An Integrated Framework for Testing Memory and Power Performance of Android Apps
碩士 === 國立臺灣大學 === 電子工程學研究所 === 105 === When it comes to software testing in the industry, memory leak bugs of mobile applications (apps) are notorious to detect. Existing techniques usually rely on the availability of source code and instrumentation techniques to generate test cases. We have develop...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2017
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/76931321937279389562 |
| Summary: | 碩士 === 國立臺灣大學 === 電子工程學研究所 === 105 === When it comes to software testing in the industry, memory leak bugs of mobile applications (apps) are notorious to detect. Existing techniques usually rely on the availability of source code and instrumentation techniques to generate test cases. We have developed a framework to automate the black-box testing of app memory leaks. Our testing procedure first builds a finite-state model from the execution traces of the app under test (AUT). Specifically, we label each state in the model with memory usages and then extend Karp’s min-cycle algorithm to a k-max cycle algorithm. Then we use the algorithm to identify those cycles in the model with large memory usage and generate test cases to execute the AUT through those cycles for several times to observe whether the memory usage will blow up the AUT. Finally, we report our implementation and experiment with some apps.
|
|---|