Why do we need to study DSP?
1. There is a lot of competition for top jobs after college
Digital Signal Processing
EC363
Prepared By
Engr. Alexander Cabatit
1 2
DSP Development
Digital Signal Processing is relatively new
developing for the past several years due
computer technology and integrated circuit (IC)
fabrication starting from:
1. MSI (medium scale integration)
2. LSI (Large scale integration)
3. VLSI (Very large scale integration) with
powerful, smaller, faster and cheaper digital
computers and special purpose digital
hardware.
3 4
Definitions
1. Signal – A single valued function of time that conveys
information. Is defined as any physical quantity that varies
with time, space, or any other independent variable(s). As
a norm, a signal carries information and the objective of
signal processing is to extract this information.
2. Spectra – Describes the frequency content of the signal.
3. Signal Processing – is extracting information from a
signal and conditioning a signal for subsequent use,
signal transformation, or altering a signal structure.
4. Systems – is a process or method or plant that can
manipulate, change, or transmit signals. They operate on
signals to produce new signals or new signal
representation. As a whole, it may be defined as an
integrated unit composed of diverse, interacting
structures to perform a desired task. The main function of
a system is to process a given input sequence to
generate an output sequence.
5 6
graduation, and hands-on DSP experience will
differentiate you from other graduates.
2. DSP technology is used in cellular phones, hard disk
drives, motors, global positioning systems, modems,
wireless basestations, GPS-capable personal digital
assistants, network routers, hearing aids, anti-lock
brakes, and much more! DVD and digital cameras are
becoming increasingly popular.
3. DSP is the fastest-growing segment of the
semiconductor market.
4. The change from conventional analog Broadcasting to
Digital Broadcasting (HDTV)
Possibilities and Limitations
Inexpensive and relatively fast digital circuit have made it
possible to construct highly sophisticated digital systems
capable of performing complex digital signal processing
functions and task. Some signals with extremely wide
bandwidths or real-time processing requiring. Analog or
optical signal processing is the only possible solution for
such signals.
The subject DSP is limited to the signals that are being
processed from analog to digital and vice versa.
ANALOG DIGITAL
DSP (Signals)
Digital signal processing (DSP) -
• is the study of signals in a digital
representation and the processing
methods of these signals. DSP and analog
signal processing are subfields of signal
processing. DSP has three major
subfields: audio signal processing, digital
image processing and speech processing.
1
 Two types of Signal
Analog - A signal with amplitude that varies continuously for all
time (amplitude and time at their respective signals).
Analog signals are continuous time signals. v(t) has a value
to each instant of amplitude or value of t
Digital - Discrete time interval signals that is quantized and then
coded. Digital can be done by sampling a continuous time
signal at isolated, equally spaced points in time. The result
is a sequence of numbers that can be represented as a
function of an index variable that takes on only discrete
values. The signal remains to be an analog signal until it is
coded then it becomes a digital signal.
7 8
4. Digital implementation of a system allows easy adjustment
of the processor characteristics during processing. Allows
flexibility in reconfiguring the digital signal process operation
by simply changing the program. For analog usually implies
changing the hardware. Adjustment in the processor
characteristics can be easily done by periodically changing
the coefficients of the algorithm representing the processor
characteristics. These adjustments are necessary in
adaptive filters and allows implementation of more complex
signal processing algorithms. Reconfiguration of an analog
system requires redesign of hardware.
5. A major advantage of digital signals over analog is the use
of digital filters to produce linear phase characteristics and
multi-rate processing. Digital circuits can be connected in
cascade without loading problems, whereas this cannot be
done easily with analog circuits.
6. Storage of digital data is very easy. Signals can be stored in
various storage media without any loss, distortion, or loss of
signal fidelity. On the other hand, analog signals deteriorate
rapidly as time progress and cannot be recovered in their
original form.
9
Disadvantages of digital processing:
1. Digital processing needs pre and post processing devices
like analog to digital and digital to analog converters and
associated reconstruction filters. This increases the
complexity of the digital system and additional cost. But at
present, digital circuits are now cheaper and more reliable. In
many cases a digital implementation of the signal processing
system is cheaper than analog implementation.
2. Digital systems suffer from frequency limitations or speed of
application. For reconstructing a sampled signal the sampling
frequency must be at least twice the highest frequency
component present in the signal. The available frequency
range of operation for a digital signal processor is primarily
dependent on the sample and hold circuit and the analog to
digital converter and as a result limited by the technology
available.
Overall, the advantages out weight the disadvantages and as
time progress, the cost of DSP is decreasing
continuously.
11
Advantages of using digital over analog
1. Digital circuits do not rely on precise values of digital signals
for their operation. Less sensitive to changes in the
component values. Less sensitive to variations in
temperature, ageing and other external parameters
2. In a digital processor, the signal and systems coefficients
are represented in binary words. This enables one to
choose any accuracy by increasing or decreasing the
number of bits in the binary word. Tolerances in analog
circuit components make it difficult to control the accuracy
of an analog processing system. Digital signal processor
provides much better control of accuracy requirements.
Accuracy requirement in the A/D Converting.• Tolerance in
analog circuit components make it extremely difficult for the
system designer to control the accuracy of an analog
processing system
3. Digital Processing of a signal facilitates sharing of a single
processor among a number of signals by timesharing. This
reduces the processing cost per signal.
7. Storage of digital data is very easy. Signals can be stored in
various storage media without any loss, distortion, or loss of
signal fidelity. On the other hand, analog signals deteriorate
rapidly as time progress and cannot be recovered in their
original form.
8. For processing very low frequency signals like seismic
signals, analog require inductors and capacitors of a very
large size whereas, digital processing does not and is more
suited for such applications.
9. In some cases, digital implementation is much cheaper than
analog systems.
10. Digital processing hardware allows programmable
operations, so be can easily modify and flexibility in system
design. A digital programmable system allows flexibility in
reconfiguring the digital signal processing operations simply
by changing the program
11. Digital signal processing can allow more sophisticated
signal processing algorithms. It is difficult to perform precise
mathematical operations on signals in analog operations 10
Important Categories of DSP
Signal Analysis: Deals with the measurement of signal
properties. It is generally a frequency-domain operations.
Some of its applications: (1) Spectrum analysis, (2) Speech
recognition, (3) Speaker verification, (4) Target detection
Signal Filtering: This task is characterized by the “signal in-
signal out” situation. It is usually time domain operation. Some
of its applications: (1) Remove of unwanted background noise,
(2) Remove of interference, (3) Separation of frequency band,
(4) Shaping of the signal spectrum
12
2
 3. Continues time signal x(t)– is a mathematically continues
Classification of Signals function Periodic signals v(t) and that function is called
1. Multichannel – signals are generated by multiple sources or continues in the time domain.
multiple sensors and can be represented in vector form.
⎧ s1 (t ) ⎫
⎪ ⎪
f1 (t ) = ⎨s2 (t ) ⎬
⎪ s (t ) ⎪
⎩ 3 ⎭ 4. Discrete-time signal x(n)– is a signal specified only at
2. Multi-dimensional – A signal that is represented by a single specific time instants. The amplitude of the discrete-time signal
independent variable, the signal is called a one-dimensional between two time instants is no zero but is just not defined.
signal. M-dimensional if its value is a function of M The smallest value of N is called the fundamental period.
independent variables.
⎧ I r ( x, y , t ) ⎫
⎪ ⎪
V ( x, y , z ) = ⎨ I s ( x, y , t ) ⎬
⎪ I ( x, y , t ) ⎪
⎩ t ⎭
13 14

