Deep-Space Optical Link Budget

First-order photon-budget feasibility check for a deep-space laser link — diffraction-limited divergence, uniform-spot geometric collection, and dominant-term ranking.

📊 Sample calculation — DSOC public scenario anchor (Dec 14, 2023): full Gaussian-beam + tolerance-budget methodology → +1.7 dB GREEN at 267 Mbps (P(lab ≥ public-anchor rate) = 31%)

Inputs

Transmit power $P_t$ (W)

Wavelength $\lambda$ (nm)

Transmit aperture $D_t$ (m)

Receive aperture $D_r$ (m)

Range units

Range $R$ In the units selected above.

Data rate $B$ (bits/s)

Optical efficiency $\eta$ Combined Tx + Rx end-to-end optical throughput.

Pointing loss $L_p$ (dB)

Atmospheric loss $L_a$ (dB)

Required photons / bit $N_{\gamma/bit,req}$

Transverse velocity $v_\perp$ (m/s) Used for scalar point-ahead angle.

Use vector point-ahead geometry Off by default. Scalar mode is sufficient for the public DSOC scenario.

Pointing tolerance (μrad) Compared to point-ahead angle.

Result

Beam divergence (half-angle)

—

Spot diameter at receiver

—

Received power

—

Photon rate

—

Photons / bit

—

Link margin

—

Point-ahead angle

—

Pointing tolerance

—

Formulas used (formula_registry.json) — click to expand

Loading…

Term-by-term loss budget

Each row is the dB cost of a single term in the budget. Sorted by magnitude — the top row is the dominant loss.

Term	Loss (dB)	Notes

Dominant term: —

Tolerance-budget loss-budget analysis

Applies a standard tolerance-budget methodology (Hemmati DESCANSO public monograph) to the DSOC engineering loss stack. Each channel has a {nominal, min, max} range from public DSOC literature (Hemmati DESCANSO; published SNSPD specs; standard atmospheric optics). Generic tolerance decomposition L_total,dB = Σ L_channel with independent uniform per-channel draws.

Target rate (Mbps) DSOC public scenario anchor: 267 Mbps.

Monte Carlo samples

Loss channels (editable)

Edit any cell to re-run the analysis live. All values are from public optical-comms literature; this page does not access internal mission parameters.

Channel	Nominal (dB)	Min (dB)	Max (dB)	Source
Σ Total	— dB	— dB	— dB	linear-dB sum (independent channels)

Predicted achievable rates (lab Gaussian-beam baseline × loss budget)

Best case (all losses at min)

—

Nominal case

—

Adverse case (all losses at max)

—

Baseline (no engineering losses)

—

Margin band vs target (link margin certification threshold)

Margin bands: GREEN ≥ +1.5 dB, YELLOW 0…+1.5 dB, RED < 0 dB. Standard link margin certification threshold on L_total.

Monte Carlo (uniform per-channel over [min, max], independent draws)

5th percentile

—

Median (50th)

—

95th percentile

—

P(lab ≥ target)

—

Channel ablation: "remove this channel, recompute predicted rate"

Standard tolerance-budget channel ablation: rank channels by the recovery you'd see if that engineering cost vanished entirely.

Removed channel	Recovery (dB)	Predicted rate (Mbps)	Uplift × baseline

Export tolerance budget

Formulas

Diffraction-limited (Airy) half-angle $$\theta_{\rm div} \;=\; 1.22\,\frac{\lambda}{D_t}$$ Spot radius at the receiver $$r_{\rm spot} \;=\; R\,\theta_{\rm div}$$ Uniform-spot geometric collection fraction $$f_{\rm collect} \;=\; \min\!\left(1,\;\frac{A_r}{\pi r_{\rm spot}^2}\right),\quad A_r = \pi(D_r/2)^2$$ Received power $$P_r \;=\; P_t\,\eta\,f_{\rm collect}\,10^{-(L_p + L_a)/10}$$ Photon energy, rate, and per-bit count $$E_\gamma = \frac{hc}{\lambda},\quad \dot N_\gamma = \frac{P_r}{E_\gamma},\quad N_{\gamma/bit} = \frac{\dot N_\gamma}{B}$$ Link margin (dB) $$M \;=\; 10\log_{10}\!\left(\frac{N_{\gamma/bit}}{N_{\gamma/bit,req}}\right)$$ Point-ahead angle $$\theta_{\rm PA} \;=\; \frac{v_\perp}{c}$$

Assumptions

Diffraction-limited (Airy) transmit beam: $\theta_{\rm div} = 1.22\lambda/D_t$ as the half-angle envelope.
Uniform-spot geometric model — beam treated as a top-hat of radius $r_{\rm spot}$ at the receiver; the receive aperture is a point compared with the spot.
Point-source aperture model for the receiver (no obscuration, central baffle, or coupling-loss model).
No turbulence, no scintillation, no adaptive-optics correction.
No detector model: no quantum efficiency, no thermal/dark-count noise, no shot-noise penalty.
Pointing and atmospheric losses are user-supplied lump dB terms.
Photon energy is the classical monochromatic value $E_\gamma = hc/\lambda$.

What this would need for mission-grade use

Show requirements

Mission-specific aperture truncation + obscuration model
Photon-counting detector quantum efficiency + dark-count + jitter model
PPM modulation + LDPC/SCPPM coding implementation
Active pointing-control loop + jitter PSD
Adaptive-optics residual model + atmospheric channel
Atmospheric absorption/scintillation profile (site-specific)
SPICE ephemeris for range + range-rate
Mission-specific link calibration

Warnings

This is a public engineering surrogate. Uniform-spot geometric approximation; does not model adaptive optics, turbulence, detector internals, or proprietary DSN / DSOC calibration.
Uniform-spot is conservative for a real Gaussian beam: a true Gaussian concentrates more power on-axis, so $f_{\rm collect}$ is typically underestimated when the receiver is well-aimed.
The collection fraction is clipped to $\le 1$ when the receiver aperture exceeds the spot — this is the near-field / oversized-aperture regime and the formula is no longer geometrically meaningful.
The 1.22 factor is the Airy first-null half-angle, not the encircled-energy radius — use as an order-of-magnitude envelope, not a precise on-target metric.
Ignores quantum-limited detector physics (PPM signaling, threshold detection, code rate, FEC overhead).
Point-ahead uses $v_\perp/c$ and ignores aberration / relativistic correction terms — adequate at $v \ll c$ only.

Source notes

Airy / diffraction-limited divergence: any standard optics text; Born & Wolf, Principles of Optics §8.
Optical-comm link-budget structure (geometric / pointing / atmospheric / optical / receiver-sensitivity split): standard treatment in laser-comm literature (e.g. Hemmati, Deep Space Optical Communications, JPL DESCANSO, public-domain edition).
Constants from CODATA / IAU; full table on the Validation page.
Not a DSN or DSOC operational tool — no proprietary calibration, no mission ephemeris.

Certificate

Each computed result can be exported as a Certificate v2 (JSON or Markdown). The deterministic result_hash depends only on inputs/constants/formulas/version; each export event also gets a unique receipt_hash that includes the timestamp.