signals and systems - faculty of engineeringtania/teaching/sas 2017...ย ยท continuous-time signal...

Post on 10-May-2020

3 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Signals and Systems

Lecture 15

DR TANIA STATHAKIREADER (ASSOCIATE PROFESSOR) IN SIGNAL PROCESSINGIMPERIAL COLLEGE LONDON

โ€ข Consider a discrete-time signal ๐‘ฅ(๐‘ก) sampled every ๐‘‡ seconds.

๐‘ฅ ๐‘ก = ๐‘ฅ0๐›ฟ ๐‘ก + ๐‘ฅ1๐›ฟ ๐‘ก โˆ’ ๐‘‡ + ๐‘ฅ2๐›ฟ ๐‘ก โˆ’ 2๐‘‡ + ๐‘ฅ3๐›ฟ ๐‘ก โˆ’ 3๐‘‡ +โ‹ฏ

โ€ข Recall that in the Laplace domain we have:

โ„’ ๐›ฟ ๐‘ก = 1โ„’ ๐›ฟ ๐‘ก โˆ’ ๐‘‡ = ๐‘’โˆ’๐‘ ๐‘‡

โ€ข Therefore, the Laplace transform of ๐‘ฅ ๐‘ก is:

๐‘‹ ๐‘  = ๐‘ฅ0 + ๐‘ฅ1๐‘’โˆ’๐‘ ๐‘‡ + ๐‘ฅ2๐‘’

โˆ’๐‘ 2๐‘‡ + ๐‘ฅ3๐‘’โˆ’๐‘ 3๐‘‡ +โ‹ฏ

โ€ข Now define ๐‘ง = ๐‘’๐‘ ๐‘‡ = ๐‘’(๐œŽ+๐‘—๐œ”)๐‘‡ = ๐‘’๐œŽ๐‘‡cos๐œ”๐‘‡ + ๐‘—๐‘’๐œŽ๐‘‡sin๐œ”๐‘‡.

โ€ข Finally, define

๐‘‹[๐‘ง] = ๐‘ฅ0 + ๐‘ฅ1๐‘งโˆ’1 + ๐‘ฅ2๐‘ง

โˆ’2 + ๐‘ฅ3๐‘งโˆ’3 +โ‹ฏ

The z-transform derived from the Laplace transform.

โ€ข From the Laplace time-shift property, we know that ๐‘ง = ๐‘’๐‘ ๐‘‡ is time

advance by ๐‘‡ seconds (๐‘‡ is the sampling period).

โ€ข Therefore, ๐‘งโˆ’1 = ๐‘’โˆ’๐‘ ๐‘‡ corresponds to one sampling period delay.

โ€ข As a result, all sampled data (and discrete-time systems) can be

expressed in terms of the variable ๐‘ง.

โ€ข More formally, the unilateral ๐’› โˆ’ transform of a causal sampled

sequence:

๐‘ฅ ๐‘› = {๐‘ฅ 0 , ๐‘ฅ 1 , ๐‘ฅ 2 , ๐‘ฅ 3 , โ€ฆ }

is given by:

๐‘‹[๐‘ง] = ๐‘ฅ0 + ๐‘ฅ1๐‘งโˆ’1 + ๐‘ฅ2๐‘ง

โˆ’2 + ๐‘ฅ3๐‘งโˆ’3 +โ‹ฏ = ฯƒ๐‘›=0

โˆž ๐‘ฅ[๐‘›]๐‘งโˆ’๐‘›, ๐‘ฅ๐‘› = ๐‘ฅ[๐‘›]

โ€ข The bilateral ๐’› โˆ’transform for any sampled sequence is:

๐‘‹[๐‘ง] =

๐‘›=โˆ’โˆž

โˆž

๐‘ฅ[๐‘›]๐‘งโˆ’๐‘›

๐’›โˆ’๐Ÿ: the sampled period delay operator

Definition Purpose Suitable for

Laplace

transform ๐‘ฟ ๐’” = เถฑ

โˆ’โˆž

โˆž

๐’™(๐’•)๐’†โˆ’๐’”๐’•๐’…๐’•Converts integral-

differential

equations to

algebraic equations.

