Fiber Optics Without Fiber

Beaming light through the air offers the speed of optics without the expense of fiber

12 min read
Fiber Optics Without Fiber

Mention optical communication and most people think of fiber optics. But light travels through air for a lot less money. So it is hardly a surprise that clever entrepreneurs and technologists are borrowing many of the devices and techniques developed for fiber-optic systems and applying them to what some call fiber-free optical communication.

Known within the industry as free-space optics (FSO), this form of delivering communications services has compelling economic advantages. Although it only recently, and rather suddenly, sprang into public awareness, free-space optics is not a new idea. It has roots that go back over 30 years--to the era before fiber-optic cable became the preferred transport medium for high-speed communication. In those days, the notion that FSO systems could provide high-speed connectivity over short distances seemed futuristic, to say the least. But research done at that time has made possible today's free-space optical systems, which can carry full-duplex (simultaneous bidirectional) data at gigabit-per-second rates over metropolitan distances of a few city blocks to a few kilometers.

These free-space systems require less than a fifth the capital outlay of comparable ground-based fiber-optic technologies. Moreover, they can be up and running much more quickly. Installing an FSO system can be done in a matter of days--even faster if the gear can be placed in offices behind windows instead of on rooftops. Using FSO, a service provider can be generating revenue while a fiber-based competitor is still seeking municipal approval to dig up a street to lay its cable.

The applications of free-space optics are many and varied. To cite just a few:

  • Metro network extensions. Carriers can deploy FSO to extend existing metropolitan-area fiber rings, to connect new networks, and, in their core infrastructure, to complete Sonet rings.

  • Last-mile access. FSO can be used in high-speed links that connect end-users with Internet service providers or other networks. It can also be used to bypass local-loop systems to provide businesses with high-speed connections.

  • Enterprise connectivity. The ease with which FSO links can be installed makes them a natural for interconnecting local-area network segments that are housed in buildings separated by public streets or other right-of-way property [photo].

  • Fiber backup. FSO may also be deployed in redundant links to back up fiber in place of a second fiber link.

  • Backhaul. FSO can be used to carry cellular telephone traffic from antenna towers back to facilities wired into the public switched telephone network.

  • Service acceleration. FSO can be also used to provide instant service to fiber-optic customers while their fiber infrastucture is being laid.

Until recently, the technology was used primarily for enterprise connectivity. It shows up mainly in local-area networks spanning multiple buildings, where right-of-way was an obstacle to leasing copper lines or fiber-optic cabling.

Over the past year or so, however, free-space optics has started to move into more mainstream service. Several free-space optics companies have begun field trials with telecommunications carriers in the United States, Europe, Asia, South America, and the Middle East.

With capital hard to come by and customers eager for high-speed data services, service providers are left in an interesting position. In the United States, an estimated 95 percent of buildings are within 1.5 km of fiber-optic infrastructure. But at present, they are unable to access it. Connecting them with fiber can cost US $100 000-$200 000/km in metropolitan areas, with 85 percent of the total figure tied to trenching and installation.

Street trenching and digging are not only expensive, they cause traffic jams (which increase air pollution), displace trees, and sometimes destroy historical areas. For such reasons, some cities, such as Washington, D.C., are considering a moratorium on fiber trenching. Others, like San Francisco, are hoping to limit disruptions by encouraging competing carriers to lay fiber within the same trench at the same time.

Things have reached this state because carriers spent billions of dollars to increase network capacity in the core, or backbone, of their networks, but have provided less lavishly at the network edges. Earlier this year, chairing the NGN Ventures Conference in Burlingame, Calif., John McQuillan, of McQuillan Ventures, Concord, Mass., a company that invests in network infrastructure companies, summed up the explosive growth of the Internet and network spending this way: "We keep thinking if we make the core of the network bigger and faster, it will solve the problem. But the real limits to growth and adoption of these new technologies come at the network edge, where people interact with the system." It is time to correct that imbalance, and free-space optics has a lot to contribute to the solution.

Simplified drawing of a single-beam LightPointe transceiver

A Single-Beam Link Head: Simplified drawing of a single-beam LightPointe transceiver shows that the received laser beam [tan] is much wider than the transmitted beam [red]. That's why the receiver lens is so much larger than the transmitter lens. Both lenses, which share the same axis, are mounted behind a glass housing with an embedded defroster for cold environments. Click on the image to enlarge.

