Chapt 9b - 9.4 Linear Systems with Discrete-Time Random...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 9.4 Linear Systems with Discrete-Time Random Inputs 315 = ∞ summationdisplay j =−∞ ∞ summationdisplay l =−∞ h [ j ] h [ l ] R XX [ n − j , n + k − l ] If X [ n ] is a wide-sense stationary discrete sequence, then we obtain R YY [ n , n + k ] = ∞ summationdisplay j =−∞ ∞ summationdisplay l =−∞ h [ j ] h [ l ] R XX [ k + j − l ] = R YY [ k ] In a manner similar to the continuous-time case, it can be shown that the power spectral density of Y [ n ] is given by S YY (Omega1) = | H (Omega1) | 2 S XX (Omega1) Example 9.5 The impulse response of a discrete linear time-invariant system is given by h [ n ] = a n u [ n ] where | α | < 1, and u [ n ] is the unit step sequence defined by u [ n ] = braceleftBig 1 n ≥ n < If the input sequence X [ n ] is a discrete-time white noise with power spectral den- sity N / 2, find the power spectral density of the output Y [ n ] . Solution The system response is given by H (Omega1) = ∞ summationdisplay n =−∞ h [ n ] e − j Omega1 n = ∞ summationdisplay n = a n e − j Omega1 n = 1 1 − ae − j Omega1 Since S XX (Omega1) = N / 2, we have that S YY (Omega1) = | H (Omega1) | 2 S XX (Omega1) = H ∗ (Omega1) H (Omega1) S XX (Omega1) = parenleftbigg 1 1 − ae j Omega1 parenrightbiggparenleftbigg 1 1 − ae − j Omega1 parenrightbigg N 2 = N 2 { 1 − ae − j Omega1 − ae j Omega1 + a 2 } = N 2 { 1 − a [ e − j Omega1 + e j Omega1 ] + a 2 } = N 2 { 1 − 2 a ( [ e − j Omega1 + e j Omega1 ] / 2 ) + a 2 } = N 2 { 1 − 2 a cos (Omega1) + a 2 } trianglesolid 316 Chapter 9 Linear Systems with Random Inputs 9.5 Autoregressive Moving Average Process The autoregressive moving average (ARMA) process is frequently used in time series analysis. It consists of two parts: the moving average process and the au- toregressive process. Let { W [ n ] , n ≥ } be a wide-sense stationary random input sequence with zero-mean and variance σ 2 W . In general the W [ n ] are assumed to be uncorrelated. An example of such a sequence is noise. The different processes are defined as follows. 9.5.1 Moving Average Process A moving average process of order q is a process whose current value Y [ n ] de- pends linearly on the q past values of the random input process. Thus, given a set of constants β , β 1 ,..., β q , the output process defined by Y [ n ] = β W [ n ] + β 1 W [ n − 1 ] + β 2 W [ n − 2 ] + ··· + β q W [ n − q ] = q summationdisplay k = β k W [ n − k ] n ≥ is called a moving average process of order q , MA( q ). The moving average process is a special case of the purely feedforward system called the finite im- pulse response (FIR) system with input X [ n ] that has nonzero mean. The general structure of FIR systems is shown in Figure 9.4, where D indicates a unit delay....
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.

Page1 / 10

Chapt 9b - 9.4 Linear Systems with Discrete-Time Random...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online