Communications
Satellite Comms Link Budget
Whether your downlink closes — EIRP, path loss, ITU-R P.618 rain fade, C/N0, achieved Eb/N0, and the link margin, with the full gains-and-losses table, for any band from S to V.
// Friis + receiver G/T + Boltzmann noise, with an ITU-R P.618 simplified rain term (P.838 k/α) and P.676 gaseous absorption. A first-order budget for antenna, power, and frequency trades — run the full ITU-R P.618 method for a licensing-grade availability figure.
This satellite link budget calculator runs the whole accounting sheet in the browser: transmit effective isotropic radiated power (EIRP), free-space path loss (FSPL), rain attenuation from the ITU-R P.618 simplified method with P.838 k/α coefficients, P.676 gaseous absorption, receiver gain-to-noise-temperature (G/T), and Boltzmann noise — down to the carrier-to-noise-density ratio (C/N₀), the achieved energy-per-bit (Eb/N₀), and the link margin. Antennas can be entered as dish diameters or as gain / G/T figures directly, and range as altitude-plus-elevation geometry or a straight slant distance. A first-order budget for antenna, power, and frequency trades — verify survivors with a full ITU-R P.618 availability run and a modem-vendor threshold.
How the link equation works & what it omits
A link budget is the accounting sheet of a radio link: it adds up every gain and subtracts every loss between the transmitter's power amplifier and the receiver's demodulator, and asks whether what arrives is strong enough to carry the data. The bottom line is the link margin — the decibels of headroom the achieved signal has above the minimum the modem needs. A positive margin means the link closes; a negative one means it does not.
The link starts with EIRP — the transmit power (in dBW) plus the transmit-antenna gain, less feed loss. A parabolic dish gain is G = η·(πD/λ)², so gain climbs with the square of both diameter and frequency (a 3 m Ka-band dish is over 50 dBi). From EIRP the signal loses free-space path loss, FSPL = 20·log₁₀(4πR/λ) — the dominant term, ~180 dB for a LEO pass and ~205 dB from GEO — plus atmospheric, rain, pointing, and polarization losses. The receiver's sensitivity is captured by G/T, its antenna gain divided by system noise temperature. Putting it together gives the carrier-to-noise-density ratio C/N₀ = EIRP − losses + G/T − k, where k = −228.6 dBW/K/Hz is Boltzmann's constant. Dividing out the bit rate, Eb/N₀ = C/N₀ − 10·log₁₀(R_b), gives the achieved energy-per-bit, and the margin is that minus the required Eb/N₀ and the modem implementation loss.
The term that makes or breaks a high-band link is rain attenuation. This tool uses the ITU-R P.618 simplified method: the specific attenuation γ_R = k·R^α in dB/km, with the frequency-dependent regression coefficients k and α from ITU-R P.838, is multiplied by the effective slant path through the rain cell (the geometric slant path scaled by a path-reduction factor that accounts for a storm cell being smaller than the beam's footprint). Below about 10 GHz — S, C, and X band — rain fade is a fraction of a decibel and links are effectively rain-immune. Above it — Ku, Ka, Q, V — γ_R climbs steeply, and at Ka (20–30 GHz) a temperate 0.01% rain rate of ~40–50 mm/hr produces a fade of tens of decibels. That is why Ka systems live or die on their rain-availability allocation and lean on adaptive coding and modulation and ground-station site diversity. A smaller gaseous absorption term (oxygen + water vapour, ITU-R P.676, peaking near the 22 GHz water line) is added on top.
What this tool does not do: it is a first-order budget, not a licensing-grade link analysis. The rain and gaseous terms are the ITU-R simplified parametrisations, not a site-specific P.618 availability run against measured P.837 rain statistics; there is no scintillation, cloud, or melting-layer term, no interference/adjacent-satellite budget, no polarization-diversity or XPD treatment, and no specific-MODCOD threshold table — you supply the required Eb/N₀ for your coding. Use the numbers to size antennas, power, and frequency and to see how much rain a link can take, then verify the survivors with a full ITU-R P.618 run and a modem-vendor threshold.
// pick a link, then dial frequency, power, antennas, and the rain rate.
Carrier & transmitter
// frequency sets the band; power + gain set the EIRP.
Path & receiver
// range + rain set the losses; G/T sets sensitivity.
Modulation & weather
// data rate + Eb/N0 set the demand; rain sets the fade.
Link margin
+16.9 dB
LINK CLOSES
Ka-band · 1,000 Mbps · 16.9 dB of headroom above the required Eb/N0.
