Short answer: size a satellite dish from the link budget, not the frequency band alone. Calculate the receive gain needed to meet the modem threshold after path losses and availability margin, then convert that gain to diameter with D = (λ/π) × √(G/η), using linear gain G and aperture efficiency η. Select a standard size that meets that requirement and, for a two-way terminal, also passes the uplink budget. The 12 GHz example below needs about 1.3 m before rain margin; it is not a universal Ku-band size. Free-space loss is the starting point, not the complete budget [1].
Table of Contents
What Sets the Dish Size You Need
Start with frequency, satellite EIRP at the site, data rate and modem threshold, receiver noise, availability target, path geometry, and operating constraints. At fixed frequency and efficiency, an extra 3 dB of antenna gain requires about 41% more diameter. Band and data rate alone are not enough to select a dish.
Frequency band
At an assumed 60% aperture efficiency, a 1.2 m dish has a calculated gain of about 41.4 dBi at 12 GHz and 45.8 dBi at 20 GHz. Free-space loss also rises by about 4.4 dB between those frequencies. Higher gain alone therefore does not establish a smaller required dish. Evaluate the actual EIRP, receiver noise and propagation conditions on each link; use Earth-space prediction methods for the site-specific atmospheric effects [2].
Satellite and coverage
Use the operator’s EIRP contour at your coordinates, for the relevant beam and carrier operating conditions, rather than the beam-peak brochure value. A 3 dB reduction in available EIRP requires 3 dB more receive gain if everything else stays fixed. That correction can change the selected diameter even though the service and frequency have not changed.
Service and data rate
With the same modulation, coding and required energy per bit relative to noise density, doubling bit rate requires about 3 dB more carrier power if noise density is unchanged. A change of modulation or coding can change that requirement. Use the modem’s specified threshold, bandwidth and implementation allowance rather than assuming every faster service needs the same gain increase.
Availability target
On a 365-day basis, 99.9% availability allows about 8.8 hours of outage per year; 99.5% allows 43.8 hours. These are outage allowances, not predictions of actual downtime. Decide how much of the service outage budget belongs to propagation rather than equipment or network failures, then calculate the atmospheric allowance for that target [2].
Path geometry
For a GEO geometry calculation with a spherical Earth radius of 6,378 km and orbital radius of 42,164 km, the overhead distance is 35,786 km. Slant range is about 38,612 km at 30° elevation, 40,586 km at 10°, and 41,127 km at 5°. Those distances add approximately 0.66, 1.09 and 1.21 dB of free-space loss relative to overhead. Atmospheric effects must be calculated separately; these distance-only penalties do not describe the full low-elevation link [2].
Regulatory limits
A transmit antenna must meet the applicable radiation-pattern requirements as well as its on-axis gain target. For example, U.S. 47 CFR 25.209 specifies earth-station antenna performance envelopes for defined services and bands; the 2025 annual edition documents those requirements. It is not a universal minimum-diameter rule. Confirm the current rules, authorization and operator acceptance requirements for the proposed terminal before ordering [3].
The Sizing Workflow
Define the service requirement
List carrier bandwidth, modulation and coding, the needed carrier-to-noise ratio, and the uptime target (99.5% or 99.9% across the year serve as common examples). The modem specification and the service contract supply them.
Collect satellite, site and receiver data
Record satellite EIRP at the site, uplink and downlink frequencies, polarization, coordinates, altitude, elevation angle, and receiver system noise temperature or noise figure. For a two-way budget, also obtain satellite receive G/T and the terminal amplifier’s usable output power. Frequency and polarization affect the specific rain attenuation coefficients; those coefficients are inputs to a path prediction, not the complete rain fade by themselves [4].
Calculate path and hardware losses
Work from free-space loss first [1]:
FSPL = 20log10(4πd/λ)
where d is the path distance and λ is the wavelength in the same units. With distance in km and frequency in GHz the working form reads:
FSPL = 20log10(d) + 20log10(f) + 92.45 (dB)
Add the remaining terms from models and measured hardware data:
- Rain attenuation and relevant scintillation effects, combined according to the applicable prediction method rather than by adding unrelated worst cases [2].
