Data transmission using an optical laser between ground station and a GEO satellite may offer very high data rates, up to 1.8 GBit/s for instance.

But what about a transmission from Earth to Mars without any repeater between the planets? Will the huge distance decrease the maximum data rate to MBit/s or even kBit/s? Using a much more powerful laser and larger aperture sizes for transmitter and receiver may increase the possible data rate, but to what extent? At least some photons per data bit are needed for an acceptable error rate.

  • $\begingroup$ This is a great question. By asking about photons per bit, it avoids things like aperture size and instead addresses the underlying principles and limitations. $\endgroup$
    – uhoh
    Commented Jan 12, 2018 at 23:18
  • 3
    $\begingroup$ Actually, that can be misleading. You can have more than one bit per photon. $\endgroup$
    – Mark Adler
    Commented Jan 13, 2018 at 16:57

3 Answers 3


No, you don't need "at least some photons per data bit". 13 bits per photon has been demonstrated with laser communications. You calculate the data rate capability the same way you do with any other wavelength, which is using the power, range, transmit and receive apertures, noise, modulation scheme, and coding gain.

This paper summarizes detailed analyses of the power and mass comparisons of spacecraft radios for the same data rate at Ka-band and laser comm at $1.55\,\mathrm{\mu m}$. The ground stations were equivalenced based on the cost of construction, which ended up being an array of radio dishes with an aperture equivalent to an $80.5\,\mathrm m$ antenna, and an optical telescope with a $10\,\mathrm m$ aperture.

Also for apples-to-apples comparison, both systems assumed the same pointing accuracy requirement for the spacecraft, with the laser terminal responsible for fine-tuning its telescope pointing with the additional accuracy needed for the smaller beam width.

The benefit is much more dramatic farther than Mars, but at maximum Mars distance, the mass of a 1 Gbps RF system would be more than twice that of a laser system, and the power required for that RF system would be 13 times the laser system. There is no question that at 1 Gbps, even at Mars, you would use a laser system.

  • 1
    $\begingroup$ That's incredible! While the IEEE paper seems paywalled, clicking view/open in this link trs.jpl.nasa.gov/handle/2014/44268?show=full shows some related slides. Does the sentence "In PPM, a single laser pulse in one of $M$ symbol slots encodes $\log_2(M)$ information bits" (where PPM is Pulse Position Modulation) concisely summarize the way this works? Since $h \nu$ is so much larger than $k_B T$, individual photons can be detected, and timed to sub-nanosecond accuracy? $\endgroup$
    – uhoh
    Commented Jan 14, 2018 at 3:38
  • 4
    $\begingroup$ Yes. Simply consider the arrival time of the single photon as carrying the information. $\endgroup$
    – Mark Adler
    Commented Jan 14, 2018 at 3:38
  • 2
    $\begingroup$ @uhoh Fundmental physics will allow you to time a photon to an accuracy roughly the same as it's inverse frequency. For 1.5 $\mu m$ that is about $5 fs$. Engineering is likely to limit you well short of this, but picosecond accuracy should be attainable. Especially as you only need to meaure relative pulse arrival times, not absolute times. $\endgroup$ Commented Apr 22, 2019 at 15:27
  • 1
    $\begingroup$ @uhoh at single photon brightness levels? JUst asking because I'm interested. $\endgroup$ Commented Apr 22, 2019 at 20:09
  • 1
    $\begingroup$ @uhoh A non-paywalled version is available at sci-hub.tw/10.1117/12.2007000. $\endgroup$
    – forest
    Commented Aug 15, 2019 at 23:51

It's generally true that the error rate in communication is proportional to the energy per bit. Rigorously you can see this through things like $E_b/N_0$ and the Shannon Hartley theorem. To compensate for the greater distance to Mars, one could either increase the transmitter power or antenna gain to maintain the same power, or transmit each bit for a longer time thus accumulating more energy in each bit.

GEO is a distance of approximately 36×103 km, while Mars is between 55×106 km and 401×106 km, depending on the relative orbits of Mars and Earth.

