Overview

In this chapter, we introduce a number of rules related to the influence of noise in the performance of optical communication systems. In this discussion, we must make a distinction between the definition of signal-to-noise ratio (SNR) as commonly used is the electrical engineering context and the analogous definition for optical systems.

In electrical circuits, especially ones that deal with communications (as opposed to conveying power), the designer must evaluate the system to know if the signals are going to be detectable. The most common way to do this is to estimate the signal level and some statistically derived measure of the noise. The SNR is then simply the ratio of two numbers: the mean signal power and the standard deviation of the noise power. For completeness, we mention that noise is expected to have a zero mean value, a Gaussian distribution, and a randomness that allows the total variance in a system to be computed by adding the variance of each of the noise terms. Remember that variance is the square of the standard deviation. Keep in mind that this definition applies to electrical signals as might be found in wires, amplifiers, and other electrical circuits. As a reminder, the SNR in electrical systems is defined by the power of the signal and the statistics of the noise.

Now, consider the case of interest in optical communications; we need to be confident that sufficient light representing the signal is present, compared with noise present in the fiber...