IEEE Signal Processing - March 2018 - 4

from the editor
Robert W. Heath, Jr. | Editor-in-Chief | rheath@utexas.edu

Research Gems Found Digging with Industry

I

ndustry involvement is a significant
area of interest for the IEEE Signal
Processing Society (SPS). It is also
a frequent topic of conversation at the
SPS Board of Governors meetings. Ac-
cording to the SPS's website (https://
signalprocessingsociety.org), 42% of
our 19,000 members are from industry.
Yet the number of attendees from indus-
try of our flagship conference, the IEEE
International Conference on Acoustics,
Speech, and Signal Processing, is much
lower, based on my best estimate from
the sessions I attended.
IEEE Signal Processing Magazine (SPM) is "the voice to the outside
world," to quote Vice President of Pub-
lications Prof. Sergios Theodoridis.
No offense to the journals in the signal
processing portfolio, but I believe that
SPM offers the most accessible content
for all signal processing members. It
is a primary mission to bring tutorial
content through feature articles, special
issues, and columns. As a result, I think
SPM is a great way to foster collabora-
tion between academia and industry.
More industry involvement is a self-
ish endeavor on my part. To be blunt,
industry provides me and my students
with practical problems that need to be
solved. My students need to solve prob-
lems so they can publish papers and
write dissertations. I need to graduate
students with M.S. and Ph.D. degrees
to fill needs in academia, industry, and
Digital Object Identifier 10.1109/MSP.2018.2793165
Date of publication: 7 March 2018

4

skeptical about the potential novelty. My
government. With closer ties to indus-
student, though, was able to formulate a
try, the problems my students solve
nice problem and solve it in an interest-
can be relevant, the work impactful,
ing and mathematical way. After ten or
and the students better trained for their
20 revisions, the journal paper was in good
positions. When industry funds collab-
shape, so we submitted it. At the time, I put
orative research, this feedback loop can
it in the category of
be efficient.
I believe that SPM offers
"mathematically inter-
Sometimes in-
esting but probably not
dustry provides re-
the most accessible
relevant" research. As
search gems, which
content for all signal
look rough to the pro-
processing members. It is time went on and, yes,
after so many revisions,
fessor. Several years
a primary mission to bring I grew to appreciate the
ago, I had a student
tutorial content through
model and ideas. One
(Dr. Omar El Ayach,
feature articles, special
of the papers we wrote
now at Google) who
together even won a
did an internship with
issues, and columns.
paper award.
Samsung in Dallas,
I want to relay to you three important
Texas. While there, he worked on a prob-
lessons that were reinforced. The lesson
lem related to multiple-input, multi-
that all good lessons come in threes, ob-
ple-output (MIMO) communication in
tained from reading a book about TED
emerging millimeter-wave communi-
talks, is free.
cation systems. After his internship,
First, industry has a wealth of practical
we worked together with his Samsung
problems. Although I was dismissing
mentors (Dr. Sridhar Rajagopal, now at
the problem being solved as boring, it
Mavenir, Dr. Shadi Abu-Surra, now at
turned out to be one of the key challenges
PHAZR, and Dr. Zhouyue Pi, now at
in millimeter-wave communication. I
Straightpath) on some papers related to
should have talked more quickly with the
his internship. Though they did not fund
Samsung folks, to see their prototype and
his Ph.D. work, I asked Omar to contin-
alleviate my concerns that the problems
ue work on these problems to bootstrap
were not that interesting. Indeed, the main
my research interests in millimeter-wave
takeaway here may be to listen more and
signal processing.
talk less. This is also good advice for per-
I have to say I was not excited at first
sonal relationships.
about the core idea in these papers. It was
Second, internships are great. They
related to hardware constraints in millime-
provide a great opportunity for students
ter-wave MIMO systems, which were not
to understand how companies work.
that important in lower-frequency systems.
There was also a flavor of older array pro-
cessing from the 1990s, which made me
(continued on page 18)
IEEE Signal Processing Magazine

|

March 2018

|
https://www.signalprocessingsociety.org https://www.signalprocessingsociety.org
Table of Contents for the Digital Edition of IEEE Signal Processing - March 2018