Continuous-time signal

and systems analysis.

Stable or unstable.

Fourier

transform ๐‘ฟ(๐Ž) = เถฑ

โˆ’โˆž

โˆž

๐’™(๐’•)๐’†โˆ’๐’‹๐Ž๐’•๐’…๐’•Converts finite

energy signals to

frequency domain

representation.

Continuous-time,

stable systems.

Convergent signals

only. Best for steady-

state.

Discrete

Fourier

transform

๐‘ฟ[๐’“๐Ž๐ŸŽ] =

ฯƒ๐’=โˆ’โˆž๐‘ต๐ŸŽโˆ’๐Ÿ ๐‘ป๐’™[๐’๐‘ป]๐’†โˆ’๐’‹๐’๐’“๐›€๐ŸŽ

๐‘‡ sampling period

ฮฉ0 = ๐œ”0 ๐‘‡ = 2๐œ‹/๐‘0

Converts discrete-

time signals to

discrete frequency

domain.

Discrete time signals.

๐‘ง โˆ’transform ๐‘ฟ[๐’›] =

๐’=โˆ’โˆž

โˆž

๐’™[๐’]๐’›โˆ’๐’Converts difference

equations into

algebraic equations.

Discrete-time system

and signal analysis;

stable or unstable.

Laplace, Fourier and ๐’› โˆ’ transforms

โ€ข Find the ๐‘ง โˆ’transform of the causal signal ๐›พ๐‘›๐‘ข[๐‘›], where ๐›พ is a constant.

โ€ข By definition:

๐‘‹[๐‘ง] =

๐‘›=โˆ’โˆž

โˆž

๐›พ๐‘›๐‘ข ๐‘› ๐‘งโˆ’๐‘› =

๐‘›=0

โˆž

๐›พ๐‘›๐‘งโˆ’๐‘› =

๐‘›=0

โˆž๐›พ

๐‘ง

๐‘›

= 1 +๐›พ

๐‘ง+

๐›พ

๐‘ง

2

+๐›พ

๐‘ง

3

+โ‹ฏ

โ€ข We apply the geometric progression formula:

1 + ๐‘ฅ + ๐‘ฅ2 + ๐‘ฅ3 +โ‹ฏ =1

1 โˆ’ ๐‘ฅ, ๐‘ฅ < 1

โ€ข Therefore,

๐‘‹[๐‘ง] =1

1โˆ’๐›พ

๐‘ง

, ๐›พ

๐‘ง< 1

=๐‘ง

๐‘งโˆ’๐›พ, ๐‘ง > ๐›พ

โ€ข We notice that the ๐‘ง โˆ’transform exists for certain values of ๐‘ง. These values

form the so called Region-Of-Convergence (ROC) of the transform.

Example: Find the ๐’› โˆ’transform of ๐’™ ๐’ = ๐œธ๐’๐’–[๐’]

โ€ข Observe that a simple equation in ๐‘ง-domain results in an infinite sequence

of samples.

โ€ข The figures below depict the signal in time (left) and the ROC, shown with

the shaded area, within the ๐‘ง โˆ’plane.

Example: Find the ๐’› โˆ’transform of ๐’™ ๐’ = ๐œธ๐’๐’–[๐’]cont.

โ€ข Find the ๐‘ง โˆ’transform of the anticausal signal โˆ’๐›พ๐‘›๐‘ข[โˆ’๐‘› โˆ’ 1], where ๐›พ is a

constant.