Applying the inverse square law, this means the power will be reduced by some factor of at least

$$ \left(36 \times 10^3 \over 55 \times 10^6 \right)^2 = 4.28 \times 10^{-7} $$

but not more than

$$ \left(36 \times 10^3 \over 401 \times 10^6 \right)^2 = 8.06 \times 10^{-9}$$

Reducing the bitrate by the same factors means the communication system given in the example would operate between 648 and 12 bits per second, all else equal.

That's not to say that laser communication to Mars couldn't work, just that the greater distance requires equipment engineered for the purpose. A Mars communication system will necessarily have higher powers, bigger apertures, and more expensive infrastructure which simply wasn't necessary for a GEO system.


Extended comment supplementing @MarkAdler's oft-cited "13 bits per photon" answer

Is it possible to extend high speed data transmission with lasers to the distance Earth to Mars?

Yes, it seems so!

Wikipedia's Psyche (spacecraft); Laser communications experiment or Deep Space Optical Communications (DSOC) unit:

The Discovery program solicitation offered mission projects an extra $30 M if they would host and test the 25 kg DSOC unit which needs about 75 Watts. It is hoped to advance DSOC to TRL 6. DSOC tests should begin about 60 days after launch. The test-runs of the laser equipment will occur over distances of 0.1 to 2.5 astronomical units (AU) on the outward-bound probe.

That probably qualifies as "...to the distance Earth to Mars" since although that could be as much as 2.7 AU (as opposed to 2.5) that's at opposition when you'd be pointing your telescopes in the general direction of the Sun. Of course with narrow band optical interference filters for the laser wavelengths, careful baffling, spatial filtering, "coronographing" and other stray light management techniques, you really could point close to the Sun if you really wanted or needed to.

JPL's Aug. 7, 2023 NASA’s Deep Space Communications to Get a Laser Boost shows the dripping-with-new-technology optical communications system attached to the side of the Psyche deep space spacecraft and discusses how it will operate in great detail.

It also importantly links JPL's (older but undated) Superconducting Nanowire Single Photon Detectors for DSOC

"The Deep Space Optical Communications (DSOC) flight transceiver is inside a large tube-like sunshade and telescope on the Psyche spacecraft, as seen here inside a clean room at JPL. An earlier photo, inset, shows the transceiver assembly before it was... Credit: NASA/JPL-Caltech"

The Deep Space Optical Communications (DSOC) flight transceiver is inside a large tube-like sunshade and telescope on the Psyche spacecraft, as seen here inside a clean room at JPL. An earlier photo, inset, shows the transceiver assembly before it was... Credit: NASA/JPL-Caltech

Groundbreaking Technologies

The transceiver riding on Psyche features several new technologies, including a never-before-flown photon-counting camera attached to an 8.6-inch (22-centimeter) aperture telescope that protrudes from the side of the spacecraft. The transceiver will autonomously scan for, and “lock” onto, the high-power near-infrared laser uplink transmitted by the Optical Communication Telescope Laboratory at JPL’s Table Mountain Facility near Wrightwood, California. The laser uplink will also demonstrate sending commands to the transceiver.

“The powerful uplink laser is a critical part of this tech demo for higher rates to spacecraft, and upgrades to our ground systems will enable optical communications for future deep space missions,” said Jason Mitchell, program executive for NASA’s Space Communications and Navigation (SCaN) program at NASA Headquarters.

Once locked onto the uplink laser, the transceiver will locate the 200-inch (5.1-meter) Hale Telescope at Caltech’s Palomar Observatory in San Diego County, California, about 100 miles (130 kilometers) south of Table Mountain. The transceiver will then use its near-infrared laser to transmit high-rate data down to Palomar. Spacecraft vibrations that might otherwise nudge the laser off target will be dampened by state-of-the-art struts attaching the transceiver to Psyche.

To receive the high-rate downlink laser from the DSOC transceiver, the Hale Telescope has been fitted with a novel superconducting nanowire single photon detector assembly. The assembly is cryogenically cooled so that a single incident laser photon (a quantum particle of light) can be detected and its arrival time recorded. Transmitted as a train of pulses, the laser light must travel more than 200 million miles (300 million kilometers) – the farthest the spacecraft will be during this tech demo – before the faint signals can be detected and processed to extract the information.

