| DARPA ISO
Sponsored Research
1999 Project Summary
|
|
| Project Website: | http://www.cs.cmu.edu/~wls/foura/ -- Additional project information provided by the performing organization |
| Quad Chart: | Quad Chart |
| Objective: | The FOUR-A Project will investigate methods for the development, analysis, and adaptation of assured software systems based on underlying techniques of advanced program analysis, annotation, and manipulation. |
| Approach: | The effort has two principal thrusts: (1) experimental
application of tools and techniques to the improvement of software dependability,
to the assurance of software dependability properties, and, as a system
evolves over time, to the preservation of those properties and their assurance,
(2) development of underlying core techniques for program analysis,
manipulation,
and annotation, including their embodiment in prototype tools for
developing and improving Java programs.
The experimental application of the prototype tools will enable exploration of how assurance and dependability can be better managed in source-level Java programs as they are developed and evolve, particularly in situations where not all of the source-code for a system is available for analysis. (In COTS-based component-oriented software engineering, it is typical not to have all code available for analysis, for reasons of trade secrecy, security, or separate development activity.) |
| Recent FY-99 Accomplishments: | (This project is a new start.)
Engineering work on the tool focused on the mechanism related to managing the audit trail of changes and on the user interface architecture. Steps were taken to initiate the vulnerability evaluation. |
| FY-00 Plans: | The overall schedule for FOUR-A is based on multiple
iterations of (1) analysis of source-code-level dependability requirements,
(2) development of manipulations and analyses, (3) engineering of the tool
prototype, and (4) application in experimental case studies. As part of
this e ort, the extent of tool coverage of Java language constructs will
be increased. Evaluative case studies will be ongoing, starting as pen-and-paper
studies and evolving to employ increasing levels of tool support.
Specific plans for FY-00 include:
|
| Technology Transition: | The principal artifact resulting from the project
is the prototype FOUR-A tool system and its components, which include the
general-purpose IR representation, a robust and flexible Java-based user
interface system based on model/view chains, and a collection of analysis
algorithms and manipulations, suitably encapsulated.
An important desiogn goal is for the tool to be easily usable in a laboratory or other evaluative setting, reducing the extent of training and technical background required for productive use. For example, many complex and deep analyses and manipulations can be ``hidden'' behind simple interfaces. The project will also produce results in the form of papers, talks, and collaborative relationships. |
| Principal Investigator: | PI Name: William L. Scherlis
Organization: Carnegie Mellon University Address: School of Computer Science, 5000 Forbes Avenue, Pittsburgh, PA 15213 Phone: 412-268-8741 Fax: 412-268-7458 email: scherlis@cs.cmu.edu Admin Contact Name: Karen Faber
|