Random Processes and Linear Systems

Abstract

The language of signal coding is primarily that of random processes and linear systems. Although signal coding systems are inherently nonlinear, the tools and techniques of linear systems play an important part in their analysis. In a mathematical model of a communication system in general and a signal coding system in particular, it is usually assumed that the data to be communicated (or stored, compressed, processed,...) is produced by a random process. Where this is not the case, there would be no need to communicate the signal---it would already be known to the potential user. In addition to the original signal, the system itself may have random components such as the addition of random noise or the introduction of digital errors in a random fashion.