Sergio Verdu, an electrical engineer at Princeton University and a co-principal investigator, is trying to find the fundamental limits of compression and data transmission. The math is easy if engineers focus on signal-to-noise ratio, but that turns out to not be the best measure of fidelity in audio and video. The human brain is very tolerant of visual noise in general, but less so of particular types of noise; for instance, sharp edges are much more important to identifying an image than color. It’s not just a question of how many bits have to be received to deliver a comprehensible message, but which bits. “In lossy compression, the bottleneck has been to come up with measures of distortion that are relevant in the real world,” says Verdu.

It’s also difficult to find the limits of compression for structural information, says Szpankowski. Proteins behave differently and transmit different information depending on their shape, and scientists would like to build computer models that would help them better understand how structure contributes to cell development or the rise of diseases. But the math becomes complex, because structural elements such as edges and vertices can add several dimensions to an equation. So far, Szpankowski says, no theory exists to provide a metric for how much information is embedded in structure. There’s not even a good way to quantify complexity; we can say a car is more complex than a bicycle, but how much more?

Shannon theory works well for point-to-point transmission or in systems with several transmitters and one receiver. But once there are two transmitters and two receivers, the problem of crosstalk arises, where a receiver can pick up a signal from the wrong transmitter. With the growth in mesh networks and mobile communications, and even high-frequency transmissions turning old copper wires into miniature antennas, the point-to-point solution isn’t adequate. “We still don’t know what are the ultimate, fundamental limits to how much information we can send robustly in the presence of noise,” Verdu says.

Quantum information
A growing challenge for information
theory is the field of quantum infor-
mation and quantum computing.
Classical information has always been
understood as a collection of bits, says
Madhu Sudan, principal researcher at
Microsoft Research New England (now
on leave from Massachusetts Institute
of Technology) and a co-principal in-
vestigator. Quantum information, by
contrast, is a continuum; the qubits
used in quantum computing can have
a value of 1 and 0 and any of the infinite
possible values in between simultane-
ously. In such a circumstance, Sudan
asks, “What does it even mean to be in-

NSF
Science and
Technology
Centers

The U.s. national science Foundation (nsF) selected the Purdue University’s science of information Center as one of the science and Technology Centers (s TCs) to receive funding in 2010. The purpose of the nsF’s ambitious s TC program is to “support integrative partnerships that require large-scale, long-term funding to produce research and education of the highest quality.” in their quest to launch the next information revolution, the Purdue scientists will collaborate with colleagues at Bryn Mawr College; howard University; Massachusetts institute of Technology; Princeton; stanford; University of California, Berkeley; University of California, san diego; and University of illinois at Urbana-Champaign.

nsF funded four other s TCs. The Center for dark energy Biosphere investigations, headed by the University of southern California, will explore sub-surface life in deep mines and aquifers and below the ocean floor and how they influence global energy cycles. The Center for the study of evolution in action, based at the University of Michigan, will develop computer models to study complex biological questions that can’t be studied in nature. emergent Behaviors of integrated Cellular systems, led by Massachusetts institute of Technology, will try to engineer biological machines. and the Center for energy efficient electronics science, led by the University of California, Berkeley, will try to develop technology that can could eventually reduce power consumption in electronics by a millionfold.

Technology

Social
Media
Trends

The use of social networking is becoming more prevalent worldwide, with people from countries of varying economic development increasingly accessing the internet to participate in networking sites. in addition, cell phone ownership has increased significantly in 16 countries (for which trends are available) over the last three years, from a median of 45% in 2007 to 81% in 2010.

These are among the findings of a new survey by the Pew research Center’s global attitudes Project. The survey, “global Publics embrace social networking,” examined technology usage in 22 countries, with 24,790 people surveyed either face-to-face or by phone. social networking is especially widespread in the U.s., Pew says, with 46% of the U.s. survey respondents saying they use social networking sites. other top-ranking countries are Poland (43%), Britain (43%), and south Korea (40%).

Pew notes that while involvement in social networking is relatively low in less economically developed nations, this is largely due to the fact that many in those countries are unable to access the internet, rather than having a disinterest in social networking.

“in middle- and low-
income countries, when
people go online they also
tend to participate in social
networking,” says richard
Wike, associate director of the
Pew global attitudes Project.
“in places like Poland, russia,
and Brazil, the vast majority of
internet users also use social
networking sites. if you look at
the two sub-saharan african
nations we surveyed, nigeria
and Kenya, relatively few people
use the internet, but among
those who do, social networking
is very popular.”
For the most part, the study
shows, men and women tend
to engage in social networking
at about the same rates. The
U.s. is the only country in which
women (52%) are significantly
more likely than men (41%) to
use social networking.

—Bob Violino