5
Random Processes
5.1 INTRODUCTION
‘Tais chapter introduces functions that depend both on the outcome of an experiment and
= time, For example, the function x(0) = a sin(wt + 6) depends on time as well as on
Se values ofa, o, and 6, which are constants for deterministic functions, However, a, 0,
=< 0 may also be specific values of the random variables A, 0, and ®, respectively.
Fesctions that depend on time as well as on one or more random variables serve as mod
<5 of signals that convey information in various systems. A deterministic signal is not a
‘Sod model forthe transmission of data, even though such a signal may vary with time,
Secause the future values of a deterministic signal are completely predictable from the
Seown past values as we discussed in Chapter 2. If information is to be conveyed, then
‘Se signal must not be known. It must not be deterministic, The signal must be, in some
“=<. unpredictable, For example, the signal might depend on an experimental outcome
sec time. Such signals are random in some sense. Thus, one purpose of this chapter is
& cefine a random process for which time is typically the independent variable, a-
‘Seuzh distance is often another independent variable in some applications, such as the
‘Sesitoring of seismic events. We will give numerous examples of random processes. In
‘S chapters that follow we will develop procedures for analyzing the response of linear
‘Sens for signal plus noise inputs. In addition, we will study some system design pro-
Ss that include maximizing the signal-to-noise ratio, estimating the signal arrival
S= cetermining the direction of arival of a signal, recognizing the presence of a par-
signal, making velocity measurements, and other applications,
RANDOM PROCESSES AND ENSEMBLES
-éeterministic functions we often encounter four types or classes of signals:
Definition 5.1. A continuous time and continuous amplitude function, also
137