Signal Processing - November 2017 - 15

to be solved with new signal processing
approaches and methodologies.
Whether signal processing will
remain a tool, per se, or whether it will
tend to integrate with social sciences
and human psychology, this is a different question, and I answer positively: we
already talk about in-body communications, where various nanometer-level
devices can be implanted in the human
body, and the human body (and mind)
will act as the transmission channel, the
catalyzer, or the receiver of various digitalized data. Digital signal processing will
still play a crucial role in such a scenario,
but can we talk about a "humanized" or
"perception-tuned" signal processing?
Another future trend is that, with the
advent of global or ubiquitous "Internet of
Everything," privacy and security threats
will increase many folds: a stalker from
abroad could gain full video access to a
remote home, or a remotely controlled
drone (big enough) could even kidnap or
injure a person. Again, signal processing mechanisms that are in use today for
solving such privacy and security threats
are probably not enough in tomorrow's
world, and new avenues of thinking need
to emerge.
Simona Lohan (elena-simona
.lohan@tut.fi) is an associate profes-
sor at the Tampere University of Tech--
nology, Finland.

Feng Liu
With the development
of mathematics and
computing techniq-ues,
the past few decades
have shown that signal
processing, more and
more, has been considered as a rel-atively independent discipline, which studies the processing of
waveforms or digits with mathematics. In
turn, the signal processing discipline has
also promoted the development of mathematics and computing techniques. As
the computing techniques further de--
velop, especially with graphic processing
units, signal processing techniques such
as deep neural networks become practical, which act more as "black boxes" but
effectively achieve our complex goal. The
future direction of signal processing

depends on the social need and the technical context.
Admittedly, mathematics plays
today, and will play in the near future, a
dominating and definitely positive role
in the development of signal processing, with its results illuminating its path.
However, it does not contradict with
the notion that signal processing will
advance with and inside the social need
and technical context in the long run.
Also, as to the old philosophy, a merited deed is influential but invisible like
rain in the early spring: "sneaks into the
tranquil night with the breeze and nurtures every spring life spontaneously
and silently."
Feng Liu (liuf@nankai.edu.cn) is a
lecturer at Nankai University, Tianjin,
China.

Lav R. Varshney

David A. Trejo Pizzo

Signal processing,
has seen much better
times in its past and
is continuing in the
present. More and
more, the future of
signal processing is
looking very bright and long-lasting in
time and in other dimensions-longlasting because any information in
nature and wherever it is available could
be looked at as a signal. Signal genesis
and its association with physical and
other systems is also always going to be
there. Therefore, one has to be optimistic about the future of signals as well as
signal processing.
Although signal processing could be
felt by many of us to be a mature topic, just
because it has been offered as a semester-wide course many times across the
globe, it is not the complete story. There
are potential areas for novel advances in
theory as well as practice. In fact, signal
processing has brought mathematics into
reality and connected it not only to physics but many other areas, for example, in
engineering, medicine, biology, and agriculture. There are challenges, say, in the
next 50 years, to provide signal processing solutions to industry problems. Some
of the future key issues would be to put
signals and systems together with other
domains, nonlinear analysis would pick
up, not only time but other dimensions

Resea rchers have
in-ten-sively investigated deep-learning algor ith ms for solving
challenging problems
in many areas such as
image classification,
speech recognition, signal processing,
and natural language processing. In my
job, I use deep-learning algorithms and
signal processing to address new challenges in energy efficiency. Understanding the behavior of consumers and the
trends in economy are essential for a
country-with signal processing we can
address this challenge better by taking
noise out of market signals and huge
databases with information about the
energy consumption. Policy makers
now will have a new tool that combines
deep learning and signal processing to
make better decisions. This means a
breakthrough for engineers, where they
become the next generation of policy
makers by solving societal problems.
Signal processing meets sociology by
listening to the signals and erasing the
outliers that are not clear with other
tools. This joint venture of social scientists and engineers is a key to make
smart cities a reality.
David A. Trejo Pizzo (dtrejopizzo@
ieee.org) is a professor, Universidad del
CEMA, Buenos Aires, Argentina.
IEEE SIGNAL PROCESSING MAGAZINE

November 2017

I believe it is folly to
doubt the creativity
a n d h a r d wo r k of
signal -processing re--
search--e rs: there are
surely new breakthroughs to come.
Whether due to new technologies leading
to new research questions, internal
knowledge shortcomings being filled, or
new scientific phenomena requiring
explanation, signal processing researchers will develop theories and techniques
to address the challenges of the future.
L av R . Va rshney (varshney@
illinois.edu) is an assistant professor
at the University of Illinois at UrbanaChampaign.

Ajay V. Deshmukh

Table of Contents for the Digital Edition of Signal Processing - November 2017