Deploying mission-critical computer systems at locations where they can only be managed remotely is increasingly becoming a necessary requirement of real-world systems design. Thus there is a practical demand for the so-called "self-aware autonomous trusted" systems capable of
- monitoring their own state in order to distinguish "good" states
from inconsistent ones,
- automatically "fixing themselves", i.e., restoring their state to
a known good one when it become "bad" (exploited) or
inconsistent, and
- attesting their current state to a remote party, so that, e.g., a
remote administrator can be reasonably sure that he is observing
actual reports of the system's state rather than an elaborate
deception produced by a rootkit. It is important that this should
be achieved without destroying the general-purpose vendor-neutral
character of the platform.
Unfortunately, modern general-purpose computing platforms fall short in most of these requirements.
ATC-08 invites research papers and case studies that suggest innovative ways to design such systems. The full CFP is available at http://www.ux.uis.no/atc08/ATC08CFP.txt
