Fiber Optics: Technology, Infrastructure, and Global Connectivity

Introduction and Historical Context

We live in an information age where broadband forms the foundation of economic growth and telecommunications is making its mark across the world. The internet, enabled by reliable communication infrastructure, drives e-commerce, education, and universal access to information. Rapidly increasing communication density and ever-expanding information flow have resulted in massive expansion of worldwide networks. The need for quick and reliable transmission and retrieval of increasing volumes of information over great distances is becoming more important every day. This paper explains the history, the fiber optic theory, and the future of worldwide fiber optic networks already in progress.

Many assume fiber optics is a technology invented in the current century, but historical examination reveals numerous concepts that led to its creation. The transition to optics coincided with the conversion from analog to digital transmission in the telephone network and with the growing necessity of computer data transmission.

Long before fiber optics was discovered, many signals were carried through copper coaxial cable. Invented by Oliver Heaviside in 1880, copper conductors were used primarily for analog signals such as high-frequency radio waves, cable television signals, and power lines (Nahin, 1988). Heaviside determined that wrapping an insulating layer around a transmission line both improved signal quality and increased the cable's physical strength. Later that year, he patented the first coaxial cable. In the early days of telephony, the general solution to increasing bandwidth was simply to add more copper cabling. The solution at the time was multiplexing, schemes that were applied to long-haul copper transmission as well as terrestrial microwave systems erected in the 1950s and 1960s. In the 1960s, the widespread use of the transistor produced smaller electronic equipment with lower power consumption that gained great popularity with the public.

The first successful intercontinental telecommunications cable was created in 1956. Transatlantic No. 1 (TAT-1) connected eastern Canada and Scotland. It took over a year to implement and carried 36 consecutive phone calls before its decommission in 1978 (Chadwick). The first fiber optic undersea cable was not implemented until 1988, by Bell Labs. They needed more bandwidth with less attenuation to compete with companies such as MCI, who were already competing for transatlantic and high-bandwidth communication (DeCusatis, 2008).

Fiber Optic Transmission Principles

In the Pacific region, TPC-5 was the first fiber cable to connect the United States and Japan. Its bandwidth speeds reached up to 5 Gb/s due to wavelength-division multiplexing. If it were not for the development of fiber optics, users would not be witnessing the telecommunications revolution experienced today. The networks that carry the bulk of data traffic across continents are all optical fiber networks. Fiber is particularly helpful for long-distance communications because light disseminates through the fiber with little loss, unlike coaxial cables.

Hayes (2006) defines optical fiber as follows: "Optical fiber is the medium in which communication signals are transmitted from one location to another in the form of light guided through thin fibers of glass or plastic. These signals are digital pulses or continuously modulated analog streams of light which represent information. These can be video information, computer information, data information, voice information, or any other type of information" (p. 15).

Fiber optic systems use light waves as the information carrier. In an analog sense, varying intensities of light represent varying current in the electrical system. Those varying currents could represent phenomena such as the human voice or a video signal. The idea of guiding beams of light is not new. As early as 1840, scientists were experimenting with the curious properties of light that allowed it to be carried in streams of water as well as glass rods bent to various shapes. However, these early experimenters realized that glass was not very clear, so light would not travel far in the material. Cladding fibers were not developed until 1951, preventing the leakage of light (Bells, 2011). In digital systems, light pulse intensities represent binary-encoded information. Light is then guided down the optical fiber core and stays there due to the difference in refractive indices between the fiber core and surrounding material.

Scientists use light waves that exhibit similar characteristics to other electromagnetic energy such as radio waves or microwaves. Using the Greek letter lambda (λ), one can indicate the wavelength of a particular signal as a cyclical variation in the energy associated with the electromagnetic wave. The wavelength of the color red, for example, is 700 nanometers (λ = 700 nm). Measurements slightly higher than this go beyond human eye perception, extending into infrared light. An easier means of discussing light speed is using the Index of Refraction, calculated by dividing the speed of light in a vacuum by the speed of light in another material:

n = (free space speed) / (medium speed)

n_glass = (300 million m/s) / (200 million m/s) = 1.5

The index is inversely related to speed. Most fiber transmission media uses glass or silica with an index of n = 1.5. Understanding how light is guided down the optical fiber requires examining refraction and reflection. Light will continue to travel in a straight direction without refraction or reflection. Refraction occurs when light rays bend when moving into a different index, such as from air to glass. Light bulbs, for example, radiate light in all directions; without proper light manipulation, the index does not change. Ray direction can be manipulated by changing the refractive index. A clear example is observing objects in water: when standing above the water, the refractive index change makes objects appear closer than they actually are.

