Deep space communications ground stations are on the ground because their transmit and receive electronics alone is bulky and heavy, not to mention their 34 and 70 meter dish antennas!

But those constraints change when it can be a 1 meter or smaller telescope with a laser diode or photodiode (or superconducting thingy) at the focal plane, exploiting the much smaller theoretical diffraction limit. For example right now in DSN Now I see DSS-14 a 70 meter dish receiving signals from Juno 8.4 GHz. That's a wavelength of 3.6 centimeters, so $(D/\lambda)^2 \approx \text{3.8} \times \text{10}^6$.

If instead a 0.85 meter telescope was used at 850 nm (a random, typical AlGaAs laser wavelength) we'd have $(D/\lambda)^2 \approx \text{10}^{12}$

Remember that the far-higher gain is also available at the other end of the link as well, so this analysis is a gross underestimate of the total improvement in link budget going optical, but that's okay because I haven't included some of the challenges.

A few watt laser is comparable to a few watt deep space probe transmitter, and only four orders of magnitude lower than a DSN transmitter (except for things lke Has DSS-43 ever been used in high power mode (>>20 kW) for an emergency situation?) so our circa-one-meter telescope "ground station" beats a 70 meter dish if it can be pointed steadily and around any intervening clouds

Hubble can be pointed steadily and around any intervening clouds, and there are several high altitude locations around the world where optical telescopes can be pointed steadily and usually around intervening clouds, often using adaptive optics. (see also the surprising answers to Why aren't ground-based observatories using adaptive optics for visible wavelengths?)

Question: Optical deep space transceiving stations of circa one meter diameter could be Earth orbit or on the ground in several places, with adaptive optics if necessary. What are the most important technical tradeoffs that will determine where they will most likely end up being deployed?

Below this answer I'd summarized the following collection of Q&A on optical communications for deep space:

"interplanetary radio communication" will go away and be replaced by optical in the not-so-distant future because a 20 kW transmitter or liquid helium-cooled front-end receiver at the end of 34 meter dish on Earth can be replaced by a few watt laser diode and an avalanche photodiode or one of those superconducting thingies at the end of a 20 or 50 cm diameter telescope. (see also)

See also Quantitatively, why will optical communication be better than X-band for deep-space communications? and How is long-distance optical communications coming along in space? and Are there plans or a program for an optical relay pathfinder for deep space? and What optical design is used by the GEDI's receiving telescope and how is the secondary held in place? (optical com will look similar)

and When will STP-3 launch with the new optical space coms test and why is it late? and What GEO relay satellite will the ISS use for end-to-end optical communication with a ground station?

DSN Now screenshot

A very heavy ground station, very likely to remain on the ground for the foreseeable future:

enter image description here

DSS-43 from here from NASA.

Summary of DSN diameters and transmit powers in this answer

  • 2
    $\begingroup$ It's notable that as far as has been publicly discussed, Starlink will have optical interlinks only for satellite-to-satellite communications. For talking to the ground, it uses radio. $\endgroup$ Commented Nov 17, 2020 at 5:00
  • 2
    $\begingroup$ If the optical deep space transceiving station is placed in an area with (almost) no clouds like the Atacama desert, it should work. $\endgroup$
    – Uwe
    Commented Nov 17, 2020 at 10:19
  • 1
    $\begingroup$ @uhoh That sounds really interesting. Can you expand on what you mean by intermittent sampling and give an example? $\endgroup$
    – ChrisR
    Commented Nov 18, 2020 at 4:34
  • 2
    $\begingroup$ SpaceX is already deploying laser interlinks on Starlink satellites, which certainly qualify as small spacecraft. They need only equip a few satellites in one or two planes with some additional links for interplanetary relay satellites, and they'll have massively redundant downlink capability through the phased array antennas used to provide internet service. $\endgroup$ Commented Nov 21, 2020 at 15:59
  • 2
    $\begingroup$ @uhoh the standard Starlink sats are equipped with enough laser links to connect to neighboring Starlinks (likely previous and next in the same plane, and some more for crossing planes), and I was mainly thinking that they'd need additional links to handle duties outside of that...though they might just redirect one of the normal links instead. I wouldn't expect the low-altitude Starlinks to directly handle interplanetary comms, but I can see the standard inter-Starlink link hardware being sufficient for reaching a relay satellite in a higher orbit. $\endgroup$ Commented Nov 21, 2020 at 23:17

1 Answer 1


One report of a study on the topic funded by NASA JPL in 2005 is "Deep-space to ground laser communications in a cloudy world". Based on global cloud statistics, they worked out the probability of success as a function of number of receiving stations and optimized their location, and found that the number needed to achieve the desired availability was well over budget. One ground station in the Atacama is good, but multiple ones all in the same desert are not, because the autocorrelation length of the cloud/no-cloud signal they found is around 600 km, so if one station in Chile was clouded out, so would the other ones probably be at the same time. You want them all at high altitude mountain observatories, but scattered over every continent, including Haleakala, Kilimanjaro if you could get permission, and other widely separated high and dry places. Another problem is turbulence, which since it is caused largely by solar heating, is worst exactly where clouds are least frequent, so you need astronomy-grade adaptive optics and probably multi-meter deformable mirrors to be able to correct the signal as it arrives. Aerosols are another consideration, which makes coastal regions less good due to sea spray, even separate from the humidity that produces clouds. My own conclusion from following this group's work for several years was that the right way to do it is laser from deep space to earth orbit, with the space-to-ground links in radio for affordable availability.

  • $\begingroup$ Thank you for the both thoughtful and thorough answer! After reading it I'm completely convinced. I do realize however that my question offered a false binary choice; there is of course a third option; telescopes hanging from high altitude balloons :-) I'm imagining something like a mashup of Project Loon + ASTHROS somehow. It could be overseen by surplus managers from JWST and the F-35, and SpinLaunch can help raise early funding. $\endgroup$
    – uhoh
    Commented Mar 2, 2021 at 8:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.