Microwave Link Budget Calculator
A terrestrial hop planned the way an operator asks the question: not “how much rain fade at 0.01 %”, but given this fade margin, how many minutes a year is the link down. Rain is modelled with ITU-R P.838-3 and P.530, not a rule of thumb.
Radios and antennas
Path profile
How the rain model works
Rain is the thing that decides whether a hop above about 10 GHz is a link or a liability, and it is where most online calculators stop at a hand-waving figure. This one follows the two recommendations that actually govern it.
ITU-R P.838-3 — specific attenuation
Attenuation per kilometre follows a power law in rain rate:
The coefficients k and α depend on frequency and polarisation. Rather than interpolating between a handful of tabulated points, this tool evaluates the log-logistic fitting functions published in the recommendation, which are defined continuously from 1 to 1000 GHz. Reproducing the recommendation's own tabulated values from those functions agrees to within 0.03 %, which is the rounding in the printed table.
Polarisation is handled properly, through the tilt relation in the recommendation rather than by picking one column:
Horizontal polarisation fades worse than vertical, because raindrops flatten as they fall and present a wider target to a horizontally polarised wave. At 18 GHz in 50 mm/h that difference is around 20 %, which is free margin for choosing vertical when nothing else constrains you.
ITU-R P.530 — path length and time percentage
A rain cell is a few kilometres across, so a long hop is never uniformly wet. P.530 handles this with an effective path length that is shorter than the geometric one, and the discount grows with distance:
That gives the attenuation exceeded 0.01 % of the time. Any other percentage follows from:
This tool inverts that relation. You are not usually asking what the fade is at some percentage — you have a radio with a fixed fade margin and you want to know what availability it buys. So the margin goes in and the percentage of the year it is exceeded comes out, reported in minutes per year because that is the number that ends up in a service level agreement.
What rain rate to use
R0.01 is the rain rate exceeded 0.01 % of an average year at your location, in mm/h. It is a local climate figure, not a guess, and the authoritative source is the ITU-R P.837 rain map. Rough orientation while you look it up:
| Climate | R0.01 | Examples |
|---|---|---|
| Dry continental / desert | 8–22 mm/h | Interior Spain, US Southwest |
| Temperate maritime | 22–32 mm/h | UK, northern Europe, US Pacific Northwest |
| Temperate continental | 30–42 mm/h | Central Europe, US Midwest |
| Mediterranean / subtropical | 40–60 mm/h | Coastal Spain and Italy, US Southeast |
| Tropical | 60–145 mm/h | Southeast Asia, Central Africa, Brazil |
These are orientation bands, not design figures. Use the P.837 value for your coordinates: the difference between 30 and 60 mm/h can halve the usable hop length at 23 GHz.
Path geometry
A link budget that closes on paper still fails if the path is obstructed. Two effects move the terrain relative to the beam.
The earth bulges between the sites, and refraction in the atmosphere modifies how much. The effective earth radius factor k captures this: 4/3 is the standard atmosphere, while sub-refractive conditions flatten the beam toward the ground and are the adverse case worth designing against.
The first Fresnel zone is the ellipsoid around the line of sight that carries most of the energy. An obstacle intruding into it costs you signal even without blocking the direct ray. The accepted rule is to keep at least 60 % of the first zone clear:
Note which way frequency runs here. A lower frequency has a fatter Fresnel zone and so needs more clearance, which catches people who assume the low bands are always the forgiving choice.
Limitations
This covers the two fading mechanisms that dominate terrestrial hop design: rain and path obstruction. It does not model multipath and ducting (P.530 has a separate method, and on long over-water or over-flat-terrain paths that mechanism can dominate at low frequencies), diffraction loss over a partially blocked path, cross-polar discrimination degradation in rain, interference from other services, or adaptive modulation trading throughput for margin. Gaseous absorption is an approximation anchored to sea level at 15 °C and 7.5 g/m³; near the 22 GHz water vapour line or the 60 GHz oxygen line, use ITU-R P.676 for your real atmosphere.
Path profiles here take a single worst obstacle. A real profile needs terrain data along the whole route, and a survey beats any of it.