|
Software Diagnostics and Conformance Testing Standards and Conformance Testing Architecture and Performance of Advanced Distributed Systems Information Storage and Integrated Systems
|
Software Diagnostics and Conformance Testing Division Contact: Mark W. Skall
NIST is improving software quality by developing reference implementations and conformance tests of standards, developing software diagnostic tools, performing research to reduce the cost of testing, developing reference data to measure the effectiveness of software tools, and improving the development of software standards. Software controls the functionality of so many devices in every business and home that even isolated failures have an adverse effect. These devices are now being connected to the Internet. A software failure in a device can now have an adverse effect on many more users beyond the device's owner. We are developing diagnostic tools and test methods for improving software quality and reducing software failures. We are using mathematical methods to improve the precision of software specifications, standards, and tests. Researchers are using these methods to automatically generate test suites from specifications. We are partnering with standards organizations to develop standards and test methods for interactive TV, the convergence of two currently separate media: TV and the Internet. In addition, we are improving the quality of health-care information systems by designing distributed architectures, developing reference implementations of those architectures, and participating in the development of standards for healthcare information. Looking further into the future, we are developing information representations and computing machines that take advantage of quantum effects, such as superposition and entanglement, which could enable exponential computational work to be done in one operation. Contact: John Barkley Standards and Conformance Testing We are designing and developing test methods for conformance tests and using these methods to produce test metrics and test suites for evaluating the quality and correctness of information technology products. Several conformance test metrics and test suites have been produced for evaluating implementations of Virtual Reality Modeling Language, language compilers, computer graphics metafile, extensible Markup Language, and the Document Object Model. In collaboration with industry, the user community, and various testing laboratories, we are working to identify new technologies and to develop measurement schemes, test collections, reference implementations, test suites, and tools necessary to advance the technology and produce usable, reliable implementations. Past research has resulted in a method for minimizing subjectivity in the evaluation of visual results in graphics testing. This method is implemented by a customizable interface, randomized self-explanatory displays, and automatic capture of results. Our recent activities focus on new methods for software testing based on statistical measures as well as automated test generation methods to develop conformance tests. Contact: Lynne S. Rosenthal NIST recognizes the need to increase and adjust its involvement in technology specifications used in the development of high-technology software. In today's information processing environment, it is becoming increasingly common for many new technology specifications to be developed in organizations outside the formal standards environment and for those technology specifications to quickly become de facto standards. Accordingly, we are seeking to be involved with organizations, both inside and outside the formal standards environment, that are developing pivotal, forward-looking technology specifications in multiple areas related to the production of high-technology software. This involvement will emphasize the development of appropriate metrics, measurement technology, and techniques for testing conformance to new high-technology software specifications. Another one of our goals is to ensure the interoperability of these specifications among both industry and the federal government. Accordingly, we work closely with federal agencies to ensure that their standards and testing requirements are met. One example of this is a joint project with the National Institute of Justice to develop a National Software Reference Library. Contact: Barbara
Guttman Architecture and Performance of Advanced Distributed Systems We are promoting the effective evaluation and efficient use of advanced computer layouts and characterizing new computer architectures to identify improved technology for applications such as electronic commerce. We also are exploring standardized, economic programming methods (for example, digital interactive television) and design of coherent evaluations that economically and reliably characterize the systems. Our work has involved the development of instrumentation and related management techniques for networked computing, communications, and user appliances. A number of instrumentation products are available; these include VLSI MultiKron® clock chips and plug-in boards, a comprehensive tuning tool (S-Check®) for improving the performance of concurrent codes, and hardware for local synchronized clocks on cluster systems. Our current projects include a reference platform and reference applications for interactive digital TV, cluster computing, network time synchronization, and assurance methods for electronic commerce. Contact: Alan Mink
Our group conducts research and development to support interoperability and standards for information storage, display, and related multimedia systems. We conduct investigations of integrated system and device performance for improving the safe preservation, fidelity of transmission, and secure access of information and associated media in or among sub-systems and the computing enterprise. In addition:
Contact: Wo Chang The biometrics project helps U.S. industry in the development of measurements and standards. We are developing testbeds to explore the metrology issues associated with optical character recognition as an approach to understand the errors induced by optical information systems. We also suggest improvements leading to reduction in errors and increased performance. Contact: Fernando Podio
Date
created:October
22, 2001 |