Communication Technology

 

 

 

Laser Communication

 

Laser communication carries data on a beam of light through open air or vacuum, with no fibre between the two ends. The physics is the same as fibre optics and the engineering is not, because a beam that travels through nothing has to be aimed. This page starts from what the technique buys you and then works through what it costs.

What is laser communication ?

Take a fibre optic link and remove the fibre. What remains is a laser at one end, a telescope and a photodetector at the other, and whatever sits between them. That is laser communication, and the industry also calls it free space optics, or FSO, when the path runs through the atmosphere.

The modulation is usually simpler than a radio engineer expects. Most links switch the laser on and off, or vary its brightness, and the receiver measures how much light arrives. That arrangement is called intensity modulation with direct detection, and it needs no local oscillator and no phase reference. Deep space links often use pulse position modulation instead, because it carries more bits for each photon delivered, and photons are the scarce resource at those distances.

More capable terminals do use coherent detection, mixing the incoming light with a local laser exactly as a radio receiver mixes with a local oscillator. That buys sensitivity and it costs a great deal of optical stability, so it appears where the link budget is tight rather than everywhere.

Three families of link use the same idea at very different scales. Terrestrial links join two buildings across a street or a campus. Inter-satellite links join spacecraft to each other. Deep space links carry data from a probe back to Earth. The hardware differs enormously and the physics does not.

  • It is fibre optics without the fibre : the same wavelengths, the same detectors and much of the same component supply chain. What changes is that the beam has to be aimed and the medium is no longer controlled.
  • Most links just switch the light on and off : intensity modulation with direct detection needs no phase reference. That is why a hobbyist can build a working optical link and a radio equivalent is much harder.
  • One idea, three very different scales : a link across a street and a link from deep space share their physics and share almost none of their engineering.

Why use light instead of radio ?

Three reasons matter here, and only one of them is really about bandwidth. The other two matter more in practice, and the third one explains why space agencies care.

The first is raw spectrum. A carrier at 1550 nm sits near 194 THz, and the usable fractional bandwidth around it is far larger than anything available in radio. No radio band offers room of that order, and the comparison is not close.

The second is licensing. Optical frequencies are not allocated and auctioned the way radio spectrum is, so a link can be deployed without buying rights to a band. For an operator facing a spectrum bill, that alone can decide the technology.

The third reason is geometry, and it is the one worth doing the arithmetic on. A beam leaving an aperture spreads by roughly the wavelength divided by the aperture diameter. Put a 10 cm aperture at 30 GHz, where the wavelength is 1 cm, and the beam spreads by about 100 milliradians. Put the same 10 cm aperture at 1550 nm and it spreads by about 15 microradians instead.

Those two numbers differ by a factor of roughly 6000, and Figure 1 sets them side by side.

The same aperture, two wavelengths RF, 30 GHz divergence about 100 mrad Optical, 1550 nm divergence about 15 microrad same aperture diameter, same range Divergence is roughly the wavelength divided by the aperture diameter, so a 10 cm aperture gives about 100 mrad at 30 GHz and about 15 microrad at 1550 nm. The two beams are drawn at different scales. The real ratio is about 6000 to 1 in angle, which is about 40 million to 1 in illuminated area.

Figure 1. Divergence falls as the wavelength falls, so the same aperture that produces a wide radio beam produces a needle at optical wavelengths. The gain in received power is the square of the ratio in angle, which is why deep space missions want optical links.

Angle is not the payoff on its own. Power arriving at the far end scales with the inverse square of the beam width. So a factor of 6000 in angle becomes a factor near 40 million in delivered power, for the same transmitted watt. No amount of extra amplifier power buys that, and no practical radio aperture matches it.

A narrow beam brings two further benefits. Nobody outside the beam can receive it, which makes interception difficult and makes an optical link attractive where confidentiality matters. And two links pointing in slightly different directions do not interfere, so a constellation can run many optical crosslinks at once with no frequency planning at all.

  • Bandwidth is the least interesting of the three reasons : it is the one always quoted, and licensing and beam geometry decide more real deployments.
  • Divergence falls with wavelength, and power gain goes as its square : the same aperture gains roughly 40 million times in delivered power moving from 30 GHz to 1550 nm.
  • A narrow beam is its own security and its own frequency plan : you cannot receive what you are not standing in, and two beams that miss each other cannot interfere.

What does the narrow beam cost ?

Every advantage above came from the same property, so every difficulty comes from it too. A beam 15 microradians wide has to be aimed to within a few microradians, and it has to stay there while both ends move. Two problems follow, and they are independent of each other.