Shooting its beam through an ordinary office window, this LightPointe optical transceiver joins two offices of the law firm Cadwalader, Wickersham & Taft located less than 100 meters apart in separate buildings in Manhattan's financial district. The 1.25-Gb/s link was installed early this year.

A free-as-air technology

Low-power infrared beams, which do not harm the eyes, are the means by which free-space optics technology transmits data through the air between transceivers, or link heads [see drawing], mounted on rooftops or behind windows [see photo]. It works over distances of several hundred meters to a few kilometers, depending upon atmospheric conditions.

Unlike most of the lower-frequency portion of the electromagnetic spectrum, the part above 300 GHz (which includes infrared) is unlicensed worldwide and does not require spectrum fees. The main limitation on its use is that the radiated power must not exceed the limits established by the International Electrotechnical Commission (Standard IEC60825-1) or, in the United States, the equivalent regulation promulgated by the Center for Devices and Radiological Health (CDRH), which is part of the Food and Drug Administration. As this is being written, by the way, IEC60825-1 is being amended, and the CDRH is expected to adopt the amended standard in the very near future, so there will be a single worldwide standard for these devices.

Commercially available free-space optics equipment provides data rates much higher than digital subscriber lines or coaxial cables can ever hope to offer. And systems even faster than the present range of 10 Mb/s to 1.25 Gb/s have been announced, though not yet delivered.

Generally the equipment works at one of two wavelengths: 850 nm or 1550 nm. Lasers for 850 nm are much less expensive (around $30 versus more than $1000) and are therefore favored for applications over moderate distances.

Given that price difference, why would a 1550-nm laser ever be chosen? The main reasons revolve around power, distance, and eye safety. Infrared radiation at 1550 nm tends not to reach the retina of the eye, being mostly absorbed by the cornea. Regulations accordingly allow these longer-wavelength beams to operate at higher power than the 850-nm beams, by about two orders of magnitude. That power increase can boost link lengths by a factor of at least five while maintaining adequate signal strength for proper link operation. Alternatively, it can boost data rate considerably over the same length of link. So for high data rates, long distances, poor propagation conditions (like fog), or combinations of those conditions, 1550 nm can become quite attractive.

As the differences in laser prices suggest, such systems are quite a bit more expensive than 850-nm links. An 850-nm transceiver can cost as little as $5000 (for a 10-100-Mb/s unit spanning a few hundred meters), while a 1550-nm unit can go for $50 000 (for gigabit-per-second setups encompassing a kilometer or two).

Various approaches

There is more than one way of designing free-space optics equipment. Some companies include switches and routers in their products; others offer a physical-layer-only solution. AirFiber and Optical Access, both of San Diego, Calif., have focused on mesh-based asynchronous transfer mode (ATM) and Internet protocol (IP) models, respectively. Protocol-independent physical-layer equipment that can be used in any network topology is the focus of LightPointe, the authors' company, also in San Diego; Optical Crossing, in Glendale, Calif.; and fSONA, in Richmond, B.C., in Canada. Terabeam Corp., in Kirkland, Wash., is unusual in being a provider of both equipment and communications services. Outside North America, a handful of other free-space optics companies in Europe and the Middle East primarily serve enterprise and campus customers.

Similarly, there are different approaches to dealing with the various factors that can affect link performance and reduce link availability below the five nines (99.999 percent) figure, the Holy Grail of carrier-class technology. By definition, carrier-class service delivers only one bad bit out of every 10 billion it carries, and statistically is out of service no more than 5 minutes and 15 seconds a year. For the rest of the 12-month period (8759 hours, 54 minutes, and 45 seconds), it will be up and running.

For free-space optics, challenges to achieving this level of performance take the shape of environmental phenomena that vary widely from one micrometeorological area to another. Included here are scintillation, scattering, beam spread, and beam wander.

Scintillation is best defined as the temporal and spatial variations in light intensity caused by atmospheric turbulence. Such turbulence is caused by wind and temperature gradients that create pockets of air with rapidly varying densities and therefore fast-changing indices of optical refraction. These air pockets act like prisms and lenses with time-varying properties. Their action is readily observed in the twinkling of stars in the night sky and the shimmering of the horizon on a hot day.

