What is the system noise temperature and how do I calculate it from individual component contributions?
System Noise Temperature Engineering
System noise temperature is the preferred noise metric for high-sensitivity systems (satellite communications, radio astronomy, deep-space links) because it provides more precise accounting of noise contributions than noise figure, especially when the antenna sees noise temperatures far from the standard 290K reference.
| Parameter | Free Space | Urban | Indoor |
|---|---|---|---|
| Path Loss Model | Friis (1/r²) | Okumura-Hata | IEEE 802.11 |
| Fading Margin | 0 dB | 10-30 dB | 5-15 dB |
| Multipath | None | Severe | Moderate-severe |
| Typical Range | Line of sight | 1-30 km | 10-100 m |
| Shadow Fading (σ) | 0 dB | 6-12 dB | 3-8 dB |
Margin Allocation
T_ant is the weighted average of the noise temperature of everything the antenna "sees," weighted by the antenna pattern: T_ant = (1/(4pi)) × integral over 4pi of T_brightness(theta,phi) × G(theta,phi) d_omega. For practical calculation: (1) Main beam contribution: the sky temperature in the main beam direction. At zenith, clear sky: 5-15K at 2-10 GHz (depending on water vapor), 20-50K at 20-40 GHz, 200-290K at 60 GHz (oxygen absorption radiates at atmospheric temperature). (2) Sidelobe contribution: the fraction of the pattern directed at the ground picks up ~290K. For a typical parabolic dish with -20 dB first sidelobe: approximately 10-15% of the total pattern is directed at the ground, contributing 30-45K. (3) Spillover: feed energy not captured by the reflector illuminates the ground. Typical spillover contribution: 10-30K for a well-designed feed. For a satellite ground station at 12 GHz pointing at 30° elevation: T_ant = T_sky(30°) + T_ground_sidelobes + T_spillover ≈ 15 + 30 + 15 = 60K. For a mobile phone antenna (omnidirectional, half the pattern sees ground, half sees sky): T_ant ≈ (290 + 50)/2 ≈ 170K.
Propagation Modeling
Any passive component between the antenna and the LNA degrades noise performance: T_feed = (L - 1) × T_physical. This is why satellite ground stations place the LNA directly at the antenna feed (at the focus of the dishes), minimizing feed loss. Feed loss penalties: 0.1 dB loss at 290K: T_feed = 6.7K. 0.3 dB loss: T_feed = 20.6K. 0.5 dB loss: T_feed = 35.4K. 1.0 dB loss: T_feed = 75K. 3.0 dB loss: T_feed = 290K. For a system with T_ant = 20K and T_rx = 30K: adding 0.5 dB of feed loss increases T_sys from 50K to 85.4K (70% degradation). Adding 1.0 dB of feed loss: T_sys = 125K (150% degradation). The sensitivity (proportional to 1/T_sys) is dramatically affected. This is why every fraction of a dB matters in the feed network of high-sensitivity systems.
- Performance verification: confirm specifications against the application requirements before finalizing the design
- Environmental factors: temperature range, humidity, and vibration affect long-term reliability and parameter drift
- Cost vs. performance: evaluate whether the application demands premium components or standard commercial grades
- Interface compatibility: verify impedance, connector type, and mechanical form factor match the system architecture
Fade Mitigation
The G/T (gain-to-noise-temperature ratio) is the standard figure of merit for a receiving system: G/T (dB/K) = G_ant (dBi) - 10×log10(T_sys). Higher G/T means better sensitivity. For a satellite ground station: G_ant = 45 dBi, T_sys = 95K: G/T = 45 - 10×log10(95) = 45 - 19.8 = 25.2 dB/K. For a mobile phone at 2 GHz: G_ant = 0 dBi, T_sys = 600K: G/T = 0 - 27.8 = -27.8 dB/K. The difference (53 dB) explains why satellite ground stations can receive signals from 36,000 km away while mobile phones need a base station within a few km.
Frequently Asked Questions
When should I use noise temperature vs noise figure?
Use noise temperature when: (1) The antenna noise temperature is significantly different from 290K (satellite links, radio astronomy, cold-sky systems). (2) You need precise noise accounting for high-sensitivity systems. (3) Comparing receiver front ends for space or scientific applications. Use noise figure when: (1) Working with standard 50-ohm test environments (the noise figure definition assumes a 290K source). (2) Comparing commercial components (datasheets specify NF, not noise temperature). Convert between them: T = 290×(10^(NF/10)-1) and NF = 10×log10(1+T/290). They carry exactly the same information; the choice is about convenience and precision for the application.
How do I reduce system noise temperature?
In order of impact: (1) Use the lowest-NF LNA possible and place it closest to the antenna (minimize feed loss). A 0.3 dB NF LNA contributes T_LNA = 20K. A 1.0 dB NF LNA: 75K. Difference: 55K. (2) Minimize feed loss: use the shortest possible waveguide/cable between antenna and LNA. Each 0.1 dB saved reduces T_feed by 6.7K. (3) Increase LNA gain: higher first-stage gain reduces the contribution of subsequent stages. 30 dB LNA gain means the second stage contributes T_2/1000 (negligible). (4) Reduce antenna sidelobe level: lower sidelobes pick up less ground noise. (5) Cool the LNA: cryogenic cooling (15-20K LNA physical temperature) reduces T_LNA to 3-5K for the best space-qualified HEMTs. Used in radio astronomy and deep-space receivers.
What is a typical G/T for different systems?
LEO satellite terminal (handheld, 1.6 GHz): G/T = -24 dB/K (low gain, high noise). VSAT terminal (1.2 m dish, 12 GHz): G/T = 15-20 dB/K. Satellite earth station (3 m dish, 12 GHz): G/T = 25-30 dB/K. Large earth station (9 m, with cryogenic LNA): G/T = 35-40 dB/K. Radio telescope (25 m, cryogenic): G/T = 50-60 dB/K. DSN 70 m antenna (Goldstone): G/T ≈ 67 dB/K at S-band (the highest-sensitivity receiving system on Earth).