- Gaseous absorption and cloud attenuation [5] [6].
- Feed and waveguide losses
- Pointing and polarization mismatch losses
Record the value, source and reference plane of each loss. Do not subtract a feed loss separately if it is already included in the antenna gain or receiver specification being used.
Find the required antenna gain
Received power = EIRP − path losses + receive antenna gain
Use dBW for both power terms, dB for losses and dBi for gain, all at consistent reference planes. Rearranging gives:
Required gain (dBi) = required received carrier power (dBW) − EIRP (dBW) + total losses (dB) + any separately allocated reserve (dB).
If the modem specifies C/N rather than received power, first calculate receiver noise over the stated noise bandwidth. Alternatively, use a consistent G/T-based budget, where G/T is antenna gain relative to system noise temperature. Include the availability-related attenuation once, and keep any additional implementation or aging reserve explicit. A larger dish does not remedy every source of service outage.
Convert gain to diameter
A parabolic reflector obeys:
G = η × (πD/λ)² (linear), or in decibels: GdBi = 10log10(G)
Here D and λ use the same length units. For this circular-aperture sizing model, η is aperture efficiency; 0.60 is an assumption used in the calculations below, not a guaranteed product specification. Obtain the antenna’s actual gain or efficiency at the operating frequency. Other aperture shapes and scanned arrays need their own effective-area and scan-loss treatment.
D = (λ/π) × √(G/η)
If the budget result is in dBi, convert it first: G = 10^(GdBi/10). Substituting a dBi number directly into the square root produces the wrong diameter.
Select a standard size and verify it
Select the next available size that meets the fully margined requirement, then repeat the budget with its specified gain. Product families such as 1.2 m quick-deploy terminals and larger earth-station antennas serve different installation needs; diameter alone does not establish compatibility. Check the relevant band and configuration within the SATCOM antenna range, then verify:
- Does the transmit side still meet its EIRP with your available amplifier?
- Do the mount and foundation suit the wind load?
- Can your pointing accuracy hold a beam that narrows as the dish grows?
A Ku-Band Worked Example
This is a calculation example with stated assumptions, not a measurement or a guarantee for any site.
Consider a GEO downlink at 12 GHz directly beneath the satellite. Distance equals the standard 35,786 km and λ stands near 0.025 m. Free-space loss becomes:
FSPL = 20log10(35,786) + 20log10(12) + 92.45 ≈ 91.08 + 21.58 + 92.45 ≈ 205.1 dB [1]
Assume 50 dBW satellite EIRP at the site, a required carrier power of −115 dBW at a defined receiver input, and 2 dB of combined clear-sky atmospheric and hardware losses to that input. The carrier threshold is an assumed input, not a modem specification. The 2 dB excludes rain attenuation and additional reserve. The clear-sky requirement becomes:
Required gain = −115 − 50 + 205.1 + 2 ≈ 42.1 dBi.
The linear gain is about 16,200. At 60% efficiency, D = (0.025/π) × √(16,200/0.60) ≈ 1.3 m before rain margin.
What changes when margin is added?
Suppose the same illustrative budget requires another 3 dB in total for rain attenuation and reserve. This is an assumed sensitivity check, not a weather prediction. The gain target becomes about 45.1 dBi and the calculated diameter about 1.85 m. Under the same efficiency assumption, a 1.8 m dish gives 44.9 dBi and falls short; a 2.4 m dish gives 47.4 dBi and exceeds the target by about 2.3 dB. This is why rounding the clear-sky result to 1.8 m is not enough to approve the link.
For an actual purchase, replace the assumed 3 dB with the site’s propagation result and allocated reserve, use the candidate antenna’s specified gain, and check the uplink separately. Keep the EIRP contour, modem threshold, loss allocation and antenna specification with the calculation so another engineer can reproduce the selection.