“Every component of DSOC exhibits new technology, from the high-power uplink lasers to the pointing system on the transceiver’s telescope and down to the exquisitely sensitive detectors that can count the single photons as they arrive,” said JPL’s Bill Klipstein, the DSOC project manager. “The team even needed to develop new signal-processing techniques to squeeze information out of such weak signals transmitted over vast distances.”

The distances involved pose another challenge for the tech demo: The farther Psyche journeys, the longer the photons will take to reach their destination, creating a lag of up to tens of minutes. The positions of Earth and the spacecraft will be constantly changing while the laser photons travel, so this lag will need to be compensated for.

“Pointing the laser and locking on over millions of miles while dealing with the relative motion of Earth and Psyche poses an exciting challenge for our project,” said Biswas.

Notably, the superconducting nanowire detector was not included on the spacecraft as well (for this demonstration) so the ground station needs to have a particularly high power laser and significantly large aperture so that a conventional photodetector can be used as the receiver on the spacecraft.

While conventional radio communications (to date, used for every deep space mission critical communications) signal to noise ratio S/N scales as the familiar

$$\frac{1}{r^{ \ 2}} \frac{1}{k_B \ T}$$

where $T$ is the temperature of the radio receiver's front-end amplifier and $r$ is the distance between transmitter and receiver, the S/N of deep space optical communications systems based on conventional photodetectors scales as

$$\frac{1}{r^{ \ 4}} f(k_B \ T)$$

i.e. an inverse fourth power of distance and a more complicated function of front-end temperature, because the front end involves a two step process - photons are (generally) first converted to electron-hole pairs in a conventional semiconductor photodiode, then the charge is collected and amplified by a charge or transimpedance. That double conversion results in the much faster drop in S/N as the signal gets weaker with distance. For more on that see

Because the superconducting nanowire receiver technology is not robustly deep-space friendly at this point in time, it seems the designers have leveraged it's extra benefits when used as both a photon counter and timer in the Mt. Palormar receiving ground station since the transmitting laser aboard Psyche will likely be only a few up to a dozen watts and the aperture is only 0.22 meters.

If I understand correctly, we can still expect a roughly $~1/r^2$ S/N scaling for both uplink and downlink because even though it's not of the superconducting nanowire variety, the optical receiver aboard Psyche is still a photon counter, rather than an analog light intensity to current converter followed by a current amplifier. So we've got on the uplink a powerful laser coupled to a large aperture telescope received by a small mirror and photon-counting photodetector (probably some kind of avalanche photodiode or other electron multiplier), and a downlink using the fancy superconducting nanowire detector and yes, the 200-inch Hale telescope atop Mt. Palomar, both hopefully providing circa $~1/r^2$ S/N scaling deep space optical links.

And in the future?

An image in this answer indicates an operation temperature for the superconducting nanowire array of "<3 Kelvin", so below that of liquid helium and perhaps even below the cosmic microwave background, so a small but very special kind of refrigerator is needed.

A system like that robust against vibrations associated with launch and conditions in deep space over long periods of time is something that will require a separate development and demonstration program before it can be used routinely for uplink as well.

It's also important to note that you can push the distance even further if you know exactly when to expect the individual photons to arrive or not arrive (again hats off to @MarkAdler's 13 bits per photon answer) and to that end it is possible that deep space atomic clock technology may be of use.

  • 1
    $\begingroup$ You might want to include the most recent test with Psyche: 25 MBit/s over 226 million kilometers. nasa.gov/missions/psyche-mission/… $\endgroup$
    – asdfex
    Commented May 24 at 16:43
  • $\begingroup$ @asdfex Wow, that is just amazing! I think it deserves a whole new answer, because rather than rambling on about future technology and what might be possible, it's a simple "Yes!" Can you consider posting that as a new answer? $\endgroup$
    – uhoh
    Commented May 24 at 21:26
  • 1
    $\begingroup$ You already mentioned the plan for such a test - you could just add a result as another short paragraph above the image. I don't think it needs a new answer - one common text might be easier to read. Let's wait for the "final" test at > 2 AU. $\endgroup$
    – asdfex
    Commented May 25 at 8:48

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.