5. Deterministic signal - Is a signal that has certainty with respect to
its value at any time. They are functions that are completely
specified in time such that the nature and amplitude of the signal
can be predicted at any time. The pattern of the signal is regular
and can be characterized mathematically.
x(t) = α(t) – Ramp function
x(t) = A sin wt
⎧ 1, n≥0⎫
x (n) = ⎨ ⎬ 8.
⎩0, otherwise⎭
6. Non-deterministic Signals or random signals– is a signal
whose occurrence is random in nature and its pattern is quite
irregular. A typical example is thermal noise in an electrical
circuit. Or the number of accidents in a year. One cannot exactly
predict what would be the figure. Uncertainty of the value of the
signal before its actual occurrence.
7. Periodic Signals – A continues signal is said to be periodic
if it exhibits periodically,
x(t) = x(t + To), - ∞ < t < ∞
Where: t – denotes the signal location
To – Is a constant. Defines the duration of
one complete cycle V(t) or the period of the signal. The
smallest value of T that satisfies the equation. A
periodic signal has a definite pattern that repeats over
and over, with a repetition period of To
Non-periodic (Aperiodic signal) - Any signal for which
there is no value of To to satisfy the mathematical condition
stated for periodic signals.
Note that a sinusoid signals can be obtained by adding 2 equal
amplitude complex exponential signals called phasors.
• Positive frequency – counter clockwise angular motion
• Negative frequency – clockwise angular motion

