Abstract | ||
---|---|---|
Function-level reuse of legacy systems plays a significant role in promoting the efficiency and accuracy for constructing a new parallel software system. However, the scale and complexity of legacy systems are always so large that we cannot determine which functional parts are worthy of reuse intuitively and effectively. In this paper, a Colored Petri Net (CPN) model based system behavior reduction and analysis approach is proposed for function-level reuse of parallel software. Specifically, a CPN model for the legacy system is available as precondition, and we first mark the input/output places and transitions on this model. Then, a model reduction approach using trace-equivalent is applied to generate an external behavior equivalent model with smaller scale. Besides, a behavior analysis approach based on linear behavior sequence to be analyzed is proposed, which could automatically get all execution fragments containing the behavior sequence from state space of the reduced model. Based on the reduction and analysis approach, we confirm effectively the reuse contents in original parallel software. The advantage and effectiveness of our method is shown through a performance analysis. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1109/COMPSACW.2012.64 | COMPSAC Workshops |
Keywords | Field | DocType |
function-level reuse,behavior sequence,analysis approach,behavior analysis approach,legacy system,linear behavior sequence,model reduction approach,cpn model,reduced model,parallel software reuse driven,cpn model reduction,external behavior equivalent model,parallel programming,software systems,petri nets,software maintenance | Petri net,Systems engineering,Computer science,Reuse,Precondition,Real-time computing,Software system,Parallel software,Software maintenance,State space,Legacy system | Conference |
Citations | PageRank | References |
0 | 0.34 | 3 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tao Sun | 1 | 120 | 15.48 |
Xinming Ye | 2 | 52 | 11.62 |
Hongji Yang | 3 | 1039 | 137.37 |
Jing Liu | 4 | 187 | 57.41 |