It's hard to build a computer network that works as required. A network contains many components, many layers, many protocols, many kinds of equipment, and many technologies. Despite the difficulty of building networks, businesses and individuals rely on networks for commerce, information, and entertainment, requiring them to be correct, efficient, reliable, and secure.
This project asks how to test networks for correct interoperation of their components. The first goal of the project is to develop the fundamental principles of network testing. The second goal is to use these principles to develop cost-effective testing methodology and tools.
The principles of network interoperability testing must help a tester answers the questions: What test cases are most cost-effective? How many test cases must be run?
This project seeks the answers to these two questions in the context of two approaches to testing, formal modeling and genetic algorithms. The approaches are being studied experimentally in the Lehman College networking laboratory, using equipment typical of enterprise networks.
Results will include evaluation of network modeling techniques and tools. They will provide methodology, algorithms, and tools for generating and executing test cases. The results will be disseminated through presentations at conferences, papers in academic journals, and making tools available to industry test labs.
The impact of the project will be to reduce the cost and improve the quality of network testing. It will also improve understanding of how to build testable networks. It will help to build future networks that are more efficient, reliable, and secure.
