z-transform derived from Laplace transform

Consider a discrete-time signal x(t) below sampled every T sec.

Lecture 15 Discrete-Time System Analysis using z-Transform

x(t ) = x0 (t ) + x1 (t T ) + x2 (t 2T ) + x3 (t 3T ) + .....

(t T) e sT (L6S13)

The Laplace transform of x(t) is therefore (Time-shift prop. L6S13):

X ( s) = x0 + x1e sT + x2e s 2T + x3e s 3T + .....

Now define

z = esT = e( + j )T = e T cos T + je T sin T

Peter Cheung Department of Electrical & Electronic Engineering Imperial College London URL: www.ee.imperial.ac.uk/pcheung/teaching/ee2_signals E-mail: p.cheung@imperial.ac.uk
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 1 PYKC 3-Mar-11

X [ z ] = x0 + x1 z 1 + x2 z 2 + x3 z 3 + .....

L5.8 p560
E2.5 Signals & Linear Systems Lecture 15 Slide 2

z-1 the sample period delay operator

Laplace, Fourier and z-Tranforms

From Laplace time-shift property, we know that z = e sT is time advance by T second (T is the sampling period). 1 sT Therefore z = e corresponds to UNIT SAMPLE PERIOD DELAY. As a result, all sampled data (and discrete-time system) can be expressed in terms of the variable z. More formally, the unilateral z-transform of a causal sampled sequence: x[n] = x[0] + x[1] + x[2] + x[3] + is given by:

X ( s) = x(t )e st dt

Continuous-time system Converts integraldifferential equations to & signal analysis; stable or unstable algebraic equations Converts finite time signal to frequency domain Continuous-time; stable system, convergent signals only; best for steady-state

X ( ) = x(t )e jt dt

X [ z ] = x0 + x1 z 1 + x2 z 2 + x3 z 3 + ..... = x[n]z n
n =0

Discrete X [n ] = x[n]e jn0T Converts finite discrete- Discrete time, otherwise 0 time signal to discrete Fourier n =0 same as FT transform N 0 samples,T = sample period frequency domain

N0 1

0 = 2 / T

The bilateral z-transform for a general sampled sequence is:

z transform
L5.8 p560
Lecture 15 Slide 3 PYKC 3-Mar-11

X [ z] =

X [ z] =
PYKC 3-Mar-11

n =

x[n]z

n =

x[n]z

Converts difference equations into algebraic equations

Discrete-time system & signal analysis; stable or unstable

E2.5 Signals & Linear Systems

E2.5 Signals & Linear Systems

Lecture 15 Slide 4

Example of z-transform (1)

Example of z-transform (2)

Find the z-transform for the signal nu[n], where is a constant. By definition Since u[n] = 1 for all n 0 (step function),

Observe that a simple equation in z-domain results in an infinite sequence of samples. Observe also that exists only for . For X[z] may go to infinity. We call the region of z-plane where X[z] exists as Region-of-Convergence (ROC), and is shown below. z-plane

Apply the geometric progression formula:

Therefore:

L5.1 p496
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 5 PYKC 3-Mar-11 E2.5 Signals & Linear Systems

L5.1 p496
Lecture 15 Slide 6

z-transforms of [n] and u[n]

z-transforms of cosn u[n]

Remember that by definition:

Since From slide 5, we know Hence

Since

Also, for

Therefore

Therefore

L5.1 p499
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 7 PYKC 3-Mar-11 E2.5 Signals & Linear Systems

L5.1 p500
Lecture 15 Slide 8

z-transforms of 5 impulses

z-transform Table (1)

Find the z-tranform of: By definition,

Now remember the equation for sum of a power series:

r
k =0

r n+1 1 r 1

Let

r = z and n = 4 z 5 1 X [ z ] = 1 z 1 z = (1 z 5 ) z 1
L5.1 p500 L5.1 p498
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 10 E2.5 Signals & Linear Systems Lecture 15 Slide 9

PYKC 3-Mar-11

z-transform Table (2)

Inverse z-transform
As with other transforms, inverse z-transform is used to derive x[n] from X[z], and is formally defined as:

Here the symbol indicates an integration in counterclockwise direction around a closed path in the complex z-plane (known as contour integral). Such contour integral is difficult to evaluate (but could be done using Cauchys residue theorem), therefore we often use other techniques to obtain the inverse z-tranform. One such technique is to use the z-transform pair table shown in the last two slides with partial fraction.

L5.1 p498
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 11 PYKC 3-Mar-11 E2.5 Signals & Linear Systems

L5.1 p494
Lecture 15 Slide 12

Find inverse z-transform real unique poles

Find inverse z-transform repeat real poles (1)

Find the inverse z-transform of: Step 1: Divide both sides by z:

Find the inverse z-transform of: Divide both sides by z and expand:

Step 2: Perform partial fraction:

Use covering method to find k and a0:

Step 3: Multiply both sides by z:

We get:

Step 4: Obtain inverse z-transform of each term from table (#1 & #6):

L5.1-1 p501

To find a2, multiply both sides by z and let z:

PYKC 3-Mar-11

E2.5 Signals & Linear Systems

Lecture 15 Slide 13

PYKC 3-Mar-11

E2.5 Signals & Linear Systems

Lecture 15 Slide 14

Find inverse z-transform repeat real poles (2)

Find inverse z-transform complex poles (1)

To find a1, let z = 0:

Find inverse z-tranform of:

Therefore, we find:

Whenever we encounter complex pole, we need to use a special partial fraction method (called quadratic factors):

Use pairs #6 & #10

Now multiply both sides by z, and let z:

We get:

L5.1-1 p502
PYKC 3-Mar-11 E2.5 Signals & Linear Systems Lecture 15 Slide 15 PYKC 3-Mar-11 E2.5 Signals & Linear Systems

L5.1-1 p503
Lecture 15 Slide 16

Find inverse z-transform complex poles (2)

Find inverse z-transform long division

To find B, we let z=0:

Consider this example:

Now, we have X[z] in a convenient form:

Perform long division:

Use table pair #12c, we identify A = -2, B = 16,

and a = -3.

Thus:

Therefore:
L5.1-1 p504

Therefore
L5.1-1 p505

PYKC 3-Mar-11

E2.5 Signals & Linear Systems

Lecture 15 Slide 17

PYKC 3-Mar-11

E2.5 Signals & Linear Systems

Lecture 15 Slide 18