Link budget
// Ka-band · λ = 15.0 mm · slant 1,123 km
39.7
EIRP (dBW)
179.5
FSPL (dB)
0.0
rain fade (dB)
27.7
G/T (dB/K)
115.4
C/N0 (dB-Hz)
+16.9 dB
margin
// gains and losses, dB — sum to the margin
| Tx power | +6.99 |
| Tx antenna gain | +33.75 |
| Tx line loss | -1.00 |
| Free-space path loss | -179.48 |
| Rain attenuation (P.618) | -0.00 |
| Gaseous absorption (P.676) | -0.36 |
| Pointing loss | -0.50 |
| Polarization loss | -0.30 |
| Rx antenna gain | +53.75 |
| Noise density (−10log Ts − k) | +202.58 |
| Data rate (−10log Rb) | -90.00 |
| Required Eb/N0 | -6.50 |
| Implementation loss | -2.00 |
| Link margin | +16.93 |
// received C = -87.1 dBW · achieved Eb/N0 = 25.4 dB · required 6.5 dB
// first-order budget
Friis + receiver G/T + Boltzmann noise, with ITU-R P.618 simplified rain attenuation (P.838 k/α) and P.676 simplified gaseous absorption. A screening budget for feasibility and antenna/power trades — for a licensing-grade availability figure, run the full ITU-R P.618 method against your site's P.837 rain statistics.
// shareable URL encodes every input. no backend.
// link budget, term by term
// ai-generated breakdown of what these numbers mean — with diagrams.
Common questions
What is a good link margin for a satellite downlink?
A common screening bar is a few decibels of positive margin after every loss is booked — but the right number depends on what you have not modelled. The default Ka-band low-Earth-orbit (LEO) preset closes with ~17 dB of clear-sky margin, and the same link in a 42 mm/hr storm cell sits at −3.3 dB: at high bands the margin question is really a rain-availability question, which is why Ka systems budget against a percentage-of-time fade statistic rather than a single clear-sky figure.
How much rain fade should I budget at Ka-band?
The heavy-rain preset — 20 GHz through a 42 mm/hr cell, a temperate-climate 0.01%-of-the-year rain rate — adds about 20 dB of attenuation. The same storm at X-band (8 GHz) costs only ~3.8 dB, and below roughly 10 GHz links are effectively rain-immune. That single frequency dependence explains a lot of system architecture: S/C/X links survive on modest margins, while Ka and above lean on adaptive coding and modulation and ground-station site diversity.
What is the difference between C/N₀ and Eb/N₀?
C/N₀ is the carrier power over the noise density — a property of the radio link alone, independent of what you send over it. Eb/N₀ divides that carrier among the bits: Eb/N₀ = C/N₀ − 10·log₁₀(R_b). The default link's ~115.4 dB-Hz of C/N₀ carrying 1 Gbit/s leaves 25.4 dB per bit; halving the data rate buys 3 dB — usually the cheapest knob when a link will not close. Then feed the rate that closes into the spacecraft data budget calculator to see whether it clears your daily generation.
Can a small satellite really downlink at 1 Gbit/s?
Yes — the default preset does it with 5 W of transmit power through a 0.3 m dish at 20 GHz, into a 3 m ground station at 25° elevation, closing 1 Gbit/s with ~17 dB of clear-sky margin. High frequency is what makes small apertures potent: gain scales as (πD/λ)², so even the 0.3 m spacecraft dish is worth ~33.8 dBi at Ka-band. The cost is the rain sensitivity above — clear-sky Ka is generous, rained-on Ka is not.
How does a GEO link budget differ from LEO?
Range, mostly: geostationary orbit (GEO) at ~38,000 km costs about 26 dB more free-space path loss than the ~1,100 km LEO slant path (205.6 vs 179.5 dB in the presets). The Ku GEO preset pays for it with 100 W of transmit power and a 1.5 m spacecraft dish, and still closes 30 Mbit/s into a 1.2 m ground terminal with ~14.7 dB of margin — even carrying ~8 dB of Ku-band rain. GEO links buy their headroom with power and aperture; LEO links get theirs from proximity, and spend it on data rate.
How do I run a CubeSat link budget?
A CubeSat link budget is the same accounting sheet with small-satellite numbers. Set the band to S (~2.2–2.4 GHz) or UHF, a transmit power of 1–2 W, and a low-gain patch or deployable antenna (~6–8 dBi) into a modest ground station — a 3–4 m dish, or a UHF Yagi array. Because most CubeSats work below ~10 GHz, rain fade is negligible, so the budget is dominated by free-space path loss over the slant range and the ground station's G/T. The two knobs that actually close a marginal pass are the downlink data rate and the ground-antenna size: halving the rate buys 3 dB, and a 3U typically closes a few hundred kbit/s to low Mbit/s at S-band. Enter your own numbers above to see where a 3U downlink closes, then feed the rate into the onboard data budget calculator to check it clears a full orbit of generated data.
References
- // ITU-R P.618-13 — Propagation data and prediction methods required for the design of Earth-space telecommunication systems (rain-attenuation method).
- // ITU-R P.838-3 — Specific attenuation model for rain for use in prediction methods (the k / α regression coefficients).
- // ITU-R P.676-12 — Attenuation by atmospheric gases and related effects (oxygen + water-vapour absorption).
- // ITU-R P.837-7 / P.839-4 — rain-rate statistics and rain-height reference for the slant path.
- // Wertz, Everett & Puschell (eds.), Space Mission Engineering: The New SMAD (2011), ch. 13 (Communications).
- // Maral & Bousquet, Satellite Communications Systems, 5th ed. (2009) — the C/N₀ link equation and rain-fade treatment.
