This preview shows pages 1–4. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 9 Linear Systems with Random Inputs 9.1 Introduction 305 9.2 Overview of Linear Systems with Deterministic Inputs 305 9.3 Linear Systems with ContinuousTime Random Inputs 307 9.4 Linear Systems with DiscreteTime Random Inputs 313 9.5 Autoregressive Moving Average Process 316 9.6 Chapter Summary 323 9.7 Problems 323 9.1 Introduction In Chapter 8 the concept of random processes was introduced. Different param eters that are associated with random processes were also discussed. These pa rameters include autocorrelation function, autocovariance function, crosscorre lation function, crosscovariance function, and power spectral density. The goal of this chapter is to determine the response or output of a linear system when the input is a random signal instead of a deterministic signal. To set the stage for the discussion, we first begin with a brief review of linear systems with deterministic inputs. Then we examine the response of linear systems to random inputs. 9.2 Overview of Linear Systems with Deterministic Inputs Consider a system with a deterministic input signal x ( t ) and a deterministic re sponse y ( t ) . The system is usually represented either in terms of its impulse func 305 306 Chapter 9 Linear Systems with Random Inputs Figure 9.1 TimeDomain and FrequencyDomain Representation of Linear System tion (or impulse response ) h ( t ) or its system response H ( w ) , which is the Fourier transform of the impulse function. This is illustrated in Figure 9.1. The system response is also called the transfer function of the system. The system is defined to be linear if its response to a sum of inputs x k ( t ) , k = 1 , 2 ,..., K , is equal to the sum of the responses taken separately. Also, the system is said to be a timeinvariant system if the form of the impulse response does not depend on the time the impulse is applied. For linear timeinvariant systems, the response of the system to an input x ( t ) is the convolution of x ( t ) and h ( t ) . That is, y ( t ) = x ( t ) * h ( t ) = integraldisplay  x ( ) h ( t ) d where the last equation is called the convolution integral of x ( t ) and h ( t ) . If we define u = t , we see that y ( t ) = integraldisplay  x ( t u ) h ( u ) du = integraldisplay  h ( u ) x ( t u ) du = h ( t ) * x ( t ) Thus, the convolution equation can be written in one of two forms: y ( t ) = integraldisplay  x ( ) h ( t ) d = integraldisplay  h ( ) x ( t ) d In the frequencydomain, we can compute the Fourier transform of y ( t ) as fol lows: Y ( w ) = F [ y ( t ) ] = integraldisplay  y ( t ) e jwt dt = integraldisplay  braceleftbiggintegraldisplay  x ( ) h ( t ) d bracerightbigg e jwt dt 9.3 Linear Systems with ContinuousTime Random Inputs 307 where F [ y ( t ) ] denotes the Fourier transform of y ( t ) . Interchanging the order of integration and since x ( ) does not depend on t , we obtain Y ( w ) = integraldisplay...
View
Full
Document
This note was uploaded on 01/05/2010 for the course STAT 350 taught by Professor Carlton during the Fall '07 term at Cal Poly.
 Fall '07
 Carlton

Click to edit the document details