15 16

9. Even Signals – If a signal exhibits symmetry in the time

Note: domain that is the signal must be identical to its
reflection about the origin. Mathematically, an even
signal satisfies the following:
the sum of two or more periodic continuous time signals need For continues time signals x(t) = x(-t)
not be periodic. They will be periodic if and only if the ratio of For a discrete signal x(n) = x(-n)
their fundamental periods is rational. In order to determine
whether the sum of two or more periodic signals is periodic 10. Odd Signals – The signal is not identical to its reflection
or not, the following steps may be used to test: about the origin, but to its negative. Exhibits anti-
1. Determine the fundamental period of the individual signals in symmetry. It satisfies the following conditions
the sum signal. mathematically:
2. Find the ratio of the fundamental period of the first signal For continues time signals x(t) = -x (-t)
with the fundamental periods of every other signal. For a discrete signal x(n) = -x (-n)
3. If all the ratios are rational, then the sum signal is also x1(t) sin ωt and x2(t) cos ωt are good examples of odd and
periodic.
even signals.
4. In the case of discrete time signals, the sum of the number
of periodic signals is always periodic because the ratio of
individual periods is always a ratio of integers, which is
rational.

17 18

3
11. Energy signals - Another classification of
signals are those that have finite energy or Digital Signals
finite average power. However, there are some
signals which can neither be considered as Steps in the process of creating digital
energy signals nor power signals. The Energy signals:
Signal is one which has finite energy and zero
1. Sampling – The process of taking
average power,
samples of the signals per sampling rate
ie x(t) is an energy signal if 0 < E < ∞.
12. Power Signal is one which has finite average 2. Quantizing – The process of assigning
power and infinite energy, ie 0 < P < ∞. If the steps or levels in the quantized signal
signal does not satisfy these 2 conditions, then 3. Coding – the process of assigning binary
it is neither an energy nor a power signal. codes to the quantized value.
19 20

Analog to digital conversion of signals

Description
A discrete time signal is defined by specifying its value only at 1. Sampler – extracts the sample values of the input
signal at the sample instants. The output of the sampler
discrete times called sampling instants. When sampling is the discrete time signal with continues amplitude.
values are quantised (assigned values) and encoded, a digital
2. Quantiser – converts the continues amplitude into a
signal is obtained. (Analog to digital converter - A/D) finite number of sample values. Each sample can be
represented by a digital word of finite word length.
3. Encoder – the final stage of AD assigns a digital word
Sampler Quantizer Encoder to each quantised sample.
To reconstruct the original signal it is enough to pass the
spectrum of x(t) and suppress the spectra of other
Continuous-time Discrete-time Discrete-time translated frequencies or simply filtering. Often referred
Digital output
continuous-amplitude
input signal
continuous-amplitude discrete-amplitude signal to as reconstruction filter. The output of the filter is
signal signal CoX(f) in the frequency domain and x(t0 in the time
domain.
Band limited – when a signal x(t) is present only for a
specific time or frequency only.
21 22

Sampling theorem Sampling error

Given a signal x(t), in order to recover the original signal • This overlap is known as aliasing or
from the samples, the sampling frequency must be
greater or equal to twice the maximum frequency in x(t). overlapping or foldover. Aliasing
The sampling theorem states that a bandlimited signal x(t) means that a signal can be impersonated
having no frequency components above the fundamental
fh frequency in hertz, is completely specified by samples by another signal. To prevent aliasing, the
that are taken at a n uniform rate greater than 2fh hertz. sampling frequency fs should be greater
The frequency equal to twice the highest frequency in
x(t) = 2fh is called the Nyquist rate. that twice the frequency fa of the
• Undersampling or down sampling. fs < 2fa . The sinusoidal being sampled. In some
results is overlapping of adjacent spectra. Higer applications anti-aliasing filter is placed
frequency components get superimposed on the lower
frequency components. before the sample and hold circuit.
23 24

4
 Signal quantising and Encoding Quantizing
A discrete time signal with continuous values of amplitude is
called sampled data signal while a continues time signal with
discrete values of amplitude is referred to as a quantised
boxcar signal.
Quantisation is the process by which the amplitude of each
sample of a signal is rounded off to the nearest possible level.
It is the conversion of a discrete time continuous amplitude
signal into a discrete time, discrete valued signal.
Encoding is done by representing each of these permissible
levels by a digital word of fixed word length.
Process of quantising produces an error called quantising error
and its simply the difference between the value of the analog
input and the analog equivalent of the digital representation.
The errors will be negligible if there are more levels and width
of these quantizing levels. For an AD, the quantiser is the only
source of error.
25 26

The End

Chapter 1

27 28