Satellite Link Budget: ITU-R Models & Monte Carlo

Why Single-Point Link Budgets Fail in the Field

A link budget produces one number: the margin between the C/N0 you receive and the C/N0 the demodulator needs. Positive means the link works, negative means it does not.

Real satellite links do not sit at one point. Rain fades the signal, transmit power drifts with temperature, the antenna points slightly off-axis, and the site's weather is a statistic rather than a constant. A single-point budget says nothing about how much of that the link can absorb.

This walkthrough uses the Satellite Link Budget tool to design a Ku-band VSAT uplink. The first lesson is less glamorous than rain fade: check the clear-sky arithmetic before anything else.

The Reference Design: Ku-Band VSAT Uplink

The terminal uploads 10 Mbit/s to a GEO satellite from Munich (48.14° N, 11.58° E) at 14 GHz.

A 1.2 m dish at 60 % efficiency has 42.7 dBi of gain at 14 GHz, and 4 W is 6.0 dBW. After feed and waveguide loss, 48 dBW is a fair terminal EIRP. The slant range comes from the GEO geometry at 38° elevation, not from the 35,786 km altitude.

Reading the Link Budget Table

The link does not close, and rain has little to do with it. Remove every atmospheric term and the budget is still 4.7 dB short:

This is the most common failure in hand-built budgets: a sign slip or a dropped term that makes a thermally impossible link look comfortable. Check the clear-sky C/N0 against the requirement before spending any effort on propagation.

Rain at 99.5 % in Munich

The tool reads Munich's climate from the ITU-R maps: $R_{0.01}$ = 32.1 mm/h (P.837-7), rain height 3.61 km (P.839-4), station height 0.51 km (P.1511-1). With the P.838-3 coefficients at 14 GHz and the P.618-13 slant-path method:

At the 99.5 % design point, rain costs under 1 dB. It grows steeply beyond that — eight times as much at 99.99 % — which is why the availability target, more than the frequency, sets how much rain margin a Ku-band link needs.

Fixing the Thermal Budget

Two changes each close the link at 99.5 %:

Cutting the rate by 5× buys 7.0 dB. The bigger terminal buys 7 dB of EIRP: 3.5 dB from the larger antenna and 3–4 dB from an 8–10 W BUC.

Checking the Monte Carlo Bands

The Monte Carlo (10,000 trials on the free tier) varies EIRP and G/T (σ = 0.3 dB), adds a pointing loss (exponential, mean 0.2 dB), spreads the rain rate log-normally (σ = 0.5) and adds a fading term (|normal|, σ = 0.4 dB). It is an uncertainty model around the ITU-R budget, not an ITU-R method.

For either fix, the p5 margin is about −1.2 dB while the nominal is +0.9 dB: 5 % of the trials end up 1.2 dB or more below threshold at the 99.5 % point. The median sits below the nominal because pointing and fading only subtract.

Design to the p5 margin, not the nominal, if the link must meet its availability reliably.

What It Takes to Reach 99.9 %

At 99.9 % the rain grows to 2.61 dB and both 99.5 % fixes fall short again (−0.70 dB with 2 Mbit/s, −0.69 dB with the 55 dBW terminal). Two ways through:

Satellite G/T is not something the terminal designer controls, but it varies across a beam by several dB. Check the G/T contour at your site rather than using the beam-peak value.

Site and Elevation

The same 2 Mbit/s design at 99.9 %, moved to other sites:

The climate at the site sets the rain margin; the latitude alone does not.

Raising the elevation angle helps less than it seems. Moving the Munich terminal from 38° to 55° (a satellite further east on the arc) cuts rain at 99.5 % from 0.97 to 0.82 dB, and with the shorter slant range and thinner gas and cloud path the margin improves by about 0.5 dB in all.

Key Design Rules from This Analysis

Close the clear-sky budget first. At Ku-band in central Europe, the thermal budget — EIRP, G/T and data rate — decides whether a 99.5 % link closes. Rain at 99.5 % is about 1 dB. Availability is where rain bites. In Munich, rain at 14 GHz grows from 0.97 dB at 99.5 % to 2.61 dB at 99.9 % and 7.99 dB at 99.99 %. Each extra nine needs a new look at the design. Design to the p5 margin. The nominal margin holds only under average conditions. Use the site, not the zone. Houston and Copenhagen differ by 3.8 dB of rain at 99.9 % for the same terminal.

Run the reference design, then change one thing at a time — the scenario URL keeps each variant.

Updated 2026-09-19. An earlier version of this article reported C/N0 = 80.8 dBHz for this design, 16 dB too high, and attributed the resulting margin to 6.8 dB of rain from a latitude-band climate. Every number above now comes from the corrected tool.

---

Related tools: Link Budget Calculator, EIRP Calculator, Noise Figure Cascade