IEEE Network - March / April 2017 - page 4

IEEE Network • March/April 2017
n several of my earlier Editor’s Notes, I mentioned security
to be a key aspect of various network-related topics. In
this editorial, I decided to focus on security, such a critical
yet sometimes not duly addressed issue in the networking
With the proliferation of mobile broadband technologies,
connectivity to the Internet has increased at an exponential
rate. At the same time, network security threats have also
increased. The common way for academic researchers and
industry-based practitioners to address these network secu-
rity threats is to borrow and extend solutions from wired
scenarios. This creates a significant problem. For example,
the IPsec, which is sufficiently secure against many known
malicious attacks, may be cumbersome for wireless net-
works, and therefore not used in radio access networks. Sim-
ilarly, security measures designed for wired network systems
that include authentication, intrusion detection, prevention,
reporting, security event management, and so forth, may not
necessarily be applicable to wireless networks. Identifying
this basic yet critical point is key to designing effective securi-
ty solutions for next generation networks.
Most of the security-related researchers in network-
ing deal with state-of-the-art intrusion detection systems,
primarily focusing on denial-of-service attack detection
and so forth. Such intrusion detection systems do have an
impact on network performance such as increased delay,
degraded throughput, and other QoS parameters. How-
ever, traditionally, network security has been considered
separately from QoS. While several attempts have been
made to jointly consider security and QoS issues, they
need to be carefully revisited in next-generation networks
such as 4G and beyond 4G. For delay-sensitive, real-time
communication, could security related operations have
any impact on network performance? Could the tradeoff
between security and QoS be theoretically modeled? For
practical networks, could hardware encryption be robust
yet fast enough to ignore such a tradeoff? Such questions
need to be taken into consideration to understand the
gravity of the problem.
In addition, we need to understand that many of the net-
work security threats are likely to happen because of the
potential misuse of cryptography applications. A secure cryp-
tographic algorithm is usually designed by making certain
assumptions. However, its implementation on a network
device, which often requires fast operations to minimize
delay, may overlook certain assumptions, leading to security
vulnerabilities. Therefore, a concerted effort between net-
work researchers, hardware manufacturers, and cryptogra-
phers is essential.
Many other security issues are expected to arise as 5G
begins to appear on the horizon. Researchers and practi-
tioners are cordially welcome to contribute their original
and creative ideas to identify the upcoming network security
issues and solve them.
Nei Kato
1,2,3 5,6,7,8,9,10,11,12,13,14,...100
Powered by FlippingBook