Research On Complex Networks Based The Least Testing Paths Generation

The research is based on the national science foundation, "The measurement method and system of aerospace embedded trusted software". The paper puts forward an approach of generating the least testing paths to find errors during software executing. The author hopes to reduce the losses, increase efficiencies and cut down the cost via this method..The complex network and software testing are introduced at first. The paper proves a large software system can be modeled by the complex network theory. And then, the paper puts forward an approach to calculate the weights to set up a weighted network. They can find out the key nodes and extract testing paths for cluster analysis aim to reduce the test data and improve the test efficiency. Because of the uncertainty during software executing, the paper uses the fuzzy theory in complex network to achieve a higher coverage with less testing cases.Finally, a tool to generate testing paths of the SPARC simulator is shown in detail and a small example is presented to explain the algorithm. The testing result of Linux shows the method is effective.