Light waves are bent, or refracted, when entering a different medium. At an air-to-glass boundary, a component of energy is also reflected. In reflection, some energy bounces back when it hits an abrupt change in refractive index. All energy at this boundary must be accounted for in both refractive and reflective components. When building optical fiber, engineers choose how the refractive index is controlled in the core and cladding. The most common indexing strategy for multimode fiber is called graded-index (GI) glass. This design uses a continuous variation of refractive index at the fiber core with gradual transition to a refractive index equal to the cladding at the edge. This approach allows more uniform bending of light energy in the core, improving bandwidth efficiency in multimode fibers.

Transmitters, Receivers, and Light Sources

A fiber optic transmitter has one essential task: to encode the electrical signal into optical format. The method of encoding is called direct modulation. It uses a varying drive current in which data is encoded as light flashes, with light output bright and dim for binary sequences. Simple circuits are required to perform direct modulation.

At the receiving end, fiber optic receivers employ direct detection, which consists of two steps. First, the photodiode (PD) converts dim optical data into weak electrical current. Since the receiving signal is detected at its lowest intensity, amplifiers or repeaters are used to boost the electrical signal to usable strength. This technique is also employed in many other applications, such as cable satellite dishes feeding signals to televisions after the raw signal is received from the antenna.

Two types of light sources can be used within transmitter modules: laser diodes and light emitting diodes (LEDs). Laser diodes are used frequently with single-mode fiber. Although lasers can be used in multimode fiber, noise problems can occur. Challenges with lasers include fragility, expense, and the requirement for sophisticated drive circuitry. They offer the best performance for optical networks with cutting-edge specifications: bandwidth up to 10 Gb/s and power up to 10 milliwatts (Oliviero & Woodward, 2009). Beams produced by laser diodes diverge much less than LED beams, similar to comparing a spotlight to a floodlight. Since lasers require more complicated components, it is important that average output remains constant to directly modulate the beam.

LEDs appear more attractive to manufacturers due to performance-versus-cost considerations. LEDs are mostly used in applications where performance requirements are moderate: bandwidth up to 300 Mb/s and power up to 10 dBm. LEDs can only be used in multimode fiber because light launched in single-mode fiber is too weak to carry; the loss of energy exceeds about 30 dBm (Morris, 2007).

Undersea Cable Networks and Global Infrastructure

In a world that must communicate to grow, efficient transmission of voice and data signals is essential. The process of connecting information systems via fiber optics to overseas locations can be extensive, and undersea cables make this possible. For example, the company Alcatel-Lucent has developed a long reputation for providing telecommunications solutions to service providers and enterprises worldwide. Specialized vessels are equipped with spools supporting up to 4,000 kilometers of optical fiber. To minimize environmental disturbance and seismic disruptions, careful implementation is employed. If a ship were to drag its anchor across a section of seabed, millions of people could be either completely severed from communications or experience sluggish speeds. The cable's exact location on the ocean floor is extremely important because the ship must follow a carefully surveyed course.

Since the first fiber optic cable, TAT-8, it has been critical for vessels to embed cables under the ocean floor to prevent shark bites. TAT-8 had a bandwidth of 280 Mbps and operated at 1,300 nanometers (Fouchard, 2002). In 1989, AT&T and British Telecom privately funded a transatlantic cable called PTAT-1, which was more resistant to environmental vulnerabilities. Because of optical fiber's signal attenuation, repeaters are installed approximately every 50 kilometers to boost the optical signal. These repeaters must withstand the deep sea's intense water pressure and challenging terrain.

