Planner › Methodology: how Signal Plan predicts Wi-Fi coverage
Methodology: how Signal Plan predicts Wi-Fi coverage
Last updated: 9 October 2026.
Signal Plan produces a preliminary predictive design: a forecast of signal strength across a floor and a suggested set of access points. This page explains the method so you can judge how far to trust it.
1. Inputs
Floor plan image and scale; design band (2.4, 5 or 6 GHz); minimum RSSI; office environment (path-loss exponent 1.9 for open office, 2.25 typical, 2.7 dense partitions, 3.1 high attenuation); mounting height; concurrent clients and Mbps per client; channel width; walls with a loss in dB; and the brand or the models you own.
2. Signal model
Predicted RSSI = access point EIRP − free-space loss at 1 m − 10 × exponent × log10(distance) − wall losses − link margin. The 1 m loss is 40.2, 47.1 and 48.7 dB at 2.4, 5 and 6 GHz. Distance includes the height difference between the ceiling-mounted access point and a client at 1.2 m. Wall loss is the dB value you pick per wall, multiplied by 1.0, 1.2 and 1.35 for 2.4, 5 and 6 GHz because higher bands lose more. The link margin is 8 dB for client antennas, body loss and fading. The model is downlink only and uses straight-line wall crossings.
3. Access point power
A client transmits at around 15 dBm, so an access point set much higher hears clients worse than they hear it. The planner sets each model's EIRP to 15 dBm plus its antenna gain, less 2 dB because datasheet figures are peak values and the average gain towards clients is lower. Where a vendor publishes the average pattern gain (for example for some Aruba models), that value is used instead of the 2 dB rule. At 6 GHz the result is capped at the indoor low-power limit (5 dBm per MHz, 30 dBm maximum).
4. Coverage, placement and capacity
Coverage is the share of a sampled grid of floor points at or above your minimum RSSI. Placement uses a greedy search: each access point is put where it adds the most weighted coverage, with a spacing penalty so units do not cluster. In automatic mode access points are added until about 95% of the sampled points meet the target. Capacity and throughput per access point are estimated from its streams, channel width and Wi-Fi generation, and the access point count is raised if either is short. These capacity figures are simplified planning values, not vendor guarantees.
5. Multiple floors (Pro)
For access points on other floors the path uses the three-dimensional distance, counts walls on the access point's floor up to the point where the path crosses the slab and walls on the receiving floor after it, and adds slab loss. Defaults are 10 dB (timber or light), 20 dB (concrete) and 26 dB (heavy reinforced) for the first slab at 2.4 GHz, scaled by band, with each extra slab adding 60% of the first because loss does not grow linearly with floor count. Slab loss is the least certain input in any prediction; calibrate it with a measurement.
6. Data quality
Access point specifications come from manufacturer datasheets and spec pages where available. Review log: 8 and 9 October 2026, 150 models reviewed across 20 brands, with corrections to streams, gains, bands and Wi-Fi generation for the models where sources disagreed with our earlier family-level estimates. Coverage by brand at 9 October 2026:
| Brand | Models | Vendor datasheet or spec page | Retailer listing quoting vendor | Family estimate |
|---|---|---|---|---|
| Cisco Meraki | 25 | 12 | 3 | 10 |
| Cisco Catalyst | 5 | 4 | 0 | 1 |
| HPE Aruba Networking | 12 | 1 | 6 | 5 |
| Aruba Instant On | 6 | 0 | 3 | 3 |
| Ubiquiti UniFi | 15 | 10 | 2 | 3 |
| Ruckus | 8 | 7 | 0 | 1 |
| Juniper Mist | 9 | 3 | 1 | 5 |
| Extreme Networks | 7 | 0 | 2 | 5 |
| TP-Link Omada | 11 | 5 | 1 | 5 |
| Fortinet FortiAP | 8 | 2 | 0 | 6 |
| Huawei AirEngine | 4 | 0 | 0 | 4 |
| Zyxel | 9 | 3 | 0 | 6 |
| NETGEAR Insight | 5 | 0 | 0 | 5 |
| Cambium Networks | 5 | 1 | 1 | 3 |
| WatchGuard | 4 | 1 | 2 | 1 |
| Ruijie Reyee | 4 | 1 | 0 | 3 |
| Grandstream | 3 | 0 | 3 | 0 |
| MikroTik | 2 | 0 | 1 | 1 |
| D-Link Nuclias | 2 | 0 | 1 | 1 |
| Generic | 6 | 0 | 0 | 6 |
7. Limits
- Real buildings differ: furniture, people, glass coatings and metal change attenuation.
- Interference from neighbouring networks is not modelled.
- Wall and slab values are planning assumptions.
- Regulatory power limits vary by country; check yours.
Validate every design with an access-point-on-a-stick survey and again after installation. See the RSSI guide and the AP placement guide. Questions or corrections: [email protected].
Open the free Wi-Fi heatmap planner
← Back to the planner · All brands · AP placement guide · RSSI guide · About · Methodology
