If you have ever sat through a satellite project post-mortem, you know the drill. Someone points to the link budget spreadsheet, shrugs, and mutters about unexpected losses. The link budget looked fine on paperâevery cell populated, every formula intactâbut the system still missed its availability target. That scenario plays out more often than most engineers care to admit.
A link budget is a quantitative accounting of every gain and loss between a transmitter and a receiver. It is not a decorative exercise or a box to tick before moving on to more interesting work. It is the document that determines whether your satellite link will close under real-world conditions or quietly fail the first time a rain cell drifts into the path. Getting it right demands more than plugging numbers into a templateâit demands understanding what those numbers represent and where they come from.
The Fundamental Equation
Every satellite link budget reduces to one core idea: the received carrier-to-noise density ratio (C/Nâ) must exceed the demodulator threshold by a comfortable margin. You can express this relationship in decibels:
C/Nâ = EIRP â Lpath â Latmospheric + G/T â k
Where:
- EIRP is Effective Isotropic Radiated Power from the transmitting earth station or satellite
- Lpath is free-space path loss
- Latmospheric captures rain, gaseous absorption, and scintillation
- G/T is the receiver figure of merit (antenna gain over system noise temperature)
- k is Boltzmann’s constant (â228.6 dBW/K/Hz)
This looks simple enough. The difficulty lies in selecting realistic values for each term and accounting for every loss mechanism that matters at your frequency and availability target.
Starting at the Transmitter
EIRP: More Than Just Power
EIRP combines transmit power and antenna gain into a single figure. If your high-power amplifier delivers 20 dBW and your antenna provides 42 dBi gain, your peak EIRP is 62 dBW. But that is the on-axis value. Off-axis EIRP density must comply with ITU-R S.580 and relevant coordination agreements. Ignoring this constraint is a fast way to incur regulatory headaches.
Also account for waveguide losses between the amplifier output and the antenna feed. A 0.5 dB loss here steals roughly 11% of your transmit power before it ever leaves the dish. In higher-frequency systems (Ka-band and above), waveguide loss can exceed 1 dB, making this a non-trivial term.
Power Back-Off
Operating a traveling-wave tube amplifier (TWTA) or solid-state power amplifier (SSPA) at saturation introduces intermodulation distortion in multi-carrier configurations. You must back off the operating point to keep distortion products below acceptable limits. Typical output back-off values range from 3 dB to 7 dB depending on the amplifier technology and modulation scheme. This back-off directly reduces available EIRP and must appear in the budget as a loss item.
Path Loss and Atmospheric Effects
Free-Space Path Loss
Free-space path loss follows the inverse-square law and scales with frequency. At 12 GHz (Ku-band) over a geostationary orbit slant range of 38,000 km, the path loss is approximately 205 dB. At 30 GHz (Ka-band), it climbs to roughly 213 dB. These are large numbers, and even small modeling errorsâsuch as using a nominal geostationary altitude instead of the actual slant range for your elevation angleâcan shift the result by a decibel or more.
Rain Attenuation
Rain fade is the single largest source of uncertainty in Ku-band and Ka-band link budgets. The ITU-R P.618 propagation model provides statistical rain attenuation values based on climate zone, frequency, elevation angle, and availability target. For a 99.99% availability link at Ka-band in a temperate climate, rain fade can exceed 15 dB. That figure alone can determine whether your link is viable.
Be honest about your availability requirement. A 99.9% availability budget (roughly 8.8 hours outage per year) will show far less rain margin than a 99.99% budget (roughly 53 minutes). Choosing the wrong target early on propagates through every downstream calculation.
Other Atmospheric Losses
Gaseous absorption (oxygen and water vapor), tropospheric scintillation, and cloud attenuation all contribute. Individually they may look smallâa few tenths of a decibelâbut they add up. At low elevation angles (below 10°), atmospheric absorption increases significantly because the signal traverses more of the atmosphere. If your link operates near the horizon, model these terms carefully.
The Receiver Side: G/T and Noise
Antenna Gain
The receiving antenna gain depends on diameter, efficiency, and frequency. A 2.4-meter dish at 12 GHz with 65% aperture efficiency yields approximately 46 dBi gain. Before plugging in a textbook efficiency value, verify what your actual antenna delivers. Feed spillover, blockage, and surface tolerances all reduce real efficiency. Measured gain data from the antenna manufacturer is always preferable to theoretical estimates.
System Noise Temperature
System noise temperature (Tsys) combines antenna noise, low-noise block downconverter (LNB) noise, and waveguide losses between the antenna feed and the LNB. A common mistake is to use the LNB noise figure alone and ignore antenna noise temperature. At C-band, antenna noise from sky and ground pickup may be 25 K or more. At Ka-band, under clear-sky conditions, antenna noise can be as low as 30 K, but rain increases both attenuation and antenna noise simultaneously.
