Outdoor coverage map
Put a site on the map, give it the radio parameters, and see the coverage it produces - radius, area and a heat map you can take away.
Site
Drag on the map to move the site.
Radio parameters
Model
Design target
Map data (c) OpenStreetMap contributors. Tiles are loaded from the public OSM server, so they may be slow or unavailable on some networks; the numbers and the exported files do not depend on them.
How this is calculated
Okumura-Hata: L = 69.55 + 26.16log(f) - 13.82log(hb) - a(hm) + (44.9 - 6.55log(hb))log(d), with a(hm) = (1.1log(f) - 0.7)hm - (1.56log(f) - 0.8) for a small or medium city and 3.2(log(11.75hm))^2 - 4.97 for a large city; suburban subtracts 2(log(f/28))^2 + 5.4 and open rural subtracts 4.78(logf)^2 - 18.33logf + 40.94. COST-231 Hata: L = 46.3 + 33.9log(f) - 13.82log(hb) - a(hm) + (44.9 - 6.55log(hb))log(d) + Cm, with Cm 0 dB for a medium city or suburban and 3 dB for a dense metropolitan centre. The 3GPP models follow TR 36.814: UMa LOS is 28 + 22log(d3D) + 20log(fc) below the break point and 28 + 40log(d3D) + 20log(fc) - 9log(dBP^2 + (hBS-hUT)^2) above it, with NLOS max(PL_LOS, 13.54 + 39.08log(d3D) + 20log(fc) - 0.6(hUT-1.5)); UMi is 32.4 + 21log(d3D) + 20log(fc) LOS and 22.4 + 35.3log(d3D) + 21.3log(fc) - 0.3(hUT-1.5) NLOS; RMa is 10.5 + 40log(d3D) + 20log(fc) - 0.6(hUT-1.5). Distances are three-dimensional in the 3GPP models, which matters at the higher bands. The cell radius is found by bisection against the target level plus the shadow fading margin, where the margin is the normal deviate for your confidence level times the standard deviation (8 dB for Hata, 4 dB for UMa LOS, 6 dB for NLOS, 7.8 dB for UMi LOS). These are empirical models with no terrain or clutter input: use them for option comparison and as a starting point, then confirm with a survey and measurement.
Common questions
How do I calculate the coverage radius of a base station?
Work backwards from the edge: the level at the edge has to be the target level minus the shadow fading margin for your confidence level, and the path loss from the site to that distance has to fit inside the budget of transmit power plus antenna gain. This tool solves that by bisection, so it works with every model it offers. The margin is what separates a 90% design from a 95% one, and it is usually several dB and a good fraction of the cell radius.
Which propagation model should I use?
Okumura-Hata below 1500 MHz, COST-231 Hata between 1500 and 2000 MHz, and the 3GPP models above 2 GHz. Each has a stated validity range, and using one outside it gives a number that looks precise and is not. If your project is at 3.5 GHz in a dense city, use UMa; the Hata models were fitted to 2 GHz data and are not extrapolated further.
Why does the cell radius change so much with the confidence level?
Because the shadow fading margin is a normal deviate multiplied by the standard deviation, and the path loss grows with distance. Raising the confidence from 90% to 99% adds about 8 dB of margin, which on a typical log-distance exponent translates into roughly a 20% smaller radius. That is often the difference between one site and two on the same street.
Does the coverage map account for terrain?
No. The models here are statistical: they capture clutter and building density as averages, and they have no idea whether a hill is in the way. For a site with terrain in the picture, run the link budget against a real terrain profile - the Longley-Rice model or the Fresnel zone tool here - and treat the map as the statistical baseline.
Can I load the coverage grid into my own GIS?
Yes. The grid export gives longitude, latitude, distance, received level and a covered flag for every cell, which imports straight into QGIS or any mapping tool. The heat map can also be saved as a PNG for a report.