CRI: Acquisition of Research Instrumentation Infrastructure for Next-Generation Broadband Communication Systems

Lead Proposal: CNS-0551686 PI: Henderson, Thomas R. Institution: University of Washington

Proposal: : CNS-0551378 PI : Riley, George F. Institution: Georgia Tech Research Corporation - GA Institute of Technology

Proposal: : CNS-0551706 PI : Floyd, Sally Institution: International Computer Science Institute

Investigators at the University of Washington, the Georgia Institute of Technology and the International Computer Science Institute will re-design, enhance and maintain the Network Simulator to address research and education challenges for the next generation of data networks. Improvements will include a new simulator architecture, new models for wireless networks, provide for software encapsulation, and integration of the tools with virtual network testbeds. The changes will enhance scalability, extensibility, and new application program interfaces that open the simulator to open source networking software. Emulation capability will be improved to allow integration with testbeds.Wireless modules will be rewritten to track advances in wireless networking. Educational scripts will be facilitated in the enhanced version. The Network Simulator is heavily used in research; these improvements will allow it to continue to be a leading resource for

Project Report

This project initiated and developed what is known as the ns-3 discrete event network simulator, a freely available network simulation tool for research and education. As the name implies, the tool is a third-generation network simulation (ns) tool intended to replace the widely-used ns-2 tool that was developed in the late 1990s. ns-3 was set up as an open-source software project, enabling the contribution of a number of research organizations worldwide. In particular, close collaboration with the French research institute INRIA turned out to be integral to the project's rapid development. ns-3 is a software tool that allows users to experiment with models of data communications networks such as the Internet or sensor networks. ns-3 allows users to conduct network experiments (such as evaluating how a new piece of networking software might behave if actually deployed in the Internet) by testing models of such software in a synthetic environment constructed by the simulator. Experimentation of this sort has several advantages, including ease of use, low cost and low barriers to entry by users, reproducibility, and adaptability to classroom use. However, the advantages above can also turn out to be liabilities if over-simplified or erroneous results are produced and the wrong conclusions are drawn. One of the main aims of the ns-3 project was therefore to improve the realism of the simulator while not losing the features that make it attractive for research and educational use. In particular, ns-3 provides new ways to combine simulations with real or virtual machines, and to allow the researcher to more easily move from working in a simulation environment to conducting network experiments in the field or on a testbed. Several techniques for allowing the reuse of software applications or network implementations directly in the simulator have been enabled. We believe that such integration also reduces the likelihood that errors are made due to over-simplified models. The broader impact of this project has been significant due to the software being developed and made available as open source. ns-3 has become one of the leading network simulators for networking research. According to web server logs, there were over 70,000 downloads during the calendar year 2010. The project conducts technical discussions and provides help for users on two public Internet mailing lists, both of which have approximately 1000 subscribers at the time of this writing. A number of recent research publications, including those at top-tier conferences, cite the use of ns-3. The tool has also been used as course-based software at several universities and is being used in industry and government research centers. The project continues to grow via additional funding and by attracting new software developers to the open source project.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Application #
0551686
Program Officer
Theodore Baker
Project Start
Project End
Budget Start
2006-07-01
Budget End
2010-12-31
Support Year
Fiscal Year
2005
Total Cost
$560,010
Indirect Cost
Name
University of Washington
Department
Type
DUNS #
City
Seattle
State
WA
Country
United States
Zip Code
98195