Personal tools
You are here: Home Graduate Study Complexity Metrics for Testing Concurrent programs

Complexity Metrics for Testing Concurrent programs

Developing concurrent programs to a high standard of
reliability is notoriously difficult due to the large number of interactions between components executing in parallel.
Testing and automated source code analysis aid the developer in finding bugs and ensuring correctness. However, it is infeasible to analyse the entire program in detail since even for small concurrent programs the number of interactions and execution paths quickly become intractable. It, therefore, becomes important to identify parts of the program that are especially error prone, and focus testing and analysis efforts on these parts.
For sequential programs, software complexity metrics help identify the error prone parts.
For concurrent software, this problem has not received enough attention.
In this project we will first devise suitable complexity metrics for concurrent programs, and then apply them to increase the efficiency of automated software analysis and testing.

Potential Supervisor: Ajitha Rajan

Document Actions