Introduction

In the preceding chapters we discussed the characteristics of optical fibers, optical sources, and optical detectors. These form the three basic units of any optical fiber communication system. In this chapter we discuss how these basic elements can be put together to build a simple point-to-point optical fiber communication link.

Let us assume that we need to transmit information between two points (see Figure 13.1). The separation between these points could range anywhere from less than a kilometer in the case of computer data links to several thousands of kilometers, for instance, in transoceanic links. In all such links, there would be a transmitter that could be either an LED or a laser diode, the transmission path consisting of optical fibers that could be either multimode or singlemode fibers, and the optical receiver that could be a PIN or APD followed by the detection electronics. The choice of the components would depend on the distance as well as on the bit rate. When the separation between the points is greater than about 50–100 km, then because of attenuation in the link or pulse dispersion, it becomes necessary to use regenerators that consist of a receiver–transmitter combination. These regenerators detect the pulse stream before the power becomes too low or the pulses become unresolvable and retime, reshape, and regenerate (and, hence, the name 3R repeater) a new set of optical pulses to be transmitted over the next part of the link. For links that are limited by loss rather than by dispersion, one can use optical amplifiers in place of regenerators.

Introduction

Although the major application of optical fibers has been in telecommunications, there is a growing application of optical fibers in sensing applications for measurement of various physical and chemical variables, including pressure, temperature, magnetic field, current, rotation, acceleration, displacement, chemical concentration, pH, and so forth. Such fiber optic sensors are finding applications in industrial process control, the electrical power industry, automobiles, and the defense sector.

One of the main advantages of a fiber optic sensor stems from the fact that optical fibers are purely dielectric and thus can be easily used in hazardous areas where conventional electrically powered sensors would not be safe. In addition, fiber optic sensors are immune to electromagnetic interference, have greater geometric versatility (i.e., they can be configured into a variety of arrangements to suit the application), and should have a very short response time. They can be multiplexed into various configurations and the information from various sensors can be transmitted over long distances by optical fibers. They can also be configured to provide spatially distributed measurements of external parameters.

In a typical optical fiber sensor, light from a source such as a laser diode or LED is guided by an optical fiber to the sensing region. Some property of the propagating light beam gets modulated by the external measurand such as pressure, temperature, magnetic field, and so forth. The modulated light beam is then sent via another (or the same) optical fiber for detection and processing.

Introduction

As discussed earlier, optical fiber communication systems are ultimately limited by either loss or dispersion. Loss leads to power levels of received signal that cannot be detected within tolerable errors, whereas dispersion leads to overlapping of adjacent pulses of light with resultant loss in resolution and, hence, information. Figure 14.1 shows a typical wavelength dependence of loss and dispersion of a single-mode fiber for both a CSF with zero dispersion around 1300 nm and a DSF with zero dispersion around 1550 nm. As evident from the figure, the effect of loss can be minimized by operating at the minimum loss wavelength around 1550 nm, whereas dispersion can be minimized by operating at the zero dispersion wavelength. Using DSFs has advantages of both minimum loss and zero dispersion. Figure 14.2 shows the maximum permissible unrepeatered length as a function of bit rate as determined by loss or by dispersion for both a CSF and a DSF (see Chapter 13 for a detailed discussion of this figure). The pulses are assumed to be Fourier transform limited. As can be seen from the figure, even at bit rates of 2.5 Gb/s, a system operating with CSFs is limited by loss rather than by dispersion. For DSF, loss-limited operation extends to almost 10 Gb/s.

In long-haul fiber optic communication systems, the effects of loss and pulse dispersion are normally overcome by using periodically spaced electronic repeaters. In these repeaters the input optical signal is first detected and converted to electrical signals.

There has always been a demand for increased capacity of transmission of information, and scientists and engineers continuously pursue technological routes for achieving this goal. The technological advances ever since the invention of the laser in 1960 have indeed revolutionized the area of telecommunication and networking. The availability of the laser, which is a coherent source of light waves, presented communication engineers with a suitable carrier wave capable of carrying enormously large amounts of information compared with radiowaves and microwaves. Although the dream of carrying millions of telephone (audio) or video channels through a single light beam is yet to be realized, the technology is slowly edging toward making this dream a reality.

A typical lightwave communication system consists of a lightwave transmitter, which is usually a semiconductor laser diode (emitting in the invisible infrared region of the optical spectrum) with associated electronics for modulating it with the signals; a transmission channel – namely, the optical fiber to carry the modulated light beam; and finally, a receiver, which consists of an optical detector and associated electronics for retrieving the signal (see Figure 1.1). The information – that is, the signal to be transmitted – is usually coded into a digital stream of light pulses by modulating the laser diode. These optical pulses then travel through the optical fiber in the form of guided waves and are received by the optical detector from which the signal is then decoded and retrieved.

Introduction

An optital detector is a device that converts light signals into electrical signals, which can then be amplified and processed. Such detectors are one of the most important components of an optical fiber communcation system and dictate the performance of a fiber optic communication link.

There are many different types of photodetectors such as photomultiplier tubes, vacuum photodiodes, pyroelectric detectors, and semiconductor photodiodes. Semiconductor photodiodes are the most commonly used detectors in optical fiber systems since they provide good performance, are compatible with optical fibers (being small in size), and are of relatively low cost. These photodiodes are made generally from semiconductors such as silicon or germanium or from compound semiconductors such as GaAs, InGaAs, etc.

In this chapter we briefly discuss the basic principle of operation of two commonly used photodiodes – namely, PIN (p-doped, intrinsic, and n-doped layers) diode and avalanche photodiodes (APD) – and study their important characteristics that are of particular relevance to optical fiber communication systems.

Principle of optical detection

The basic principle behind photodetection using semiconductors is optical absorption. When light is incident on a semiconductor, the light may or may not get absorbed depending on its wavelength. If the energy hv of a photon of the incident light beam is greater than the bandgap of the semiconductor, then it can be absorbed, leading to generation of e–h pairs (see Section 11.4.2). When an electric field is applied across the semiconducting material, the photogenerated e–h pairs are swept away, leading to a photo current in the external circuit.

Introduction

In any optical waveguide such as a planar waveguide or an optical fiber, guidance of light takes place through the phenomenon of total internal reflection. For this to happen, the guiding region has to have a refractive index larger than the surrounding regions. In a class of waveguides called leaky waveguides, the low index surrounding region has a finite thickness comparable to the penetration depth of the guided field, and beyond this distance the medium has an index equal to or greater than that of the guiding region. In such a case, the waves do not undergo total internal reflection and, thus, the reflection coefficient is less than unity. Such a phenomenon is known as frustrated total internal reflection (FTIR). Hence, in such waveguides, there are no perfectly guided modes. On the other hand, such waveguides have leaky modes that are characterized by a finite loss coefficient. Such leaky modes find applications in the realization of many devices such as in-line fiber polarizers (see Section 17.5.1). We also see in Section 24.5 that a bent waveguide is a leaky waveguide, and the loss due to bending can be understood as due to a leakage mechanism.

In Section 24.2 we discuss the leakage loss calculations using the ray picture, and in Section 24.3 we discuss the concept of quasimodes. In Section 24.4 we discuss the matrix method to numerically obtain leakage loss. Finally, in Section 24.5 we discuss the bending loss in optical waveguides.