FSO communications systems deal with scintillation by sending the same information from several separate laser transmitters. These are mounted in the same housing, or link head, but separated from one another by distances of about 200 mm. It is unlikely that in traveling to the receiver, all the parallel beams will encounter the same pocket of turbulence since the scintillation pockets are usually quite small. Most probably, at least one of the beams will arrive at the target node with adequate strength to be properly received. This approach is called spatial diversity, because it exploits multiple regions of space. In addition, it is highly effective in overcoming any scintillation that may occur near windows. In conjunction with a design that uses multiple and spatially separated large-aperture receive lenses, this multi-beam approach is even more effective.

Dealing with fog, more formally known as Mie scattering, is largely a matter of boosting the transmitted power, although spatial diversity also helps to some extent. In areas with frequent heavy fogs, it is often necessary to choose 1550-nm lasers because of the higher power permitted at that wavelength. Also, there seems to be some evidence that Mie scattering is slightly lower at 1550 nm than at 850 nm. However, this assumption has recently been challenged, with some studies implying that scattering is independent of the wavelength under heavy fog conditions. Nevertheless, to ensure carrier-class availability for a single FSO link in most non-desert environments, the link length should be limited to 200-500 meters.

Free-space optics systems, when deployed through a network, can be engineered to provide the high availability desired by carriers. It's done by limiting link lengths in accordance with known local weather patterns. For example, LightPointe, which is in lab and field trials with more than a dozen carriers around the world, recently concluded a trial in Denver, Colo., in which its 1.25-Gb/s system obtained 99.997 percent availability over a three-month period at a challenging time of the year: November-February.

Other atmospheric disturbances, like snow and especially rain, are less of a problem for free-space optics than fog.

Swaying buildings

One of the more common difficulties that arises when deploying free-space optics links on tall buildings or towers is sway due to wind or seismic activity. Both storms and earthquakes can cause buildings to move enough to affect beam aiming. The problem can be dealt with in two complementary ways: through beam divergence and active tracking. The techniques are effective, as evidenced earlier this year during an earthquake in Seattle, where Terabeam has its free-space optical service up and running. The company reported that only a few of its links lost connections, and those for only a short time.

With beam divergence, the transmitted beam is purposely allowed to diverge, or spread, so that by the time it arrives at the receiving link head, it forms a fairly large optical cone. Depending on product design, the typical free-space optics light beam subtends an angle of 3-6 milliradians (10-20 minutes of arc) and will have a diameter of 3-6 meters after traveling 1 km. If the receiver is initially positioned at the center of the beam, divergence alone can deal with many perturbations. This inexpensive approach to maintaining system alignment has been used quite successfully by FSO vendors like LightPointe for several years now.

If, however, the link heads are mounted on the tops of extremely tall buildings or towers, an active tracking system may be called for. More sophisticated and costly than beam divergence, active tracking is based on movable mirrors that control the direction in which the beams are launched. A feedback mechanism continuously adjusts the mirrors so that the beams stay on target.

These closed-loop systems are also valuable for high-speed links that span long distances. In those applications, beam divergence is not a good approach. By its very nature, it reduces the beam power density just when receivers, being less sensitive at high data rates, need all the power they can get.

Beam wander arises when turbulent eddies bigger than the beam diameter cause slow, but large, displacements of the transmitted beam. It occurs not so much in cities as over deserts over long distances. When it does occur, however, the wandering beam can completely miss its target receiver. Like building sway, beam wander is readily handled by active tracking.

This office is a node on a metropolitan-area ring

Last-Mile Connectivity: Working via a hub building, free-space optics can connect each of the three buildings at the left to a competitive local exchange carrier's central office at 100-Mb/s. This office is a node on a metropolitan-area ring, which is connected to a regional ring by means of conventional fiber-optics equipment. Click on the image to enlarge.

Some case studies

In one free-space optics business case, a competitive local exchange carrier (CLEC) has an agreement with a large property management firm to provide all-optical 100-Mb/s Internet access capability to several buildings located in an office park. The carrier is building its network by leasing regional dark fiber rings and long-haul capacity from a wholesale fiber provider. It has identified a potential hub, or point-of-presence, less than a kilometer from the office park and within sight of one of its central offices. The CLEC currently has no fiber deployed to target customer buildings [see drawing].

