Part 2 — How to Design a Wi-Fi Geolocation Network: Cisco Meraki AP Layout Step-by-Step
In Part 1 of this series, we saw how a non-optimized access point (AP) layout can create “dead zones” in a store—areas where devices aren’t consistently heard by enough APs to be located reliably. In Part 2, we’ll focus on the practical side: how to design an AP mesh that makes Wi-Fi geolocation stable and predictable, using Cisco Meraki hardware as a reference.
Wi-Fi geolocation is often described as “triangulation.” In practice, it’s usually based on comparing signal strength (RSSI) measurements from multiple APs. That means the layout matters a lot more than most people expect: overlap, AP spacing, ceiling height, walls, shelving, and even the content of the space can all change results.
This guide is meant to be a best-practice reference you can apply quickly on a floor plan. Every building is unique though, so final deployment plans should always be reviewed and validated by qualified network/system engineers.
1. Building the optimal network mesh
When you design Wi-Fi for geolocation, you’re designing for something slightly different than “good Wi-Fi” for browsing. The goal isn’t only coverage—it’s consistent overlap.
A simple way to think about it: wherever a customer stands, their device should be picked up by at least three APs with a signal that’s strong enough to be useful. That “three APs everywhere” concept is the foundation of stable Wi-Fi location.
You’ll sometimes see signal targets mentioned in dBm. If dBm isn’t your world, don’t worry—you don’t need to become an RF expert to draw a good design. The only thing to remember is that for geolocation you generally want a fairly strong signal to multiple APs; a commonly used target is around –67 dBm or stronger (stronger means the number is closer to zero, like –60 being stronger than –75). In this article, we’ll keep the focus on a layout method that naturally pushes you toward that outcome, then you validate it in the real world during survey and tuning.
2. Triangulation Pre-requisites
So what does “location-ready” look like on paper?
A good starting point for many retail environments is to keep APs roughly 12–15 meters (40–50 feet) apart. When you do that with a consistent pattern, you end up creating the overlap needed so most points in the store will be heard by three nearby APs.
In more technical details, you have to keep the RSSI ≥ –67 dBm, and the SNR ≥ 25. The AP spacing numbers (12m, 15m, etc.) are just a convenient way to aim for this outcome, but a more precise result can be achieved by doing an on site survey to analyze the RSSI and SNR.
Designs that increase AP density like this also come with real RF challenges: channel reuse, minimum bitrates, and transmit power often need to be tuned more carefully than in a basic Wi-Fi deployment. Cisco Meraki has strong documentation on high-density environments, and it’s worth reading if you’re going beyond “standard coverage” and aiming for precision.
3) Step-by-step guide using hexagons
First rule: Stop using circles to position APs.
Circles look intuitive, but they’re surprisingly hard to use when you want a clean, repeatable layout. Where exactly do you place the next circle so overlap is uniform? How do you avoid under-covered pockets? It quickly becomes guesswork.
Instead, we’ll build a honeycomb pattern and represent each AP’s “planning range” with a hexagon. Hexagons tile perfectly, which makes it much easier to place APs consistently and spot weak zones before you ever install hardware.
Now let’s build the template.
What you need
A floor plan you can trust
Start with a floor plan that’s up to date and at scale (meters or feet). It should clearly show walls, partitions, large obstacles, and “signal killers” like storage rooms, elevators, thick concrete cores, or dense shelving—anything that could realistically affect radio propagation.
If your drawing tool supports it, lock the scale so your shapes don’t slowly drift as you copy/paste and align the grid.
Choose your target AP spacing (your “grid size”)
Before drawing the first hexagon, decide what spacing you’re designing around. This is your planning distance between APs—simple, practical, and easy to apply on a map.
-
15m (~50ft) is a common starting point for open retail spaces with relatively clean sightlines.
-
12m (~40ft) is a common adjustment for complex shapes, heavier obstructions, or environments where signals drop faster than expected (dense shelving, lots of metal, back-of-house zones).
-
10m (~33ft) is a good “high confidence” option when you want maximum stability, or when the site is especially challenging (lots of partitions, tall/packed shelving, mixed materials, or you’re aiming for tighter location behavior).
A simple way to pick: start at 15m, and only tighten to 12m or 10m when the building layout—or your accuracy goals—demand it.
Step 1 — Draw the AP hexagon
Let’s start with a simple reference size that’s easy to remember and easy to draw.
Create a square that’s about 17 meters (56 feet) tall. Inside that square, draw a regular hexagon centered vertically so the hexagon touches the top and bottom of the square. On the left and right sides, you’ll naturally get a small equal margin (roughly a little over 1 meter / 3–4 feet on each side). Don’t overthink the math—this is a planning tool, not a lab instrument.
This hexagon will be our “AP cell.”
Step 2) Create the first AP triangle
Duplicate the hexagon twice and slide the three hexagons together until they “lock” into place. There’s essentially one clean way to do it—when you have it right, the hexagons touch naturally without gaps or overlaps.
At that point, the centers of the three hexagons form an equilateral triangle with a side length of about 15 meters (50 feet). This triangle is exactly what you want: it represents three APs that will work together to locate devices in the space between them.
Step 3 — Add a practical wall buffer
Design walls margins to ensure we don’t put AP too close from walls and to prevent signal bleeding outside of the store.
Typically draw a safe area of 4-6m (13-20ft) around the store walls.
