Introduction to Digital Signal Processing
by Derrick Corea Technosoft Innovations, IncIntroduction :
Digital Signal Processing Applications is already part of our daily
lives. It is present in cell phones, in photographic cameras, televisions,
automobiles, computers, medical equipment, automation systems, control,
communication, etc. Few people, however, are aware of the exact size of the
science involved and the complex technology behind it all. For us users, it's
all very simple and easy to use. Just press a button.
In this sequence of
articles, I will cover this topic very superficially and explain some basic and
important topics about digital signal processing. I do not intend to present a
course on this subject. That would be very long and tedious. Anyone who really
wants to study this subject can consult the following book:
Discrete-Time Signal Processing -
Analog vs. Digital
All natural systems or processes are in essence analogous.
The development of electronics went the same way: analogical in its essence.
Thus, signal conditioning, filtering, and even mathematical operations were
implemented analogously. There were no computational resources that could be
used efficiently and economically viable for the implementation of more complex
equipment. FIGURE 1 illustrates an analog electronic circuit with some
complexity.
Like everything else that is natural, electronic components
are also subject to variations in environmental conditions such as temperature,
humidity, aging, pressure, chemical attack, corrosion, etc. It was necessary to
use complex compensation procedures to correct the errors resulting from these
variations.
Stock photo of 'Printed-circuit board- It is photographed by
close up'
With the creation of the first integrated circuits around the
early 1960s and the subsequent creation of silicon-integrated microprocessors
from 1971 onwards, the way was opened for thinking about digitizing real-world
signals and processing them digitally. The most obvious advantage of the
digitized signal is that this signal no longer suffers from the variations to
which the analog signals are subjected. You can reduce the analog interfaces to
a minimum, enough to convert the analog signal to digital accurately and let
the microprocessor handle and process that signal. The same is true in case you
need to exit to the real world, converting it to an analog signal. At the same
time, there was a still unexplored opportunity to develop technologies for
digitized signals.
High slide JS
Continuous
x Discrete
Analog signals are by nature continuous over time. They are
theoretically constituted by a sequence of infinite points. When we digitize
these signals, they go through a process known by sampling and are transformed
into a finite sequence of discrete points. This transformation is performed by
components known as Analog / Digital converters or simply A / D converters. The
transformation of the digital signals into analog is carried out by electronic
components known as Digital / Analog converters, or D / A converters. In FIGURE
3, an example of the process of transforming the continuous analog signal into
a discretized signal can be seen.
Continuo_ Discrete
FIGURE 3: Transformation of analog signal into discrete
The resulting effect due to the sampling of the signal is
visible: The discretized signal is not exactly the same as the analog signal.
In FIGURE 3 the discretized signal is a coarse reproduction of the original
signal. This has a number of consequences and implications. On the other hand,
turning a continuous signal into a discrete signal brings the advantages of
being able to tailor the signal size to be processed to the limited memory
capacity of the processors. The memory of the processor, however large, is
always finite, as is its processing power. Here we have pointed out some of the
problems that must be taken into account and resolved when designing digital
signal processing systems.
After all, what is Digital Signal Processing?
To the layman it may seem that digital signal processing is
only possible with the aid of DSPs, silicon integrated processors specially
developed for this function. This is an understandable mistake. DSPs are
processors with a hardware architecture developed to carry out typical
operations of the implementation of digital signal processing procedures. Many
manufacturers have adopted an architecture known as Harvard Architecture, where
there are separate data and addressing paths for data and instruction memories,
a much more modern concept than the traditional Von Neumann architecture. In
Harvard Architecture it is possible to address and access in a single
instruction more than one memory location simultaneously, in a single machine
cycle. Although they have a differentiated architecture, the DSPs can be used
as microcontrollers, so much so that there is pressure from some manufacturers
of these components, that these DSPs are also adopted for conventional
operations. Some have even been renamed DSCs (Digital Signal Controllers), i.e.
digital signal controllers.
The digital signal processing itself is a technology, a
science, a number of abstract concepts which results in application of
computational algorithms for performing specific operations on digital data.
Some examples of digital signal processing:
·
Digital filtering of signals;
·
Speech recognition and synthesis;
·
Treatment and recognition of patterns in images;
Digital radio and TV
Digital signal processing algorithms can be run on DSPs,
specially developed to optimize the performance of these algorithms and enable
real-time operations. However, in suitable applications and in suitable
applications, it is possible to perform digital signal processing in common
microprocessors with low performance or even offline in a personal computer,
table t, etc.
Sponsor Ads
Created on Aug 16th 2019 04:38. Viewed 290 times.