When fiber was compared with free-space optics, deployment costs for service to the three buildings worked out to $396 500 versus $59 000, respectively. The fiber cost was calculated on a need for 1220 meters: 530 meters of trunk fiber from the CLEC's central office to its hub in the office park plus an average of 230 meters of feeder fiber for each of the runs from the hub to a target building, all at $325 per meter. Free-space optics is calculated as $18 000 for free-space optics equipment per building and $5000 for installation. Supposing a 15 percent annual revenue increase for future sales and customer acquisition, the internal rate of return for fiber over five years is 22 percent versus 196 percent for free-space optics.

In planning communications networks, much money can be saved by building the network piecemeal, adding to it as warranted by customer demand. Free-space optical networks lend themselves to such a scalable model much better than fiber-based networks do. With fiber, the cost of digging a trench is so high that it makes sense to install as much fiber as possible while the trench is open. With FSO, only the equipment absolutely needed at any time needs to be deployed. As new customers are signed up, the equipment needed to support them is installed. This demand-based approach lowers the capital expenditure required to grow the customer base and allows the service provided to immediately begin recovering costs associated with the network equipment capital outlay. In this scenario, not only the service provider but also the customer wins, because he or she can be instantly online and start to benefit from the higher bandwidth network connection.

All the same, while free-space optics advantageously bypasses the need to dig up streets for fiber-optic lines, its exposure to weather variations will remain its No. 1 challenge. In an outdoor lab trial in 2000, XO Communications, Reston, Va. (the broadband voice and data service provider formerly known as Nextlink), used free-space optics equipment to protect some of its fiber systems from accidentally severed cables. In addition to providing redundancy for ground-based fiber, the FSO equipment was used to close off a Sonet ring and to connect additional buildings to the ring. In this scenario, free-space optics was complementary to fiber-optic cable.

The same applications have been demonstrated pairing free-space optics with local multipoint distribution systems (LMDS) radio communications networks. According to the XO Communications trial, the diversity (in both transport medium and traversed path) that comes from backing up fiber with FSO may provide better protection than backing fiber up with additional fiber.

Clearly, then, FSO is not the ideal choice for all communications applications. Equally clearly, it has important roles to play both as a primary access medium and as a backup technology. Key to its success will be a realistic analysis of historical weather patterns in combination with customers' needs for network availability. With proper planning, path blocks like window washers and rooftop maintenance workers can also be dealt with, and the technology will be able to realize its great potential.

Driven by the need for high-speed local-loop connectivity and the costs and difficulties of deploying fiber, the interest in free-space optics has certainly picked up dramatically among service providers worldwide. The technology will likely migrate deeper into the network. The authors believe that FSO could be the ultimate solution for high-speed residential access. Instead of hybrid fiber-coax systems, hybrid fiber-laser systems may turn out to be the best way to deliver high data rates directly to the home. At that point, it will certainly be true to say that technology has caught up with the idea of providing high-capacity last-mile connectivity via free-space optics as envisioned at the end of the 1960s. But in the meantime free-space optics continues to accelerate the vision of all-optical networks cost effectively, reliably, and quickly with freedom and flexibility of deployment.

Michael J. Riezenman, Editor

About the Author

HEINZ A. WILLEBRAND AND BAKSHEESH GHUMAN are both with LightPointe Communications Inc., San Diego, Calif. Willebrand, founder and chief technology officer, holds several patents on the use of wavelength-division multiplexing technologies for extremely fast local-loop access. Ghuman is the chief marketing officer.

To Probe Further

The two dominant professional societies are probably the best overall source of information on all aspects of optics: SPIE, The International Society for Optical Engineering, whose Web site is at, and the Optical Society of America, at

Lightreading, at is devoted exclusively to following optical technology and companies.

White papers on laser eye safety and Sonet ring closure written by engineers from several companies are available through LightPointe's Web site:

An industry organization, The Free-Space Optical Alliance, has recently been formed, but as this issue goes to press, it has no Web site. For information on the alliance, send an e-mail to Maha Achour at


This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions