Adaptive Security and Separation in Reconfigurable Hardware
Award No:
0524771
Project Title:
CT-T: Collaborative Research: Adaptive
Security and Separation in Reconfigurable
Hardware
Investigators:
Cynthia Irvine (NPS)
Ryan Kastner (UCSB)
Tim Levin (NPS)
Thuy Nguyen (NPS)
Timothy Sherwood, PI (UCSB)
Institution:
University of California Santa Barbara (UCSB)
Naval Postgraduate School (NPS)
Website:
http://www.cs.ucsb.edu/~arch/RCsec
Description of Graphic Image:
The figure above illustrates the ubiquity of
reconfigurable hardware devices in everything
from aircraft to cell phones, yet little is known
about the security of these devices. The
impetus for smaller systems leads inevitably to
including different functions on the same chip.
If those different functions have different levels
of trust (for example the anti-lock brakes and
the MP3 player in your car) new methods are
required to ensure that they are kept separate.
Project Description and Outcome –
The RCsec project is developing the first method for composing secure systems on
reconfigurable hardware. The team from UCSB and NPS has extended the latest
advancements from hardware and system security research to provide a new approach for
secure data management using reconfigurable hardware. We address this problem with a set
of novel security primitives, a new architecture for establishing separation on the device, and
the complementary use of both static and dynamic techniques for ensuring the controlled
sharing between cores.
Blurring the line between software and hardware, reconfigurable hardware offers the raw high
speed of custom-built circuits and the flexibility of a general-purpose (e.g., desktop) processor.
The flexibility of reconfigurable hardware is a boon for embedded system developers, who can
now rapidly prototype and deploy solutions that include a variety of “soft cores” – which execute
programs from different third-party vendors – with performance approaching that of custom
silicon designs. However, the various cores may have divergent levels of trustworthiness and,
unlike traditional computers in which resources are managed by an operating system (such as
Windows), reconfigurable hardware typically gives each core fine grain control over the
underlying hardware resources. Given that reconfigurable hardware is already trusted to carry
out many critical tasks such as flight control and anti-lock braking, an adversary or rogue nation
could potentially craft a soft core to intercept or even interfere with the secure operation of the
device. Our results show how reconfigurable hardware can be built to provide a trustworthy
processing environment for a whole class of current and future computing systems.
Please select the Primary Strategic Outcome Goal that BEST DESCRIBES the highlight:
Foster research that will advance the frontiers of knowledge, emphasizing areas of greatest
opportunity and potential and establishing the nation as a global leader in fundamental
transformational science and engineering.
Secondary Strategic Outcome Goals that BEST DESCRIBES the highlight:
Learning: Cultivate a world-class, broadly inclusive science and engineering workforce, and
expand the scientific literacy of all citizens.
This work is notable because:
This work is the first to address the security of systems built with reconfigurable hardware.
Does this Highlight Represent Transformative or Multidisciplinary Research?
This work is high risk since it is the first time the security of reconfigurable systems has been
carefully considered (1).
Most information assurance professionals lack the hardware
background required to analyze the security of these reconfigurable systems, while most
embedded system designers treat security as an afterthought. Our research has the potential
to transform the way a large, and growing, class of systems is engineered (3). We place a large
deal of trust in these systems, and a failure due to an attack would have disastrous
consequences (4).