SBIR-STTR Award

With over 20,000 nationwide users of dozens of unique experimental facilities and high performance computing resources, DOE has a long history of successful network research, advanced network deployment, and advanced applications support. The heavily collaborative and computing nature of these DOE researchers drives the requirements for sophisticated allocation systems for accessing distributed resources such as supercomputers, high-end storage systems with extremely large scientific data sets, unique on-line facilities, and massive, multidimensional datasets in tele-immersive environments. However, allocation systems for these differentiated network services are not currently available. This project will develop an allocation system for production networks like DOE’s Energy Sciences Network (ESnet). Phase I will survey current DOE allocation processes, develop a requirements list for Esnet, and build a prototype allocation system for differentiated services on ESnet.

Commercial Applications and Other Benefits as described by the awardee:
Commercial applications for allocations for differentiated services exist in both corporate intranets as well as the public Internet. It is anticipated that this allocation system could provide the basic platform for this service

A management policy is needed for allocating Internet Protocol Quality of Service (IP QoS) among the users of DOE's ESnet (Energy Sciences Network) and for deploying that policy into bits per-second. This project will develop a software program, structured as a policy management tool, to allow less sophisticated network personnel to intelligently configure their IP QoS using human policies. Phase I demonstrated the concept of translating human policy into QoS policies. Design information was gathered from the literature as well as from QoS researchers and standards bodies. The findings narrowed the design of the policy management tool. The Phase II project will further develop software to transform human policy information into data that can be formatted into network-ready policy information. The software structure will include a wizard feature to simplify the translation of human policy to QoS-speak, a relational database to translate that policy into accurate and useful configuration files for improved network operation, an input interface component, and an output interface component.

Commercial Applications and Other Benefits as described by the awardee:
The target market for this software product is enterprise internets and the Internet. The software should provide a sim-ple way to configure various network devices for QoS and to validate what was deployed. Router, switch, and policy server vendors should seek such tools to enhance their IP QoS product lines.