Step 4 — Drop the hex grid onto the map
Take your 3-hex starter pattern and place it on the floor plan where accuracy matters most. For some stores that’s near the entrance and main aisles; for others it might be the highest-traffic departments or the zones where you’ll use geolocation features (navigation, analytics, proximity messaging, etc.).
Once the anchor point is chosen, everything becomes mechanical: you keep snapping hexagons next to each other like tiles.
Step 5 — Fill the room with the honeycomb
Keep assembling hexagons outward, filling the usable area while staying out of the buffer zone. As you build, you’ll notice something powerful: the layout almost designs itself. You’re no longer guessing where the next AP should go—the pattern makes the decision for you.
When you reach irregular spaces (L-shapes, alcoves, backrooms), keep the pattern as consistent as possible and only bend it when you truly need to.
Step 6 — Optimize to remove weak pockets
After filling the space, step back and look for “thin areas”—zones where the honeycomb leaves a pocket with fewer nearby APs than the rest of the store.
A simple optimization trick is to swap the row pattern. For example, if your rows look like 3-4-3-4-3 APs, try 4-3-4-3-4. Often, adding only one AP (or simply shifting the pattern) dramatically improves uniformity and reduces the chance of dead zones.
Step 7 — Rescale if the building shape is complex
Some buildings just don’t cooperate: narrow corridors, diagonal walls, dense shelving, stock rooms, metal structures, or multi-zone layouts can make a “15m mesh” feel too loose.
In those cases, it’s perfectly reasonable to tighten the planning distance—moving from about 15m (50ft) down to 12m (40ft) or even 9m (30ft) —so you can fit more APs and keep the overlap consistent.
When you increase density, you’ll typically need more careful RF tuning (channels, transmit power ranges, minimum bitrates) so the network stays stable and doesn’t create interference problems, as mentioned in the Cisco High Density Wi-Fi Deployment Guide, “AutoRF tries to reduce the TX power uniformly for all APs within a network but in complex high density network it is necessary to limit the range and the values for the AP to use. To better support complex environments, minimum and maximum TX power settings can be configured in RF profiles.”
Step 8) Adjustments
Real buildings rarely match a perfect grid, so it’s normal to make small adjustments. The key is to bend the pattern gently—not stretch it until you create weak pockets.
Try to keep the mesh as uniform as possible, especially between neighboring APs. If you’re designing around a 15m (50ft) grid, aim to stay close to that distance most of the time, and treat anything above roughly 18m (60ft) as a “last resort” exception. Once spacing gets too large, overlap drops quickly, and the chances of inconsistent geolocation increases.
When you adjust positions, watch the triangles formed between nearby APs. A healthy mesh creates triangles that look fairly balanced. If you start getting long, skinny triangles, it usually means one side is doing too much work and overlap won’t be as stable. A simple rule of thumb is to keep the longest side no more than about 1.2× the shortest side.
Finally, keep an eye on the building outline—alcoves, L-shapes, backrooms, and long corridors can create “pockets” that sit outside the natural perimeter of your AP layout. Those edge areas are where accuracy tends to taper first, so if a zone can’t reliably “see” enough nearby APs, the fix is usually simple: add one AP, or tighten spacing locally to pull that pocket back into the mesh.
On the left, the triangle has a ratio of 1.13 (17/15=1.13 < 1.2), this is acceptable
On the right, the triangle has a ratio of 2, it is too skinny and this might affect both the precision of geolocation, and the Wi-Fi quality
How Eye-In makes this faster
To speed up the planning phase, our partner Eye-In Network Mapping Software includes a built-in hex-grid pattern so you can draft a location-ready layout in minutes, then refine it with survey validation.
4) Budget considerations
The trade-off is simple: more APs usually means better location stability, but it also increases hardware, installation, cabling, switching capacity, and long-term maintenance.
Let’s take another example of approximately 3,000m2 (32,000sq2)
We designed 2 examples of configurations:
-
A with 13 APs (regular density with 15m between APs)
-
B with 23 APs (high density with 12m between APs)
Layout A might use a more standard spacing and land around a dozen APs. Layout B might tighten the mesh for better geolocation consistency and land closer to twenty-plus APs, but this layout might be almost twice as expensive.
So which one should you choose? The “best” answer depends on your goals. If geolocation accuracy is a core feature (wayfinding, proximity experiences, detailed traffic analytics), the higher-density layout is usually worth it. If location is a secondary feature, a lighter layout may be the right compromise.
What's next
At this point, you have a practical way to design a location-ready AP layout without guesswork. By using a simple hex-grid approach, you can build a mesh that keeps spacing consistent, maintains strong overlap, and reduces the risk of dead zones—especially in the areas where retail analytics matter most. Even before any fine tuning, a clean layout like this dramatically improves the reliability of Wi-Fi geolocation data and makes heatmaps, dwell zones, and customer journeys far more trustworthy.
But a good drawing isn’t the whole story. In real stores, the final accuracy also depends on what you install and how you install it: antenna type, access point model, mounting height, whether APs are hidden above tiles, how shelving blocks line-of-sight, and how precisely the floor plan and AP positions are entered in the Meraki dashboard.
In Part 3 of this series, we’ll focus on the practical deployment side. We’ll cover which hardware models are best suited for Wi-Fi geolocation, why omnidirectional antennas are usually the safest choice, and the installation best practices that make or break location stability—so the mesh you designed on paper performs the same way once it’s live.
Sources: