Scribd
Upload
69 views13 pages

Lecture 11 - Discrete Time Systems (Slides)

This document summarizes a lecture on discrete time linear systems. It discusses how discrete time systems take a discrete input sequence and produce an output sequence through some internal operation. It describes linear and shift-invariant discrete systems. It provides examples of basic building blocks like scaling, addition, and delay. It discusses moving average filters and how they can be described by difference equations and transfer functions in the z-domain. It also introduces finite impulse response (FIR) filters and recursive filters, and compares their characteristics and responses.

Uploaded by

Buddhadev Ghosh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
69 views13 pages

Lecture 11 - Discrete Time Systems (Slides)

This document summarizes a lecture on discrete time linear systems. It discusses how discrete time systems take a discrete input sequence and produce an output sequence through some internal operation. It describes linear and shift-invariant discrete systems. It provides examples of basic building blocks like scaling, addition, and delay. It discusses moving average filters and how they can be described by difference equations and transfer functions in the z-domain. It also introduces finite impulse response (FIR) filters and recursive filters, and compares their characteristics and responses.

Uploaded by

Buddhadev Ghosh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Lecture 11

Discrete Time Systems

Prof Peter YK Cheung

Dyson School of Design Engineering

URL: www.ee.ic.ac.uk/pcheung/teaching/DE2_EE/
E-mail: [email protected]

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 1


Linear Discrete time systems
! A discrete time system takes in a sequence of discrete values x[n[ at the
input and produces an output sequence y[n] through some internal
operation or transformation T{.}

𝑥[𝑛] T{‧} 𝑦[𝑛] = 𝑇{ 𝑥 𝑛 }


𝑥[𝑛 − 𝑘] 𝑦[𝑛 − 𝑘]
! The system is LINEAR if it obeys the principle of superposition:
𝑦 𝑛 = 𝑇 𝑎& 𝑥& 𝑛 + 𝑎) 𝑥) 𝑛 = 𝑎& 𝑇 𝑥& 𝑛 + 𝑎) 𝑇 𝑥) 𝑛
! The system is shift-invariant if:

𝑦 𝑛 = 𝑇 𝑎& 𝑥1 𝑛 − 𝑘 + 𝑎) 𝑥) 𝑛 − 𝑘
= 𝑦[𝑛 − 𝑘]

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 2


Shift-invariant Discrete time systems
! Furthermore, a system is shift-invariant if delaying the input x[n] by k
samples results in the same output y[n], but delayed also by k.
𝑥[𝑛] 𝑦[𝑛] = 𝑇{ 𝑥 𝑛 }
T{‧}
𝑥[𝑛 − 𝑘] 𝑦[𝑛 − 𝑘]
! In this course, we only consider linear shift-invariant discrete systems.

T{‧}

T{‧}

T{‧}

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 3


Basic building blocks in a discrete linear system
! Scaling

𝑥[𝑛] 𝑦 𝑛 =𝛼𝑥 𝑛
a
! Adding
𝑥[𝑛] 𝑝 𝑛 = 𝑥 𝑛 + 𝑦[𝑛]
S

y[𝑛]
! Delay (i.e. Dk = time shift by k sample periods)

𝑥[𝑛] 𝑦 𝑛 =𝑥 𝑛−𝑘
Dk

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 4


Moving average filter
! Consider the following discrete time system: delay taps
𝑥[𝑛 − 1] 𝑥[𝑛 − 2] 𝑥[𝑛 − 3]

𝑥[𝑛] 𝑥[𝑛] D D D
F{.}
𝑦[𝑛]

0.25

0.25

0.25

0.25
signal flow diagram
difference equation
S
𝑦 𝑛 = 0.25 (x[n] + x[n-1]+x[n-2]+x[n-3])

! This system take the current and the previous 3 input samples, and average
them. This is also known as a moving average filter.

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 5


Example - COVID cases in UK

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 6


z-transform and difference equation
! According to Lecture 10 slide 9, if the z-transform of x[n] is X[z]:
?
𝑥 𝑛 → 𝑋[𝑧]
then,
?
𝑥 𝑛 − 𝑘 → 𝑋[𝑧] 𝑧 BC
! In other words, delaying a signal x[n] by k sample period is equivalent to
multiplying its z-transform X[n] with z-k.
! We can apply this important property of z-transform (known as the shift
property) to the difference equation relating the input sequence to the
output sequence:
𝑦 𝑛 = 0.25 (x[n] + x[n-1]+x[n-2]+x[n-3])
𝑌 𝑧 = 0.25 𝑋 𝑧 + 𝑋 𝑧 𝑧 B& 𝑧 + 𝑋 𝑧 𝑧 B) 𝑧 + 𝑋 𝑧 𝑧 BE
𝑌 𝑧 = 0.25(1 + 𝑧 B& + 𝑧 B) + 𝑧 BE )𝑋 𝑧
! This is the z-domain version of the difference equation in terms of z-k, where
k is delay in unit of sample.
PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 7
Transfer function in the z-domain
! Take the results from the previous slide and re-arrange:

𝑌 𝑧 = 0.25 𝑋 𝑧 + 𝑋 𝑧 𝑧 B& 𝑧 + 𝑋 𝑧 𝑧 B) 𝑧 + 𝑋 𝑧 𝑧 BE
𝑌 𝑧 = 0.25(1 + 𝑧 B& + 𝑧 B) + 𝑧 BE )𝑋 𝑧

𝐻 𝑧 = 𝑌 𝑧 /𝑋[𝑧] = 0.25(1 + 𝑧 B& + 𝑧 B) + 𝑧 BE )


! As in the case of Laplace transform, in the z-domain,
transfer function = output / input
! This moving average filter takes the average of the current data sample
x[i], and the previous three samples x[i-1], x[i-2] and x[i-3], to produce the
output y[i].
! The averaging function has a smoothing effect – that is, it performs the
function of a lowpass filter.

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 8


Frequency Response of this filter
! Here is the frequency response of this moving average filter:

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 9


General FIR filters
! Instead of using the same coefficient values in the moving average filter,
one could use different coefficients at different delay taps.
! The number of delay taps can be increased to N.
! This will implement a filter function of the form as difference equation:
𝑦 𝑛 = 𝑏I 𝑥 𝑛 + 𝑏& 𝑥 𝑛 − 1 + 𝑏& 𝑥 𝑛 − 2 + ⋯ + 𝑏KB& 𝑥 𝑛 − (𝑁 − 1)
! In z-domain form: KB&

𝑌 𝑧 = (𝑏I +𝑏& 𝑧 B& + 𝑏) 𝑧 B) + 𝑏E 𝑧 BE + ⋯ + 𝑏KB& 𝑧 B KB& )𝑋 𝑧 = 𝑋[𝑧] M 𝑏C 𝑧 BC


CNI
KB&
𝑌𝑧
𝐻𝑧 = = 𝑏I + 𝑏& 𝑧 B& + 𝑏) 𝑧 B) + 𝑏E 𝑧 BE + ⋯ + 𝑏KB& 𝑧 B KB&
= M 𝑏C 𝑧 BC
𝑋𝑧
CNI
! By choosing different coefficients b0, b1, b2 …., bN-1, one can implement
different types of filters: lowpass, bandpass, highpass etc.
! Such a filter will have N terms in the impulse response, where N is the
number of signal taps x[n], … x[n-(N-1)]. Therefore it is also known as a
finite impulse response filter (FIR) of order N.

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 10


Recursive Filter
! FIR filters derives the current output from current and previous inputs
! Such a filter does not make use of previous outputs – that is, it does not
rely on past information
! Recursive filter is different – it derives the current output from both input
and previous output samples.
! Here is one of the simplest recursive filter:

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 11


Step response of Recursive Filter
! Let us consider the response of this filter to a step input:

x[0:9] = [0 1 1 1 1 1 1 1 1 1]

y[0:9] = [0 0.2 0.36 0.49 0.59 0.67 0.74 0.79 0.83 0.87 ]

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 12


Frequency Response of Recursive Filter
! If we computer the magnitude response of this filter, we will get the
following characteristics:

PYKC 1 March 2021 DE2 – Electronics 2 Lecture 11 Slide 13

You might also like