Chapter 26 — Key Takeaways (Communications, Navigation, and Data Handling)

A one-page reference. Reread this before an exam, or before you size any spacecraft link.

Form Use it to find
$P_r = P_t + G_t + G_r - L_{\text{fs}} - L_{\text{other}}$ received power (dBW) from the whole chain
$\text{EIRP} = P_t + G_t$ the transmitter's effective isotropic radiated power (dBW)
$L_{\text{fs}} = 20\log_{10}\!\left(\dfrac{4\pi d}{\lambda}\right)$ free-space path loss (dB) — the inverse-square term
$N_0 = kT_s$; in dB, $-228.6 + 10\log_{10}T_s$ noise spectral density (dBW/Hz)
$\dfrac{P_r}{N_0} = P_r - N_0$ the link's carrier-to-noise-density (dB-Hz) — a rate
$\dfrac{E_b}{N_0} = \dfrac{P_r}{N_0} - 10\log_{10}R$ energy-per-bit margin at data rate $R$ (bit/s)

Link closes when received $E_b/N_0 \ge$ the coded receiver's requirement (≈ 2.3–2.5 dB with strong coding). Margin $= (E_b/N_0)_{\text{received}} - (E_b/N_0)_{\text{required}}$; keep it positive.

Decibel survival kit

Ratio In dB Quantity Reference
×10 +10 dB dBW 1 watt
×2 +3.01 dB dBm 1 milliwatt
×1000 +30 dB dBi isotropic antenna (gain)
÷2 −3 dB dB-Hz ratio per hertz (a rate)

Power in dBW $= 10\log_{10}(P/1\ \text{W})$. Never use $10\log_{10}$ for path loss — it is squared, so $20\log_{10}$.

Antennas — gain and beamwidth (two views of $D/\lambda$)

Quantity Formula Units
Dish gain $G = \eta\left(\dfrac{\pi D}{\lambda}\right)^2$ ratio → dBi
Half-power beamwidth $\theta_{3\text{dB}} \approx 70^\circ\,\dfrac{\lambda}{D}$ degrees
Gain–beamwidth link $G \approx \dfrac{29{,}000}{\theta_{3\text{dB}}^2}$ ($\theta$ in deg) —

$\eta \approx 0.5$–$0.7$ (spacecraft ~0.6, big ground dishes ~0.7). High gain = narrow beam = hard pointing. Carry a low-gain (near-isotropic) antenna for safe mode.

The Deep Space Network

  • Three complexes ~120° apart in longitude — Goldstone, Madrid, Canberra — for continuous coverage as Earth rotates. Each has one 70 m and several 34 m antennas.
  • Station quality is one number: $G/T$ (dB/K). Raise it with a bigger dish (more $G$) or colder receiver (less $T$). A DSN 70 m at X-band: $G_r \approx 74\ \text{dBi}$, $G/T \approx 61\ \text{dB/K}$.
  • Arraying combines dishes; gain rises with total collecting area, $\Delta G = 10\log_{10}(\sum D_i^2 / D_{\text{ref}}^2)$ — a few dB that decide a distant encounter.

Bands, coding, and the data-rate scaling

Band Downlink Trade
S ~2.3 GHz robust, wide beam, low rate
X ~8.4 GHz the deep-space workhorse
Ka ~32 GHz highest rate; hard pointing, rain loss
  • For fixed dishes, higher $f$ wins: $P_r \propto f^2$ (two gains beat one path loss).
  • Data rate $\propto 1/d^2$ (from $P_r \propto 1/d^2$): $R_2/R_1 = (d_1/d_2)^2$.
  • Coding gain ≈ 7 dB (convolutional + Reed-Solomon) → worth ×5 in rate; modern turbo/LDPC approach the Shannon limit $E_b/N_0 > -1.6\ \text{dB}$.
Technique Measures How Precision
Ranging distance $d = c\tau/2$ (round-trip time) ~meters
Doppler radial velocity $\Delta f/f = -v/c$ sub-mm/s
Delta-DOR plane-of-sky angle two-station interferometry, differenced vs a quasar ~few nrad (~km at Mars)

These three feed orbit determination (Ch. 13), which fuses them into a trajectory; that trajectory drives GN&C (Ch. 27). Round-trip light time $= 2d/c$ (≈ 9 h to Pluto) forces autonomy.

Radiation-hardened computers (theme #2 in silicon)

Threat What it does Defense
Single-event upset (SEU) flips a bit (soft error) EDAC memory, triple-modular redundancy
Single-event latchup (SEL) parasitic short; can destroy chip rad-hard fab (SOI), current limiting
Total ionizing dose (TID) slow transistor degradation (rad/Gy) shielding, rad-hard process

Rad-hard = larger features + redundancy → ~15–20 years behind consumer chips, on purpose. Example: RAD750, ~200 MHz, hundreds of krad, ~\$200k; the priority is availability, not speed.

Numbers worth memorizing

  • New Horizons at Pluto: $P_r \approx 1.4\times10^{-18}\ \text{W}$, ~1 kbit/s, RTT ~9 h.
  • Deep-space X-band path loss at tens of AU: ~300+ dB.
  • Boltzmann in dB: $-228.6\ \text{dBW/K/Hz}$. Every ×10 in power = 10 dB; every ×2 = 3 dB.
  • Received power and data rate both scale as $1/d^2$; for fixed dishes, $P_r \propto f^2$.

Common pitfalls

Pitfall Reality
Using $10\log_{10}$ for path loss It is $20\log_{10}(4\pi d/\lambda)$ — the ratio is squared.
"More power is the fix." Power is linear in dB and scarce; gains are squared — enlarge dishes / array / cool the receiver.
"Doubling distance halves the signal." It quarters it ($1/d^2$).
Forgetting antenna efficiency $\eta$ Omitting $\eta \approx 0.6$ overstates gain by ~2 dB.
Confusing $P_r/N_0$ (dB-Hz) with $E_b/N_0$ (dB) Subtract $10\log_{10}R$ to go from the first to the second.

Mission / astrotools additions this chapter

  • MDR: added your mission's communications note — band, data rate and volume, ground stations, antenna, and transmitter power (drawn from the Ch. 25 power budget).
  • comms.py: watts_to_dbw(p), fspl_db(d, f), dish_gain_db(D, f, eff), link_budget(...) — all in decibels.