Waveguide and cable losses ahead of the LNB are doubly harmful: they attenuate the signal and add thermal noise proportional to their physical temperature. A 0.5 dB loss at 290 K adds roughly 35 K to the system noise temperature. This is why LNBs are mounted as close to the feed as physically possible.
Link Margin: The Gap Between Theory and Reality
Link margin is the difference between your available C/Nâ and the threshold required by your modulation and coding scheme. A 3 dB margin means the link can tolerate twice the noise or half the signal power before it falls to threshold.
How much margin is enough? That depends on your confidence in the input data and your tolerance for outage. For well-characterized links with measured hardware parameters, 3 dB of clear-sky margin is a common starting point. For less certain scenariosânew frequency bands, untested propagation environments, or hardware with limited test dataâ5 dB or more is prudent.
Document what your margin accounts for. A margin labeled “implementation margin” should cover hardware aging, pointing errors, and polarization mismatch. A margin labeled “propagation margin” should cover rain fade and other atmospheric effects. Mixing these categories into a single unexplained number makes the budget opaque and impossible to audit.
Common Mistakes That Break Budgets
Using peak values instead of worst-case values. Amplifier datasheets quote saturated output power, but your link will operate at a backed-off point. Antenna gain specifications refer to the boresight, while your actual pointing error reduces effective gain.
Ignoring polarization mismatch. Cross-polarization discrimination matters in frequency-reuse systems. If your receive antenna is misaligned by just 1° of polarization rotation, you may lose 20 dB of isolation from the co-polarized signal.
Confusing availability with reliability. Availability is the fraction of time the link meets its performance threshold. Reliability is the probability that the hardware does not fail. These are different metrics. A link can have 99.99% availability and still experience hardware failures that take it offline for days.
Omitting implementation losses. Modem implementation loss accounts for differences between theoretical demodulator performance and actual hardware performance. Typical values range from 0.5 dB to 2 dB depending on the modulation order and code rate. Skipping this term makes your budget optimistic.
A Worked Example
Consider a Ku-band downlink from a geostationary satellite to a 1.8-meter earth station in a temperate climate, targeting 99.7% availability:
- Satellite EIRP: 52 dBW (saturated)
- Output back-off: 3 dB
- Free-space path loss: 205.5 dB
- Rain attenuation (0.3% exceedance): 4.2 dB
- Gaseous absorption: 0.2 dB
- Earth station antenna gain: 44.5 dBi
- System noise temperature: 150 K (21.8 dBK)
- G/T: 22.7 dB/K
- Boltzmann constant: â228.6 dBW/K/Hz
C/Nâ = (52 â 3) â 205.5 â 4.2 â 0.2 + 22.7 â (â228.6) = 90.4 dBHz
If your DVB-S2 demodulator requires 62.5 dBHz for 8PSK 3/4, you have a margin of 27.9 dBâcomfortable, but mostly consumed by rain margin. Under clear-sky conditions, that margin climbs to 32.1 dB, which illustrates why you must distinguish between clear-sky and rain-attenuated budgets.
Document Everything
A link budget is a living document. Every value should carry a source: a datasheet reference, a measurement report, an ITU recommendation, or a stated assumption. When someone reviews your budget six months from now, they should be able to trace each number back to its origin. Spreadsheets without comments are a liability. An unexplained 2 dB entry labeled “misc loss” is an invitation for someone to remove it or misinterpret it.
Version control matters. When hardware specifications change or you update a propagation model, record the revision. A budget that silently shifts between meetings erodes trust and introduces errors.
FAQ
What is the minimum link margin I should design for?
There is no universal minimum. For well-characterized systems with measured hardware data, 3 dB of clear-sky margin is a reasonable baseline. For links at Ka-band or higher, where rain fade can be severe, your rain margin may need to exceed 10 dB depending on the availability target. Always separate clear-sky margin from rain margin so you can evaluate each independently.
Should I use ITU-R models or local measurement data for rain attenuation?
Use local measurement data whenever it is available and statistically significant. ITU-R models like P.618 provide good estimates based on climate zone and frequency, but they are statistical approximations. If you have years of beacon measurements at your site, that data will reflect your specific propagation environment more accurately. When local data is unavailable, ITU-R models are the standard reference.
How do I handle links that cross multiple climate zones?
Satellite links traverse a shared atmospheric path, but the rain rate along that path is not uniform. Use the ITU-R rain height model to calculate the effective path length through the rain layer, and apply the appropriate rain rate for the climate zone at the earth station location. The path is dominated by conditions near the surface, so the climate zone at the earth station typically has the strongest influence. For long, low-elevation paths, consult ITU-R P.618 for guidance on path-reduction factors.