design them in isolation and you end up re-cabling.
Designing for roaming and seamless handoff
Coverage and capacity get the attention, but for any space where people move — corridors, warehouses, airport terminals, hospital wards — roaming is what users actually feel. As a device moves between access points, the handoff must be fast and invisible, or a voice call drops and a video call stutters. Good roaming starts in the survey: overlapping cells sized so a client always has a strong next AP to move to, with power tuned so it lets go of the old one at the right moment.
Fast-roaming standards and a controller that coordinates the access points do the rest, steering clients to the best radio and easing them between cells. Plan roaming explicitly for mobile environments rather than hoping coverage overlap is enough — it rarely is by accident.
Segmentation and security in the wireless plan
A wireless design is also a security design. Different user groups — staff, guests, IoT devices, payment terminals — should land on separate VLANs with policies appropriate to each, so a compromised guest device cannot reach corporate systems. Guest access in particular needs a captive portal and isolation, and in India it carries log-retention obligations you must design in from the start.
Bake segmentation into the survey so the SSIDs, VLANs and security policies are decided alongside coverage, not bolted on afterward. A controller then enforces the same policy consistently across every access point in the design.
Common survey mistakes
Three mistakes recur. The first is designing for coverage when the problem is capacity — a few high-power APs blanket a lecture hall with signal yet collapse under hundreds of devices. The second is inaccurate floor-plan inputs, where unmarked walls and materials produce a predictive design that fails on site. The third is skipping validation, mounting everything from a model and discovering the dead spots only after users do.
Avoiding them is mostly discipline: classify the problem honestly, mark materials accurately, validate RF-hostile spaces before you cable, and run a post-install survey to confirm. The hardware rarely lets you down; the planning is where networks are won or lost.
- Designing for coverage when the real problem is capacity
- Feeding the model inaccurate floor-plan inputs
- Skipping validation in RF-hostile spaces
- Turning every radio to maximum power instead of tuning cells
Outdoor and special environments
Outdoor coverage — courtyards, loading yards, stadium concourses, station platforms — follows different rules from indoor design. Signal travels further with fewer walls, but weather, mounting height and interference from neighbouring networks all come into play, and access points must be rated for the environment. Immunity’s Lotus Alpha range includes hardened outdoor models for exactly these spaces.
Special indoor environments need their own thought: cold storage attenuates signal heavily, warehouses with tall metal racking create reflections and shadows, and heritage buildings with thick stone resist penetration. Each of these is a case where an on-site or AP-on-a-stick validation pays for itself, because no predictive model fully captures how an unusual space treats RF.
Tools, deliverables and what to expect
A professional survey produces tangible deliverables, not just an opinion. Expect a predictive design with access-point placements on the floor plan, a channel and power plan, a bill of materials listing APs, switches, PoE and cabling, and — where validation was done — heat maps of measured coverage and capacity. Those documents become the blueprint the install team follows and the baseline you validate against afterward.
Modern surveys increasingly lean on the management platform itself. Once the network is live, a cloud controller continuously measures the very metrics a survey checks — coverage, interference, client experience — and tunes the network as conditions change. Send us your floor plans and user counts, and we will turn them into a full design package and show how Net Cloud keeps it optimised after go-live.
Documenting the design for the install team
A survey only delivers value if the people mounting access points follow it faithfully. That means clear documentation: a floor plan marking each AP’s position and mounting height, the channel and power plan, the cabling and PoE requirements, and notes on any awkward locations identified during validation. A precise as-designed package removes the guesswork that otherwise creeps in on site and produces the coverage gaps a survey was meant to prevent.
It also creates a baseline to validate against afterward. When the post-installation survey walks the live network, the install team and the design can be compared directly, and any deviation — an AP mounted in the wrong spot, a missed cable run — is caught and corrected rather than quietly degrading the network everyone signed off.
Keeping the design alive after go-live
Spaces change: walls go up, furniture moves, headcount grows, and a new neighbour’s network appears on your channels. A survey captures a moment; a healthy wireless network keeps adapting after it. Ongoing monitoring of RF health, interference and client experience lets the network retune channels and power automatically as conditions shift, so the performance validated on day one does not slowly erode.
This is where a one-time survey becomes continuous optimisation. Immunity’s Net Cloud watches the very metrics a survey checks and adjusts the live network as the environment evolves, turning your initial design into a baseline the platform actively maintains. Send us your floor plans and we will produce the predictive design — and show how it stays optimised long after the access points are mounted.
Validate, tune and keep watching
The survey is not finished when the access points are mounted. A post-installation survey walks the space with the real network live, confirms coverage and capacity match the design, and feeds adjustments back into channel and power settings. After that, the network should keep watching itself: a cloud control plane continuously monitors RF health, client experience and interference, and tunes automatically as the environment changes.
That is exactly what Immunity’s Net Cloud does — turning a one-time survey into ongoing optimisation, so the network you validated on day one keeps performing as furniture moves, walls go up and user numbers grow. If you would like a predictive design for your space, send us your floor plans and we will produce an AP plan, channel scheme and bill of materials to match.