How Band Choice Changes the Size You Need
Read the table below as a sizing warning: it assumes the standard 35,786 km GEO distance with 60% efficiency [1]:
| Band (example frequency) | Free-space loss | 1.2 m dish gain |
|---|---|---|
| C (4 GHz) | ~195.6 dB | ~31.8 dBi |
| X (8 GHz) | ~201.6 dB | ~37.8 dBi |
| Ku downlink (12 GHz) | ~205.1 dB | ~41.4 dBi |
| Ku uplink (14 GHz) | ~206.5 dB | ~42.7 dBi |
| Ka downlink (20 GHz) | ~209.5 dB | ~45.8 dBi |
| Ka uplink (30 GHz) | ~213.1 dB | ~49.3 dBi |
For fixed satellite EIRP, distance, receive diameter and efficiency, the frequency-dependent gain increase cancels the free-space-loss increase in this simplified comparison. The table does not prove that Ka-band always needs a larger dish or that the highest frequency sets the size. Actual beam EIRP, noise, service threshold, atmospheric attenuation and transmit constraints decide which budget is limiting [2].
What Size Matches Your Gain Target at 12 GHz
These are calculated gains at 12 GHz and 60% efficiency, not measured product data. Choose a row at or above your fully margined gain target, then verify the candidate’s specified performance. The gain difference is relative to a 1.2 m dish, not a site-specific rain allowance.
| Dish diameter | Calculated gain (dBi) | Gain relative to 1.2 m |
|---|---|---|
| 0.6 m | ~35.3 | −6.0 dB |
| 0.9 m | ~38.9 | −2.5 dB |
| 1.2 m | ~41.4 | 0.0 dB |
| 1.8 m | ~44.9 | +3.5 dB |
| 2.4 m | ~47.4 | +6.0 dB |
| 3.7 m | ~51.1 | +9.8 dB |
Doubling diameter adds about 6 dB at unchanged efficiency. That is a gain increase, not a guarantee of a particular availability level.
Sizing Checks Buyers Often Skip
A rough full half-power beamwidth estimate is 70λ/D degrees; the coefficient depends on illumination. At 12 GHz it gives about 1.5° for a 1.2 m dish and 0.7° for a 2.4 m dish. This is beamwidth, not an allowable pointing error. Compare the actual pattern with installation accuracy, wind deflection and any tracking requirement.
The feed horn affects illumination and aperture efficiency, while losses in the waveguide chain depend on the hardware and reference plane. Check those specifications separately. For custom antenna feed systems, confirm feed-to-reflector compatibility rather than assuming a matching change will recover a fixed number of decibels.
C-band smoothness does not guarantee Ka-band service. Small surface errors scatter power from the main beam at short wavelengths. Rate the antenna by its specified band as well as diameter, or the dish underperforms its nameplate size.
On a two-way link, check usable amplifier power, transmit feeder loss and antenna gain together against the uplink requirement and applicable limits. If future capacity matters, budget that future service explicitly; an arbitrary extra size is not a substitute for calculating the requirement.
The Short Answer to Keep
Choose the smallest standard dish that meets the fully margined receive and transmit budgets, required antenna patterns, and installation constraints. Keep the assumptions with the result: without the site EIRP, receiver threshold, propagation allowance and actual antenna gain, a diameter is only a preliminary estimate.
Sizing a link now? Send your band, site, and uptime target.
[1] ITU-R P.525, Calculation of free-space attenuation (P.525-5 in force).
[2] ITU-R P.618, Propagation data and prediction methods for Earth-space systems (P.618-14 in force).
[3] U.S. Government Publishing Office, 47 CFR 25.209, Earth station antenna performance standards, 2025 annual edition.
[4] ITU-R P.838, Specific attenuation model for rain (P.838-3).
[5] ITU-R P.676, Attenuation by atmospheric gases and related effects (P.676-13).
[6] ITU-R P.840, Attenuation due to clouds and fog (P.840-9).
