Research Challenges in High Confidence Systems

Research Challenges in High Confidence Systems
Introduction

	Overview
	Background
	Workshop Organization
	Endnotes

Overview

On August 6-7, 1997, the National Science and Technology Council's Committee on Computing, Information, and Communications (CCIC) sponsored an invitational workshop, Research Challenges in High Confidence Systems (HCS). The workshop was held at the Institute for Defense Analyses, Alexandria, Virginia, and chaired by Ms. Teresa Lunt, Program Manager, Information Survivability, Information Technology Office/Defense Advanced Research Projects Agency.

The objectives of the workshop were to explore research topics that would enable the creation of new technologies for developing and assessing high confidence systems, to recommend integrated research challenges to stimulate and focus high confidence systems research, and to suggest national goals and benefits to encourage U.S. Government interagency commitment to a national research agenda in high confidence systems.

To achieve its objectives, the workshop considered the findings of the 1995 Committee on Information and Communications (CIC)¹, and explored in more detail the gap between the needs for high confidence and the HCS state-of-the-art for high confidence, as well as the research needed to begin to fill in that gap.

Background

The 1995 CIC report defines a high confidence system as one in which the designers, implementers, and users have a high degree of assurance that the system will not fail or misbehave due to errors in the system, faults in the environment, or hostile attempts to compromise the system. Such a system can be expected to behave appropriately within an operational context envisioned by its creators. High confidence, then, is a measure of predictability that a system will behave within established expectations. Highlights from the 1995 CIC report include:

There must be clear benefit from such systems: their use "must exceed the costs of not using them, failure, and misuse." Initially, only those systems for which the consequences of failure are very serious are likely to receive the attention and extraordinary efforts of assuring correct and safe operation. However, the research goal should be to not only enable the achievement of high confidence systems but also to raise the possibility that the technologies developed to create such systems are eventually usable to increase the levels of confidence that can be placed in all operational computing systems.
"...high confidence systems of the future require integration of many properties including functional correctness and safety, fault tolerance, time-critical response, and security." The report recommended that a national research agenda be developed by using high confidence research as an integrative, inter-disciplinary incentive for these related areas.
Research is needed to develop measures of confidence, methods for predicting confidence, and measures for understanding the costs of high confidence.
The 1995 CIC report also recommended research on system and component engineering methods, including assurance methods, for building useful high confidence systems in cost-effective ways, as well as large-scale experiments to demonstrate that high confidence systems can be built cost effectively.

Workshop Organization

The 1997 CCIC workshop was conducted in two parts: Problem Exploration and Solution Exploration. It used two groups of panelists to help stimulate these explorations. The first group comprised participants who possessed high confidence systems problems; the second comprised participants who possessed technologies or research ideas that could contribute to addressing the problems. The panelists for each session are identified in Appendix A. The panelists each briefly presented either the problems faced by their agency (or field) or the solutions that might contribute to addressing such problems.

At the end of the first day, the problems were further explored to develop the challenges, issues, and priorities for HCS research. During the second day, the solution was further explored to identify possibilities, approaches, and goals for HCS research and development (R&D).

Endnotes

The July 1995 workshop was sponsored by the Committee on Information and Communications (CIC) (succeeded by the Committee on Computing, Information, and Communications (CCIC)) under the National Science and Technology Council of the National Coordination Office for High Performance Computing and Communications. A report on this 1995 workshop, America in the Age of Information, is available at http://www.ccic.gov/ccic/cic_forum_v224/cover.html.