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.