Signal Processing - March 2017 - 14

special reports
John Edwards

Stepping Up Security with Signal Processing
Innovative tools and approaches address threats on multiple fronts

ignal processing is helping to make
life a little less scary. Cyberattacks,
street crime, terrorist strikes, and
other types of public threats have the
potential to devastate lives, destroy businesses, and even bring down governments. Finding ways to defuse potential
threats, as well as pinpoint perpetrators
before they can act, is important for
maintaining public security as well as
ensuring a better quality of life for people
worldwide. According to a recent report
issued by the Institute for Economics and Peace, a think tank based in Sydney, Australia, the global economic impact
of violence was US$13.6 trillion in 2015, or
13.3% of the world's gross domestic product.
With the ultimate goal of creating a
safer and more stable society, researchers worldwide are developing an array of
innovative security tools and approaches
that are rooted in or assisted by signal
processing. "Security is a wild, wild west
frontier where signal processing is one
of the tools for explorer," observes Yang
Cai, a senior systems scientist at Carnegie
Mellon University's CyLab Institute and
the director of CyLab's Visual Intelligence Studio.

multiple sources. To help website operators thwart such potentially paralyzing
attacks, a CyLab research team, led by
Prof. Cai, has developed a novel interactive visual analytics tool that scrutinizes
large volumes of network metadata, log
files, and security event logs to discover
hidden data and suspicious dynamic
behavioral patterns lurking within a large
network infrastructure (Figure 1).
Visual Analytics for Discovering
Dynamic Patterns in Network Data-
code-named Meteor-enables semantic
representations of cyberdynamic compo-

nents, such as packet and protocol sequences, attack behaviors, insider threats,
and sequential patterns, as well as access
points and cyber events. A three-dimensional (3-D) navigation interface enables
users to browse through and analyze a
large volume of cyberdata, including network traffic and historical data.
The researchers have already used
Meteor to study a typical malware distribution network (MDN), a connected
set of maliciously compromised domains
that's used to disseminate compromised
software to unknowing computers and

Visualizing cyberattacks
Distributed denial of service (DDoS)
attacks pose a major threat to websites
worldwide. A DDoS attack attempts
to make an online service unavailable
by overwhelming it with traffic from
Figure 1. Prof. Yang Cai (left) and research assistant Sebastian Peryt examine a visualization at the
Digital Object Identifier 10.1109/MSP.2016.2638478
Date of publication: 3 March 2017

Carnegie Mellon University CyLab Visual Intelligence Studio. (Photo courtesy of Carnegie-Mellon
University.)
IEEE Signal Processing Magazine

March 2017

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