โ€ข By definition:

๐‘‹[๐‘ง] =

๐‘›=โˆ’โˆž

โˆž

โˆ’๐›พ๐‘›๐‘ข โˆ’๐‘› โˆ’ 1 ๐‘งโˆ’๐‘› =

๐‘›=โˆ’โˆž

โˆ’1

โˆ’๐›พ๐‘›๐‘งโˆ’๐‘› = โˆ’

๐‘›=1

โˆž

๐›พโˆ’๐‘›๐‘ง๐‘› = โˆ’

๐‘›=1

โˆž๐‘ง

๐›พ

๐‘›

= โˆ’๐‘ง

๐›พ

๐‘›=0

โˆž๐‘ง

๐›พ

๐‘›

= โˆ’๐‘ง

๐›พ1 +

๐‘ง

๐›พ+

๐‘ง

๐›พ

2

+๐‘ง

๐›พ

3

+โ‹ฏ

โ€ข Therefore,

๐‘‹[๐‘ง] = โˆ’๐‘ง

๐›พ

1

1โˆ’๐‘ง

๐›พ

, ๐‘ง

๐›พ< 1

=๐‘ง

๐‘งโˆ’๐›พ, ๐‘ง < ๐›พ

โ€ข We notice that the ๐‘ง โˆ’transform exists for certain values of ๐‘ง, which consist

the complement of the ROC of the function ๐›พ๐‘›๐‘ข[๐‘›] with respect to the

๐‘ง โˆ’plane.

Example: Find the ๐’› โˆ’transform of ๐’™ ๐’ = โˆ’๐œธ๐’๐’–[โˆ’๐’ โˆ’ ๐Ÿ]

โ€ข We proved that the following two functions:

โ–ช The causal function ๐›พ๐‘›๐‘ข ๐‘› and

โ–ช the anti-causal function โˆ’๐›พ๐‘›๐‘ข[โˆ’๐‘› โˆ’ 1] have:

โ– The same analytical expression for their ๐‘ง โˆ’transforms.

โ– Complementary ROCs. More specifically, the union of their ROCS

forms the entire ๐‘ง โˆ’plane.

โ€ข Observe that the ROC of ๐›พ๐‘›๐‘ข ๐‘› is ๐‘ง > ๐›พ .

โ€ข In case that ๐›พ๐‘›๐‘ข ๐‘› is part of a causal systemโ€™s impulse response, we see

that the condition ๐›พ < 1 must hold. This is because, since lim๐‘›โ†’โˆž

๐›พ ๐‘› = โˆž, for

๐›พ > 1, the system will be unstable in that case.

โ€ข Therefore, in causal systems, stability requires that the ROC of the systemโ€™s

transfer function includes the circle with radius 1 centred at origin within the

๐‘ง โˆ’plane. This is the so called unit circle.

Summary of previous examples

โ€ข By definition ๐›ฟ 0 = 1 and ๐›ฟ ๐‘› = 0 for ๐‘› โ‰  0.

๐‘‹[๐‘ง] =

๐‘›=โˆ’โˆž

โˆž

๐›ฟ ๐‘› ๐‘งโˆ’๐‘› = ๐›ฟ 0 ๐‘งโˆ’0 = 1

โ€ข By definition ๐‘ข ๐‘› = 1 for ๐‘› โ‰ฅ 0.

๐‘‹[๐‘ง] = ฯƒ๐‘›=โˆ’โˆžโˆž ๐‘ข ๐‘› ๐‘งโˆ’๐‘› = ฯƒ๐‘›=0

โˆž ๐‘งโˆ’๐‘› =1

1โˆ’1

๐‘ง

, 1

๐‘ง< 1

=๐‘ง

๐‘งโˆ’1, ๐‘ง > 1

Example: Find the ๐’› โˆ’transform of ๐œน[๐’] and ๐’–[๐’]

โ€ข We write cos๐›ฝ๐‘› =1

2๐‘’๐‘—๐›ฝ๐‘› + ๐‘’โˆ’๐‘—๐›ฝ๐‘› .

โ€ข From previous analysis we showed that:

๐›พ๐‘›๐‘ข[๐‘›] โ‡”๐‘ง

๐‘งโˆ’๐›พ, ๐‘ง > ๐›พ

โ€ข Hence,

๐‘’ยฑ๐‘—๐›ฝ๐‘›๐‘ข[๐‘›] โ‡”๐‘ง

๐‘งโˆ’๐‘’ยฑ๐‘—๐›ฝ, ๐‘ง > ๐‘’ยฑ๐‘—๐›ฝ = 1

โ€ข Therefore,

๐‘‹ ๐‘ง =1

2

๐‘ง

๐‘งโˆ’๐‘’๐‘—๐›ฝ+

๐‘ง

๐‘งโˆ’๐‘’โˆ’๐‘—๐›ฝ=

๐‘ง(๐‘งโˆ’cos๐›ฝ)

๐‘ง2โˆ’2๐‘งcos๐›ฝ+1, ๐‘ง > 1

Example: Find the ๐’› โˆ’transform of ๐œ๐จ๐ฌ๐œท๐’๐’–[๐’]

โ€ข Find the ๐‘ง โˆ’transform of the signal depicted in the figure.

โ€ข By definition:

๐‘‹ ๐‘ง = 1 +1

๐‘ง+1`

๐‘ง2+1

๐‘ง3+1

๐‘ง4=

๐‘˜=0

4

๐‘งโˆ’1 ๐‘˜ =1 โˆ’ ๐‘งโˆ’1 5

1 โˆ’ ๐‘งโˆ’1=

๐‘ง

๐‘ง โˆ’ 11 โˆ’ ๐‘งโˆ’5

๐’› โˆ’transform of 5 impulses

๐’› โˆ’transform Table

No. ๐’™[๐’] ๐‘ฟ[๐’›]

๐’› โˆ’transform Table

No. ๐’™[๐’] ๐‘ฟ[๐’›]

Inverse ๐’› โˆ’transform

โ€ข As with other transforms, inverse ๐‘ง โˆ’transform is used to derive ๐‘ฅ[๐‘›]from ๐‘‹[๐‘ง], and is formally defined as:

๐‘ฅ ๐‘› =1

2๐œ‹๐‘—เถป๐‘‹[๐‘ง]๐‘ง๐‘›โˆ’1๐‘‘๐‘ง

โ€ข Here the symbol ืฏ indicates an integration in counter-clockwise

direction around a closed path within the complex ๐‘ง-plane (known as

contour integral).

โ€ข Such contour integral is difficult to evaluate (but could be done using

Cauchyโ€™s residue theorem), therefore we often use other techniques to

obtain the inverse ๐‘ง โˆ’transform.

โ€ข One such technique is to use the ๐‘ง โˆ’transform pairs Table shown in the

last two slides with partial fraction expansion.

Find the inverse ๐’› โˆ’transform in the case of real unique poles

โ€ข Find the inverse ๐‘ง โˆ’transform of ๐‘‹ ๐‘ง =8๐‘งโˆ’19

(๐‘งโˆ’2)(๐‘โˆ’3)

Solution

๐‘‹[๐‘ง]

๐‘ง=

8๐‘ง โˆ’ 19

๐‘ง(๐‘ง โˆ’ 2)(๐‘ โˆ’ 3)=(โˆ’

196)

๐‘ง+

3/2

๐‘ง โˆ’ 2+

5/3

๐‘ง โˆ’ 3

๐‘‹ ๐‘ง = โˆ’19

6+

3

2

๐‘ง

๐‘งโˆ’2+

5

3

๐‘ง

๐‘งโˆ’3

By using the simple transforms that we derived previously we get:

๐‘ฅ ๐‘› = โˆ’19

6๐›ฟ[๐‘›] +

3

22๐‘› +

5

33๐‘› ๐‘ข[๐‘›]

Find the inverse ๐’› โˆ’transform in the case of real repeated poles

โ€ข Find the inverse ๐‘ง โˆ’transform of ๐‘‹ ๐‘ง =๐‘ง(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3

Solution

๐‘‹[๐‘ง]

๐‘ง=

(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3=

๐‘˜

๐‘งโˆ’1+

๐‘Ž0

(๐‘งโˆ’2)3+

๐‘Ž1

(๐‘งโˆ’2)2+

๐‘Ž2

(๐‘งโˆ’2)

โ–ช We use the so called covering method to find ๐‘˜ and ๐‘Ž0

๐‘˜ = เธญ(2๐‘ง2 โˆ’ 11๐‘ง + 12)

(๐‘ง โˆ’ 1)(๐‘ง โˆ’ 2)3๐‘ง=1

= โˆ’3

๐‘Ž0 = เธญ(2๐‘ง2 โˆ’ 11๐‘ง + 12)

(๐‘ง โˆ’ 1)(๐‘ง โˆ’ 2)3๐‘ง=2

= โˆ’2

The shaded areas above indicate that they are excluded from the entire

function when the specific value of ๐‘ง is applied.

Find the inverse ๐’› โˆ’transform in the case of real repeated poles cont.

โ€ข Find the inverse ๐‘ง โˆ’transform of ๐‘‹ ๐‘ง =๐‘ง(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3

Solution

๐‘‹[๐‘ง]

๐‘ง=

(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3=

โˆ’3

๐‘งโˆ’1+

โˆ’2

(๐‘งโˆ’2)3+

๐‘Ž1

(๐‘งโˆ’2)2+

๐‘Ž2

(๐‘งโˆ’2)

โ–ช To find ๐‘Ž2 we multiply both sides of the above equation with ๐‘ง and let

๐‘ง โ†’ โˆž.

0 = โˆ’3 โˆ’ 0 + 0 + ๐‘Ž2 โ‡’ ๐‘Ž2 = 3

โ–ช To find ๐‘Ž1 let ๐‘ง โ†’ 0.

12

8= 3 +

1

4+

๐‘Ž1

4โˆ’

3

2โ‡’ ๐‘Ž1 = โˆ’1

๐‘‹[๐‘ง]

๐‘ง=

(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3=

โˆ’3

๐‘งโˆ’1โˆ’

2

๐‘งโˆ’2 3 โˆ’1

(๐‘งโˆ’2)2+

3

(๐‘งโˆ’2)โ‡’

๐‘‹ ๐‘ง =โˆ’3๐‘ง

๐‘งโˆ’1โˆ’

2๐‘ง

๐‘งโˆ’2 3 โˆ’๐‘ง

(๐‘งโˆ’2)2+

3๐‘ง

(๐‘งโˆ’2)

Find the inverse ๐’› โˆ’transform in the case of real repeated poles cont.

๐‘‹ ๐‘ง =โˆ’3๐‘ง

๐‘งโˆ’1โˆ’

2๐‘ง

๐‘งโˆ’2 3 โˆ’๐‘ง

(๐‘งโˆ’2)2+

3๐‘ง

(๐‘งโˆ’2)

โ€ข We use the following properties:

โ–ช ๐›พ๐‘›๐‘ข[๐‘›] โ‡”๐‘ง

๐‘งโˆ’๐›พ

โ–ช๐‘›(๐‘›โˆ’1)(๐‘›โˆ’2)โ€ฆ(๐‘›โˆ’๐‘š+1)

๐›พ๐‘š๐‘š!๐›พ๐‘›๐‘ข ๐‘› โ‡”

๐‘ง

(๐‘งโˆ’๐›พ)๐‘š+1

[โˆ’2๐‘ง

๐‘งโˆ’2 3 = (โˆ’2)๐‘ง

๐‘งโˆ’2 2+1 โ‡” (โˆ’2)๐‘›(๐‘›โˆ’1)

222!๐›พ๐‘›๐‘ข ๐‘› = โˆ’2

๐‘›(๐‘›โˆ’1)

8โˆ™ 2๐‘›๐‘ข[๐‘›]

โ€ข Therefore,

๐‘ฅ ๐‘› = [โˆ’3 โˆ™ 1๐‘› โˆ’ 2๐‘›(๐‘›โˆ’1)

8โˆ™ 2๐‘› โˆ’

๐‘›

2โˆ™ 2๐‘› + 3 โˆ™ 2๐‘›]๐‘ข[๐‘›]

= โˆ’ 3 +1

4๐‘›2 + ๐‘› โˆ’ 12 2๐‘› ๐‘ข[๐‘›]

Find the inverse ๐’› โˆ’transform in the case of complex poles

โ€ข Find the inverse ๐‘ง โˆ’transform of ๐‘‹ ๐‘ง =2๐‘ง(3๐‘ง+17)

(๐‘งโˆ’1)(๐‘ง2โˆ’6๐‘ง+25)

Solution

๐‘‹ ๐‘ง =2๐‘ง(3๐‘ง + 17)

(๐‘ง โˆ’ 1)(๐‘ง โˆ’ 3 โˆ’ ๐‘—4)(๐‘ง โˆ’ 3 + ๐‘—4)

๐‘‹[๐‘ง]

๐‘ง=

(2๐‘ง2โˆ’11๐‘ง+12)

(๐‘งโˆ’1)(๐‘งโˆ’2)3=

๐‘˜

๐‘งโˆ’1+

๐‘Ž0

(๐‘งโˆ’2)3+

๐‘Ž1

(๐‘งโˆ’2)2+

๐‘Ž2

(๐‘งโˆ’2)

Whenever we encounter a complex pole we need to use a special partial

fraction method called quadratic factors method.

๐‘‹[๐‘ง]

๐‘ง=

2(3๐‘ง+17)

(๐‘งโˆ’1)(๐‘ง2โˆ’6๐‘ง+25)=

2

๐‘งโˆ’1+

๐ด๐‘ง+๐ต

๐‘ง2โˆ’6๐‘ง+25

We multiply both sides with ๐‘ง and let ๐‘ง โ†’ โˆž:

0 = 2 + ๐ด โ‡’ ๐ด = โˆ’2

Therefore,

2(3๐‘ง+17)

(๐‘งโˆ’1)(๐‘ง2โˆ’6๐‘ง+25)=

2

๐‘งโˆ’1+

โˆ’2๐‘ง+๐ต

๐‘ง2โˆ’6๐‘ง+25

Find the inverse ๐’› โˆ’transform in the case of complex poles cont.

2(3๐‘ง+17)

(๐‘งโˆ’1)(๐‘ง2โˆ’6๐‘ง+25)=

2

๐‘งโˆ’1+

โˆ’2๐‘ง+๐ต

๐‘ง2โˆ’6๐‘ง+25

To find ๐ต we let ๐‘ง = 0:

โˆ’34

25= โˆ’2 +

๐ต

25โ‡’ ๐ต = 16

๐‘‹[๐‘ง]

๐‘ง=

2

๐‘งโˆ’1+

โˆ’2๐‘ง+16

๐‘ง2โˆ’6๐‘ง+25โ‡’ ๐‘‹ ๐‘ง =

2๐‘ง

๐‘งโˆ’1+

๐‘ง(โˆ’2๐‘ง+16)

๐‘ง2โˆ’6๐‘ง+25

โ€ข We use the following property:

๐‘Ÿ ๐›พ ๐‘› cos ๐›ฝ๐‘› + ๐œƒ ๐‘ข[๐‘›] โ‡”๐‘ง(๐ด๐‘ง+๐ต)

๐‘ง2+2๐‘Ž๐‘ง+ ๐›พ 2 with ๐ด = โˆ’2, ๐ต = 16, ๐‘Ž = โˆ’3, ๐›พ = 5.

๐‘Ÿ =๐ด2 ๐›พ 2+๐ต2โˆ’2๐ด๐‘Ž๐ต

๐›พ 2โˆ’๐‘Ž2=

4โˆ™25+256โˆ’2โˆ™(โˆ’2)โˆ™(โˆ’3)โˆ™16

25โˆ’9= 3.2, ๐›ฝ = cosโˆ’1

โˆ’๐‘Ž

๐›พ= 0.927๐‘Ÿ๐‘Ž๐‘‘,

๐œƒ = tanโˆ’1๐ด๐‘Žโˆ’๐ต

๐ด ๐›พ 2โˆ’๐‘Ž2= โˆ’2.246๐‘Ÿ๐‘Ž๐‘‘.

Therefore, ๐‘ฅ ๐‘› = [2 + 3.2 cos 0.927๐‘› โˆ’ 2.246 ]๐‘ข[๐‘›]

โ€ข Since ๐‘ง = ๐‘’๐‘ ๐‘‡ = ๐‘’ ๐œŽ+๐‘—๐œ” ๐‘‡ = ๐‘’๐œŽ๐‘‡๐‘’๐‘—๐œ”๐‘‡ where ๐‘‡ =2๐œ‹

๐œ”๐‘ , we can map the

๐‘  โˆ’plane to the ๐‘ง โˆ’plane as below.

โ–ช For ๐œŽ = 0, ๐‘  = ๐‘—๐œ” and ๐‘ง = ๐‘’๐‘—๐œ”๐‘‡. Therefore, the imaginary axis of the

๐‘  โˆ’plane is mapped to the unit circle on the ๐‘ง โˆ’plane.

= Im(s)

Mapping from ๐‘  โˆ’plane to ๐‘ง โˆ’plane

Mapping from ๐‘  โˆ’plane to ๐‘ง โˆ’plane cont.

โ€ข For ๐œŽ < 0, ๐‘ง = ๐‘’๐œŽ๐‘‡ < 1. Therefore, the left half of the ๐‘  โˆ’plane is mapped

to the inner part of the unit circle on the ๐‘ง โˆ’plane (turquoise areas).

โ€ข Note that we normally use Cartesian coordinates for the ๐‘  โˆ’ plane

๐‘  = ๐œŽ + ๐‘—๐œ” and polar coordinates for the ๐‘ง โˆ’plane (๐‘ง = ๐‘Ÿ๐‘’๐‘—๐œ”).

Mapping from ๐‘  โˆ’plane to ๐‘ง โˆ’plane cont.

โ€ข For ๐œŽ > 0 , ๐‘ง = ๐‘’๐œŽ๐‘‡ > 1 . Therefore, the right half of the ๐‘  โˆ’plane is

mapped to the outer part of the unit circle on the ๐‘ง โˆ’plane (pink areas).

Find the inverse ๐’› โˆ’transform in the case of complex poles

โ€ข Using the results of todayโ€™s Lecture and also Lecture 9 on stability ofcausal continuous-time systems and the mapping from the ๐‘  โˆ’plane tothe ๐‘ง โˆ’plane, we can easily conclude that:

โ–ช A discrete-time LTI system is stable if and only if the ROC of itssystem function ๐ป(๐‘ง) includes the unit circle, ๐‘ง = 1.

โ–ช A causal discrete-time LTI system with rational ๐‘ง โˆ’transform ๐ป(๐‘ง) isstable if and only if all of the poles of ๐ป(๐‘ง) lie inside the unit circle โ€“i.e., they must all have magnitude smaller than 1. This statement isbased on the result of Slide 5.

Example: homework

โ€ข Consider a LTI system with input ๐‘ฅ[๐‘›] and output ๐‘ฆ[๐‘›] related with thedifference equation:

๐‘ฆ ๐‘› โˆ’ 2 โˆ’5

2๐‘ฆ ๐‘› โˆ’ 1 + ๐‘ฆ ๐‘› = ๐‘ฅ[๐‘›]

Determine the impulse response and its ๐‘ง โˆ’transform in the followingthree cases:

โ–ช The system is causal.

โ–ช The system is stable.

โ–ช The system is neither stable nor causal.

top related