Home > Articles > Networking > Wireless/High Speed/Optical

  • Print
  • + Share This
This chapter is from the book

This chapter is from the book

The Communication System

Any communication system has three basic parts—a transmitter, a receiver, and a channel. The channel can be a hard-wired channel, like a transmission line, or it can be a soft-wired channel, like free space. Optical communication systems are no different. Light carries the information in an optical communication system. Optical communication systems use both types of channels. One popular system that uses a soft-wired system is the remote. Whether it is a garage door opener, a ceiling fan remote, or a TV/VCR remote, all use free space as the channel. What is common for all three? It is that each uses a relatively short channel length over which it sends optical information. Contrast that with sending an optical message under the Atlantic Ocean or with sending a message (phone call) cross-country. Sending an optical signal cross-country by using free space is impractical. The atmosphere heavily attenuates a high frequency (light wave) signal. In addition dust, rain, and other airborne particles prevent the use of free space as a practical channel. So the fiber cable is the most practical choice as the channel for sending optical messages. Figure 1-8 shows a typical fiber optic system.

Figure 1-8 Typical Fiber Optic Communication System.

The message contains the information; the source emits the carrier. In our example of smoke signal communication, the fire is the source emitting the (smoke) carrier. The message is the information sent by the interruptions caused by the blanket, cloth, or whatever. The interruptions modulate the smoke trail. Without the interruptions there is no message. A lone trail of smoke carries no information (only that there is a fire). There must be modulation. There is digital and analog modulation. For example, a TV remote transmits digital modulation. A high frequency carrier signal is modulated by a series of 1s and 0s. Figure 1-9(a) is an example of digital modulation. The carrier is a high frequency signal.

Figure 1-9 Digital and Analog Modulation.

Digital information is impressed on the carrier. The message is 10100. Analog modulation occurs when a high frequency carrier is modulated by an analog message, as shown in Figure 1-9(b). The message is low frequency compared to the carrier. For fiber optics the carrier frequency is light. The source is a light emitting diode (LED) or a laser diode. You might ask, "What kind of light do they emit?" Good question! The light depends on the material the diode is made from. For example, a GaInP diode will emit red light at a wavelength around 660 nm. An AlGaAs laser diode will emit around 700 nm. For a digital modulation the input current turns the LED on but for analog modulation a bias current keeps the LED on at all times. In both cases the optical power generated by the LED is proportional to the input current, as shown in Figure 1-10.

Figure 1-10 LED Optical Output Power versus Input Current Curve for Both Digital and Analog Modulation.

Both digital and analog modulations are shown on the same plot. Note that the input digital current turns the LED on but that the LED is on already when the analog input current is applied. That is, a dc bias current, Idc must apply at all times in the analog modulation case. Otherwise, the negative-going input current would be cut off by the LED.

We next compare the laser diode with the LED. We note that the optical power versus input current is not linear as with the LED, as seen in Figure 1-11.

Figure 1-11 Laser Diode Optical Output Power versus Input Current Curve for Both Digital and Analog Modulation.

Therefore, there needs to be a dc bias current, Idc, for both types of modulation. This bias current is near the knee of the curve for digital modulation, and the bias current is higher for the analog. It is important that the analog modulation operate on the linear portion of the curve to avoid harmonic distortion. As in Figure 1-10, both inputs are graphically projected to the right, showing the output power verses time plots. In summary, both LEDs and laser diodes are used for digital and analog modulation.

Next, let's check out detectors—light detectors.

The principles of light detection or photodetection are simply light energy striking an electronic component and causing the electric current to flow. The electronic component can be a vacuum electron tube like a photomultiplier tube or a semiconductor device like a photodiode. Properties of interest are rise time, response, and frequency response. The rise time is defined as the time it takes for output of the device to go from (for a step input) 10% to 90% of the final value. This is shown in Figure 1-12.

Figure 1-12 Rise Time of a Device.

The rise time is related to the frequency response of the device. Actually, the devices act like low pass filters. Because of the junction and parasitic capacitances, the high frequencies are attenuated. The low pass filter has a magnitude frequency response of the form


where is the peak message power in watts, is the peak message current in amperes, and is the carrier lifetime. Equation (1.26) is in normalized form. When , , which is the same as . So is called the frequency. Since power is proportional to the square of the current, we can relate the ratio to power. In other words, difference in current is and in power is ½. Thus,

where is the rise time in seconds. The rise time is limited by the high frequency performance of the electronic component.

Responsivity of a detector is defined as the ratio of output to input. It can be written



where v and i are the output voltage and current, respectively, and P0 is the average input optical power. So there is voltage responsivity and there is current responsivity.

The third characteristic is the frequency response of the device. It should be clear that responsivity is a function of frequency. The intent is to use the detector where the losses are minimal.

Detectors can be vacuum tube type or semconductor type. The size of the semiconductor diode makes it more suitable for fiber systems. Avalanche photo-diodes, PIN, and pn semiconductors are three examples of semiconductor diodes. Let's examine the last one first. The pn junction diode is relatively slow (the rise time is long), and the responsivity is poor compared to the avalanche and PIN diodes. The avalanche and PIN diodes use the same materials and so their frequency responses are the same and we discuss them together.

To get started: a typical pn semiconductor diode has a region where the p material and n material are joined. This is called a depletion region and is illustrated in Figure 1-13.

Figure 1-13 Typical Depletion Region.

There are essentially no free charges here and the resistance is high. When reverse biased, the depletion region increases or the potential energy barrier between p and n regions increases. An incoming photon will give enough energy to free an electron from the valence band to the conduction band, causing current to flow. As a light detector the semiconductor pn junction is reverse biased. Incoming photons generate electron-hole pairs, causing current to flow. The disadvantage of the pn diode is the slow rise time to an applied optic step input. The rise time is in the order of microseconds. Figure 1-14 shows a PIN diode where p and n regions are separated by an intrinsic semiconductor material, usually silicon. (Hence the name PIN.)

Figure 1-14 PIN Diode.

Since the intrinsic region is wide, the probability that the incoming photon will be absorbed there is higher than in the pn diode. This decreases the rise time and increases efficiency. Shown in Figure 1-15 is a typical current-voltage characteristic for a PIN diode. Naturally, as temperature increases, so does dark current.

Figure 1-15 Typical PIN Characteristic Curve.

Another device used as a detector is the avalanche photodiode (APD). The APD is similar to a PIN diode but has better responsivity. Additionally, it has an internal gain, though the rise time is similar to that of the PIN diode. Table 1-4 compares PIN and APD for two materials. The rise times are close, but the responsivities and gains are quite different.

Table 1-4 Comparison of PIN and APD Diodes



Rise Time [ns]

Current Responsivity [A/W]






















In addition to LEDs (sources) and photodetectors (receivers) there may be a need for amplification of weak signals. We next discuss optical amplifiers.

Optical signals experience loss when certain activities take place, for example, losses in splitters or couplers or simply loss of signal over long fibers. Under those conditions it is necessary that the optical signal be amplified. The higher frequencies of (optical) amplification give it an advantage of greater bandwidth. In addition, wavelength division multiplexing (WDM) is possible. (WDM is the same as frequency division multiplexing [FDM]. At optical frequencies, the label WDM is preferred.)

Optical amplification is done with semiconductor and fiber amplifiers. The amplifier can be located near the source, in front of the receiver, or somewhere along the transmission line, depending on the requirements. In the first case, if the light source level needed boosting, the amplifier would be located just after the light source. This is a power amplifier. In the second case, the signal-to-noise ratio may be low by the time it reaches the receiver. The amplifier would be located in front of the receiver. This is a preamplifier. In the third case, lengthy transmission lines between light source and receiver will attenuate the optical signal. An in-line amplifier—or more than one—would be placed to boost the optical signal. These are called line amplifiers.

The first type of amplifier to be discussed is the semiconductor optical amplifier. These amplifiers are actually lasers that operate below oscillation. They have gain, but not enough to oscillate. They can amplify a wave inserted into the medium, but sometimes they will unintentionally amplify another optical signal when more than one channel exists on the line. This is called crosstalk. Another disadvantage of semiconductor optical amplifiers is the high noise figure that is the ratio of signal to noise at the input to signal to noise at the output of the amplifier. Some advantages are wide bandwidth, high gain, and their smaller size that makes them easily integrable with other devices. A model of a semiconductor amplifier is shown in Figure 1-16. The cavity could be a Fabry-Perot laser cavity.

Figure 1-16 Semiconductor Optical Amplifier.

The external current pump injects current into the cavity. The current excites hole-electron pairs to amplify the optical signal.

The second type of amplifier we discuss is the fiber optic amplifier. In particular, the erbium-doped fiber amplifier (EDFA) is a popular fiber amplifier. A typical EDFA is shown in Figure 1-17.

Figure 1-17 Erbium-Doped Fiber Amplifier.

The fiber is a couple of meters in length. The two isolators suppress reflections. The EDFA is stimulated through a higher optical frequency pump. The pump excites fiber additives that amplify optical signals entering the fiber. The advantages of an EDFA are a low noise figure, a high efficiency from pump to amplified signal, suitability for long transmission lengths, and polarization independence. Some disadvantages are crosstalk, relatively large, spontaneous light emission (with no input), and gain saturation. Summarizing key characteristics of optical amplifiers, we are interested in gain, gain efficiency, bandwidth, saturation, and noise. Gain is defined as the ratio of output power to input power. Gain efficiency rates gain as a function of input power. Bandwidth is the range of frequencies at which we will use the amplifier. We have said little about saturation, but it is the point after which gain does not increase even though we increase input power. Finally, noise is any unwanted signal that interferes with our receiving the desired signal.

That brings us to the purpose of this book—to study the applications of "the desired signal." The signal is the optical signal. Light has always fascinated people—scientists and nonscientists alike. Electric storms, rainbows, sunsets, shining stars, colorful animals and vegetation, and many other attractive natural phenomena have evoked human responses of awe and worship. We have utilized natural phenomena and our scientific minds to research and develop outstanding technologies.

How about infrared binoculars? That is really not a new technology. What is new is the relatively low cost and light weight. The old binoculars are heavy and costly. They require cooling to cryogenic temperatures that lead to limited battery life. Of course, they are heavy as a result. Prototype infrared binoculars that use a microbolometer array to replace the cooled detectors are being developed. The microbolometer sensor array needs no cooling.

What about new solar cells? A major detriment to use of solar cells for space communication is efficiency. New triple junction solar cells with an efficiency of 26% are being produced. Naturally, the higher efficiency increases satellite payload possibilities.

What about solar power for domestic use? Clearly, the cost of oil affects our interest in solar power. As the price of oil exceeds $30 per barrel, interest in alternative energy sources increases. The photovoltaic market is growing rapidly. Worldwide production of photovoltaics is expected to reach 1,200 MW a year by the year 2010. The demand will exceed production. There appears to be a shift from crystalline silicon (photovoltaic cells) to polycrystalline silicon and amorphous thin-film silicon. Other technologies being tried are cadmium telluride and copper indium gallium diselenide. Two even newer technologies for producing solar electricity are organic-dye-stabilized titanium dioxide and polymer photovoltaics.

Finally, whether light technology is old or new is secondary to the fact that for light technology the game is on!

  • + Share This
  • 🔖 Save To Your Account