Free-space optical links, Part 2: Technical issues

Free-space optical links provide a reliable, cost-effective, and quick way to link two fixed communication nodes located a few kilometers apart.

The first part of this article established the basic operating principles and rationale for free-space optical links. This part looks at some practical issues associated with their implementation.

Q: How do you aim the optical transmitter at the receiver?

A: Since the receiver is, by definition, visible to the transmitter, it may seem easy enough to simply look and aim. However, we are talking about an ultra-precise straight line over kilometers in many cases. To facilitate aiming, many units come with a built-in telescopic imaging system similar to a spotting scope on a rifle and coupled with a video camera and display, which the installation crew uses to align and adjust the mounting angle.

Q: What are some possible “mechanical” impairments?

A: Buildings and towers, even if firmly attached to Earth, are not without motion. There are issues of structure swaying, vibration, temporary and even semi-permanent shifts due to thermally induced effects, minor seismic events, traffic-induced micro-motion, nearby mass-transit vibrations, and more. To overcome these difficulties, the “negative” issue of beam divergence is turned to an advantage.

Q: What is this beam divergence?

A: Although popular literature usually portrays the laser beam as a thin beam which retains that thinness over the entire travel distance, in reality, the laser beam spreads or diverges the same as a regular non-laser light does. Of course, for a laser with its coherent output, this divergence is significantly less. Depending on the specific optical emitter and the lens arrangement, the laser beam will spread to some extent.

Q: How much divergence are we talking about?

A: A beam with a typical 1 milliradian (0.057°) spread diverges to one meter over a distance of one kilometer, Figure. Basic math and the small-angle trigonometry approximation make it easy to determine the divergence area at the receiver for other angles and distances (Figure 1).

Fig 1: Beam divergence is both a power limiting degrading signal-to-noise ratio and a useful feature to boost link integrity despite motion and other impediments. (Image: Purdue University)

This divergence presents a classical engineering tradeoff. On one side, you want as little divergence as possible so the receiver can see and capture more of the optical emitter’s energy. However, this has two implications: first, the beam intensity at a distance may exceed allowed regulatory limits. Second, a wider beam means the system will be far less sensitive to misalignment due to vibration and motion. Reduced beam intensity at the receiver means reduced signal-to-noise ratio (SNR) and higher bit error rate (BER) as a general rule.

Q: Are there other issues?

A: Of course. Under the broad topic of path-link attenuation are the effects of water vapor, rain, fog, smoke particles, dust, air turbulence, and other sources of varying losses. These reduce the amount of light reaching the receiver. Each of these factors has a different mechanism. For example, smoke particles will block some light, while rain and fog will reflect and scatter the light, creating an optical multipath.

Q: What’s the impact, in practical terms?

A: The optical-signal attenuation depends on the amount and type of impediment. Further, the attenuation is also a function of the wavelength of the light used for the link. Some of the relationships are not obvious: larger droplets of rain are less of a problem than smaller droplets and fog.

There have been many detailed studies, both theoretical and based on field measurements, on attention due to these factors. For example, during dense fog conditions, when the visibility is less than 50 meters, attenuation can be more than 350 dB/km [Note “visibility” means that only 2% of the light reaching the receiver].

During heavy fog, attenuation at all wavelengths (usually 850, 950, and 1550 nm are used) closely track and follow the same pattern, implying that specific attenuation is independent of the choice of operating wavelength (Figure 2).

Fig 2: In heavy fog conditions, the attenuation versus distance degradation is about the same for all standard FSO laser wavelengths. (Image: Arvix)

In contrast, when the visibility range is high (say, 6 km), specific attenuation is far less for 1550 nm compared to 850 nm and 950 nm (Figure 3).

Fig 3: In lighter fog conditions, the attenuation versus distance degradation differs substantially for the standard FSO laser wavelengths. (Image: Arvix)

Q: Is there more to the attenuation and impairment issue?

A: Yes, absolutely more. It is a multifaceted topic with many considerations, comprehensive research and data, and no possibility of a simple summary. The References discuss the issue in much more detail.

Q: What are some of the options to overcoming these impediments and sources of signal attenuation?

A: As with most engineering issues, there are several options. The obvious one is to increase source power, but this is not a practical solution in many cases. Not only is it more difficult and costly to increase laser power (there are thermal issues, for example), but once specific laser-output levels are exceeded at various wavelengths, the FSO system is no longer “safe” to the human eye per various regulatory mandates. Therefore, this is an acceptable solution in many cases.

Q: So, what can be done?

A: The solutions usually involve either space and frequency (wavelength) diversity and sophisticated data encoding via various error correcting codes(ECC). All of these techniques are transparent to the end user. Some systems also include dynamic beam alignment to maintain proper aim.

Q: What is meant by space diversity here?

A: It’s a form of path diversity. The easiest way to provide this is to have a multiple-lens arrangement are the receiver (typically with four to six lenses arranged in a circle and spaced about 20-30 cm apart. This increases the chance that the received beam paths will be somewhat independent of each other, and so provides redundancy in the path and reduces the impact of one (or more) paths having high attenuation.

The “cost” of doing this is a larger enclosure, more lenses, multiple optical sensors, and some way of electrically combining the sensor outputs; some designs instead use a mirror arrangement to optically combine the revied beams and direct them a single optical sensor.

Q: Is that the only space-diversity approach?

A: No. Some more costly and complex designs instead use multiple optical sources (lasers) working in parallel, also arranged to create multiple optical paths. These signals are then combined at the revied by one or more lens arrangements. This approach is more costly because optical sources are more costly than optical sensors.

Q: What about wavelength diversity?

A: Wavelength diversity is the same as frequency diversity used in RF systems, except that designers think and work with wavelength as the unit rather than the frequency in the optical world. Doing so is just a matter of a simple, well-known numerical conversion since wavelength λ × frequency = c (speed of light) (Figure 4).

Fig 4: RF engineers work and think in terms of both frequency and wavelength, while optical engineers use wavelength almost exclusively; they are linked by a simple formula using the speed of light (c) in vacuum or air. (Image: Science Notes)

To provide wavelength diversity, multiple laser sources are used with a spread of wavelengths. A typical system will use two, three, or even four different wavelengths. Since the attenuation factors such as rain, fog, dust, and others are a function of wavelength, using this diversity increases the signal strength and SNR at the receiver.

The subject of sources of attenuation, their detailed optical and physical mechanisms, and their severity is a major topic and has been extensively studied, analyzed, measured, and documented.

Q: How do you decide which mitigation approach to use?

A: It’s done by modeling the many relevant factors such as distance, allowed maximum power, and receiver sensitivity. Also, local conditions must be factored in, including expected atmospheric conditions (dry versus humid), expected wind and turbulence, air cleanliness or dirt, and any data associated with these environmental issues.

This part has looked at issues related to a practical terrestrial FSO link. The final part looks at a family of commercially available systems, including dynamic beam alignment to overcome vibration issues.

Related EE World Content

Is Li-Fi “To Be” or “Not To Be”? Part 3 – Status
Calgary signs on for Terabeam FSO gear
LightPointe receives FSO patent

References