Pointing, acquisition and tracking

Fifteen microradians is about one thousandth of a degree. At a range of 1000 km that beam is roughly 15 m across, so an aiming error of a few microradians misses the receiver entirely. The problem has a name in the industry, Pointing Acquisition and Tracking, usually shortened to PAT, and a laser terminal spends most of its complexity on it.

Acquisition comes first and it is the hard part. Neither end knows precisely where the other is, so a wide beacon beam sweeps a search area until the far end detects it and answers. Only then does the link narrow down to the data beam. That sequence is the optical equivalent of the initial synchronization a radio link performs, and it takes seconds rather than microseconds.

Tracking then has to hold the lock. Spacecraft vibrate, buildings sway in wind, and a platform jitter of a few microradians is enough to break an optical link that a radio link would never notice. Terminals answer with a fast steering mirror driven by a position sensor, correcting far faster than the spacecraft body can be moved.

What the atmosphere does

The second problem applies only to links that pass through air, and it is the reason terrestrial FSO has never displaced fibre. Space to space links avoid all of it, which is a large part of why optical crosslinks became routine in orbit before they became routine on the ground.

Fog is the worst case by a wide margin. Fog droplets are comparable in size to the optical wavelength, so they scatter it strongly, and attenuation in dense fog reaches hundreds of dB per kilometre. No link budget survives that. Rain damages an optical link far less than intuition suggests. Raindrops are large compared with the wavelength, so they scatter it less selectively than they scatter millimetre waves.

Turbulence causes the second effect. Pockets of air at different temperatures bend the beam slightly, and by different amounts across the aperture. The received power fluctuates and the beam wanders off boresight. Engineers call the fluctuation scintillation, and it is the same effect that makes stars twinkle.

Both effects are about availability rather than capacity. The link runs at full rate or it stops. A radio link therefore sits beside most FSO deployments that need carrier grade availability, and takes over when the optical path closes. That hybrid arrangement is common enough to be the default rather than the exception.

  • Acquisition is harder than tracking : a wide beacon has to find the far end before the narrow data beam can be used, and that search takes seconds.
  • Platform jitter breaks optical links that radio would ignore : a few microradians of vibration is nothing to a radio beam and is the whole beam width to an optical one.
  • Fog stops the link, rain mostly does not : droplet size against wavelength decides which one matters, and it is the opposite of the intuition an RF engineer brings.
  • The atmosphere costs availability, not throughput : an optical link runs at full rate or not at all, which is why a radio fallback sits beside most terrestrial deployments.

Where is it actually used ?

The technology has moved from demonstration to deployment unevenly, and the pattern follows the previous section exactly. Where there is no atmosphere the case is strong, and where there is atmosphere the case depends on what the alternative costs.

The table below sorts the four places an engineer is most likely to meet an optical link.

Application

Typical range

Why light rather than radio

Inter-satellite crosslink

Hundreds to thousands of km

No atmosphere, no spectrum licence, and many links can run at once without frequency planning.

Deep space downlink

Millions of km and beyond

The beam gain is the only practical way to close a link at that range with a spacecraft sized aperture.

Optical feeder link

Ground station to satellite

Moves the capacity bottleneck off congested RF feeder spectrum.

Terrestrial FSO

Hundreds of metres to a few km

Fibre-like capacity with no trench to dig, installed in a day rather than a quarter.

Inter-satellite crosslinks are the clearest success. A large constellation needs its satellites to talk to each other rather than route everything through the ground. An optical crosslink is licence free, interference free and light enough to fly. Constellation operators now treat optical crosslinks as ordinary equipment rather than as a demonstration.

Deep space is where the argument is strongest and the engineering hardest. NASA demonstrated the case in 2013 with the Lunar Laser Communication Demonstration, which returned 622 Mbit/s from lunar orbit, far beyond what a comparable radio link delivers. Later missions have pushed optical links much further out, and the reason is always the beam gain rather than the bandwidth.

Terrestrial FSO is the one that has never become common. It solves a real problem, which is reaching a building that fibre does not serve, and the fog availability problem keeps it in niches. One adjacent technology is worth separating from it. Li-Fi uses modulated LED lighting indoors rather than a laser outdoors. It shares the idea of carrying data on light and almost none of the engineering discussed here.

  • Orbit first, ground later : the applications without atmosphere matured first, which is exactly what the physics predicts.
  • Deep space uses optics for gain, not for bandwidth : at those ranges the narrow beam is the only way to deliver enough power, and the data rate is a consequence.
  • Terrestrial FSO is limited by availability, not by capability : the capacity is there and the fog is the reason it stays a niche.
  • Li-Fi is a different thing : indoor, LED based and diffuse, so it shares the headline and not the problems.

YouTube

Readings