Widespread internet availability enables many new applications. However, the United States has fallen behind in high-speed internet deployment. Despite the Department of Defense's history of developing the internet, the U.S. is not perceived as internet-savvy. According to the internet monitoring company Akamai, broadband internet speeds in the United States are estimated at about 25 percent of South Korea's speeds, which leads the global market (Sutter, 2010). Some countries have committed to reducing latency to just a few milliseconds. Japan and the United Kingdom are planning to reduce latency to approximately 60 milliseconds. The Arctic Fibre and Arctic Link are expected to be routed through the Arctic islands of Canada, making their way to a main distribution frame in Tokyo. The completed undersea link is expected to cost between $600 million and $1.5 billion. Fiber optics is becoming increasingly focused on decreasing latency and increasing redundancy worldwide. The demand for real-time activity is critical for banks and stock markets, where a variation of a few milliseconds can increase or decrease billions of dollars. Latency reduction will also benefit other technologies that rely heavily on the internet, such as universities and hospitals (Anthony, 2012).

In western European countries and the U.S., telecommunications cables are mandated to go underground. Specific laws dictate where and how cables must be installed. In developed countries, it would be very expensive to change the entire infrastructure, which is why many developed nations still lack fiber optic connections. In Romania, for example, when high-speed internet emerged, there were no laws governing installation, so companies pulled wires above ground.

The company Google Inc. implemented a solution in Kansas City, Missouri by allowing the community to share high-speed internet at gigabit speeds. The goal was to find a location where the company could make an impact on the community and develop working relationships with local government, community organizations, and utility companies. However, this represents only the beginning for the average internet user (Malik, 2010).

Fiber-to-the-Home and Broadband Deployment

In the industry, the most popular internet service in the United States is Digital Subscriber Line (DSL). Most DSL services for consumers average from 256 kbit/s to 40 Mbit/s downstream. An idea was created to allow fiber to reach homes and businesses, called FTTX (Fiber-to-the-X). Since the long-distance and local loop cabling for companies are mostly fiber, FTTX completes the fiber telephone system. FTTX normally uses very high bit rate digital subscriber line (VDSL) at downstream rates of tens of megabits per second. As copper infrastructure continues to age, phone companies are realizing that alternatives to FTTX are inadequate for future bandwidth needs. DSL has already peaked due to aging copper infrastructure that cannot support higher-speed systems.

Running each fiber line directly from the provider would be impractical, yet several deployment methods exist for FTTX. An Active Star network would be one solution. In this approach, fiber is split from a centralized node with a drop cable leading to the customer. However, this still requires an active link to every home.

Another approach to reduce expenses is called curbing. Curbing involves running a local switch and using existing copper cabling to connect the customer via DSL. The problem is that the frequency capacity of copper cable is inversely proportional to link length: longer cables have lower frequency capacity. This limits the distance to approximately 15,000 feet. Although this method is inexpensive, it will not truly provide fiber-level speeds to customers.

The optimal architecture for FTTX is the Passive Optical Network (PON). In this method, an optical splitter divides the downstream signal from the central office to each subscriber up to 32 times (Keiser, 2006). The cost of downstream electronics is shared among all connected users. For upstream transmission, the system uses a single digital link with cheap components, keeping costs down. The passive optical network provides the broadband-to-home performance expected at the lowest possible cost. PON systems can use cascaded couplers, thus conserving cabling. The optical line terminal (OLT) in the central office connects to a fiber distribution hub (FDH) that houses the optical splitter serving one campus. A drop cable then goes to an optical network terminal (ONT) at the subscriber's location.

Like all telecommunications systems, extensive standards govern PON architectures. Broadband Passive Optical Network (BPON) uses digital systems for voice and data but analog for video, similar to cable television. If desired, the system can upgrade the analog video signal to IPTV. Gigabit PON and Ethernet PON work similarly to BPON but are fully digital. Even the cable television industry has its own fiber service version: Radio Frequency over Glass (RFOG), a local version of a hybrid fiber-coax system with cable modem service.