AETHERIX
← All lessons
Phase 4 · Link Budgets & Orbital Mechanics Day 22 of 42 7 min read Thursday, August 13, 2026

RF Link Budgets (Ka/X/S/UHF)

View source

🎯Learning Objective

Compute a complete RF link budget for Ka-band, including system noise temperature, carrier-to-noise ratio, and Eb/N0 — and explain why RF is AETHERIX's backup to optical despite lower data rates. This addresses exam Learning Objective 4.

📖The Core Concept

While optical provides the throughput, RF provides the reliability. RF signals at Ka-band (26.5 GHz), X-band (8.4 GHz), S-band (2.3 GHz), and UHF (401 MHz) penetrate clouds, tolerate pointing errors orders of magnitude larger than optical, and have decades of flight heritage. AETHERIX's strategy is optical primary, RF backup — RF is always available when optical fails.

The RF link budget chain differs from optical in three critical ways:

  1. Antenna gain is lower. RF parabolic dishes have much lower gain-to-size ratio than optical telescopes. A 3-meter spacecraft dish at Ka-band produces ~+52 dBi; a 34-meter DSN dish produces ~+80 dBi. Compare this to the +113 dBi of a 22 cm optical telescope. The FSPL is lower for RF (because frequency is lower), but the net effect is that RF cannot match optical's received power.

  2. Noise matters. Unlike optical (photon-counting), RF links are noise-limited. The key metric is the system noise temperature: Tsys = Tant + T₀·(NF_linear − 1), where Tant is the antenna noise temperature (deep-space sky is cold, ~50 K), T₀ is the reference temperature (290 K), and NF is the receiver noise figure. With NF = 2 dB (1.585 linear), Tsys = 50 + 290 × (1.585 − 1) ≈ 220 K. Thermal noise power is N = k·T·B, where k is Boltzmann's constant (1.38×10⁻²³ J/K) and B is bandwidth.

  3. Eb/N0 replaces SNR. RF link quality is measured as Eb/N0 (energy per bit to noise density ratio): Eb/N0 = C/N − 10·log₁₀(B/Rb), where C/N is carrier-to-noise ratio and Rb is bit rate. The required Eb/N0 depends on modulation and coding. AETHERIX uses a conservative 10 dB default (assuming legacy uncoded or light coding). Production systems with LDPC or turbo codes would recover ~8 dB.

Ka-band link budget at average distance (225M km):

  • TX power: 20 W (+43 dBm), solid-state power amplifier
  • TX antenna: 3 m dish → ~+52 dBi
  • EIRP: ~+94 dBm
  • FSPL at 26.5 GHz, 225M km: ~−290 dB
  • RX antenna: 34 m DSN dish → ~+80 dBi
  • System temperature: ~220 K
  • Noise power in 12 MHz bandwidth: ~−128 dBm
  • Received power: ~−133 dBm
  • C/N: ~−5 dB → with Eb/N0 adjustment: margin is modest at average distance

Why RF is the backup, not the primary:

RF Ka-band achieves 2–6 Mbps versus optical's 10–200 Mbps. The beam divergence at Ka-band (~0.1°) is ~100× wider than optical (~10 μrad), meaning most of the transmitted energy misses the receiver. Pointing requirements are relaxed (<0.5° vs <10 μrad for optical), making RF operationally simpler. And RF penetrates clouds and rain — when all three optical ground stations are clouded out, the RF link at any DSN complex still works.

The combined availability argument:

If optical site availability = 95.7% (three diverse sites) and Ka-band RF availability = 99.0%, then:

A_combined = 1 − (1−0.957)(1−0.99) = 1 − 0.043×0.01 = 99.96%

This exceeds the 99.9% target. Neither layer alone meets it — optical alone (95.7%) fails the SLA; RF alone (99%) barely meets it. The hybrid exceeds it by two orders of margin.

📊Diagrams & Data

Band selection — the engineering trade
flowchart TD
    F["Choosing a band"]
    F --> UHF["UHF ~400 MHz<br/>proximity links<br/>robust, low rate"]
    F --> S["S-band ~2 GHz<br/>telemetry + command<br/>very reliable"]
    F --> X["X-band ~8 GHz<br/>deep-space workhorse"]
    F --> KA["Ka-band ~32 GHz<br/>high rate, weather sensitive"]
    F --> OPT["Optical 1550 nm<br/>highest rate<br/>needs precise pointing"]
    UHF --> TRADE["Gain and bandwidth rise with frequency;<br/>pointing tolerance and weather margin fall"]
    S --> TRADE
    X --> TRADE
    KA --> TRADE
    OPT --> TRADE
Higher frequency buys antenna gain and bandwidth but costs pointing accuracy and rain/weather margin.

🔬In AETHERIX

The RF link budget is implemented in src/infrastructure/rf_link_budget.py.

The RFLinkBudgetCalculator class is initialized with a carrier frequency. Four frequency constants are defined: - KA_BAND_FREQ_HZ = 26.5e9 (26.5 GHz) - X_BAND_FREQ_HZ = 8.4e9 (8.4 GHz) - S_BAND_FREQ_HZ = 2.3e9 (2.3 GHz) - UHF_FREQ_HZ = 401e6 (401 MHz)

Key methods: - calculate_free_space_loss_db(distance_km): −20·log₁₀(4π·d·f/c) — note this uses frequency rather than wavelength, but is mathematically identical. - calculate_antenna_gain_dbi(diameter_m, efficiency): 10·log₁₀(η·(π·D·f/c)²). - calculate_system_temperature(antenna_temp_k, noise_figure_db): Tant + T₀·(NF_linear − 1) with T₀ = 290 K. - calculate_noise_power_dbm(system_temp_k, bandwidth_hz): 10·log₁₀(kTB) + 30. - calculate_rf_link_budget(...): Complete budget from EIRP to Eb/N0 and link margin.

The calculate_mars_earth_link(scenario) method uses representative deep-space parameters: - Ka-band: 20W TX, 3m spacecraft dish, 34m DSN RX, 10 Mbps data rate, required Eb/N0 = 10 dB. - UHF (surface relay): 10W TX, 0.3m antenna, 5m RX, 256 kbps, required Eb/N0 = 9 dB.

The RFLinkBudget dataclass captures all 23 fields including EIRP, received power, noise power, C/N, Eb/N0, and link margin. The __str__ method produces a formatted ASCII-art report with OPEN/CLOSED status.

📐Key Numbers & Formulas

Band Frequency Use Case Max Data Rate
Ka 26.5 GHz High-rate deep-space downlink 2–6 Mbps
X 8.4 GHz Deep-space standard / conjunction fallback 0.5–2 Mbps
S 2.3 GHz TT&C, emergency <1 Mbps
UHF 401 MHz Mars surface relay 64 kbps – 1 Mbps

Key formulas:

FSPL = −20·log₁₀(4π·d·f/c)           [dB]

G_antenna = 10·log₁₀(η·(π·D·f/c)²)  [dBi]

Tsys = Tant + T₀·(10^(NF/10) − 1)   [K], T₀ = 290 K, Tant ≈ 50 K

N_noise = 10·log₁₀(k·Tsys·B) + 30   [dBm]

Eb/N0 = C/N − 10·log₁₀(B/Rb)       [dB]

Margin = Eb/N0_achieved − Eb/N0_required  [dB]

Ka-band at 225M km: Tsys ≈ 220 K, bandwidth ≈ 12 MHz, noise power ≈ −128 dBm.

🔗Standards & References

💡How the Examiner Will Probe This

QWhy is RF the backup, not the primary?

RF Ka-band achieves 2–6 Mbps vs optical's 10–200 Mbps. The beam divergence is ~100× wider, so most energy misses the receiver. But RF penetrates clouds and tolerates pointing errors 500× larger. It's the guaranteed-delivery layer when optical fails. Combined availability of optical + RF exceeds 99.9%, while neither alone does.

QCalculate the system noise temperature.

Tsys = Tant + T₀·(NF_linear − 1) = 50 + 290×(10^(2/10) − 1) = 50 + 290×0.585 ≈ 220 K. The deep-space sky is cold (50 K), but the receiver's noise figure adds thermal noise. Lower Tsys = better sensitivity — this is why DSN uses cryogenic LNAs.

QWhat happens during solar conjunction?

Ka-band degrades significantly due to solar corona plasma. X-band (8.4 GHz, lower frequency) is more resistant to coronal effects and becomes the conjunction fallback. But even X-band is degraded — the Lagrange relay path (ES-L4/L5) maintains 50–70% availability during conjunction, far better than any direct link.

✅Self-Check Questions

📁Deep Dive Resources

  • Source code: src/infrastructure/rf_link_budget.py — RFLinkBudgetCalculator, RFLinkBudget, factory functions create_ka_band_calculator() and create_uhf_calculator()
  • Demo script: python src/infrastructure/rf_link_budget.py (prints Ka-band and UHF budgets)
  • Topic summary: interview_prep/topic_summaries/link_budget.md — RF Link Budget section
  • Cheat sheet: interview_prep/cheat_sheets/formulas.md — Link Budget Equations
  • Design rationale: docs/DESIGN_RATIONALE.md — §4.4 (Combined availability derivation)