Research On Software Fault Location Based On Biological Consensus Sequence

When software has a failing running, the problem of how to locate a software programm is worthing exploring. Although the fault location software technology has some experience in this study can draw on, but how do the various methods to achieve efficient, rapid, automated fault diagnosis fault location is still key issue.In this paper, we summary localization of software failure at home and abroad, specially model-based fault location technique, analyze problems in the present study and select the use of fault running sequence and one success running sequence for fault may cause not to find error point as a key point. Under the support of the techniques, such as software fault diagnosis, software running sequences and biological sequence alignment, combination of "Nearest Neighbor" thinking, introduced gene biology the principle of alignment, and then a algorithm is implemented, which is based on biological consensus sequence. The experiments prove the correctness and validity. This paper includes the following two aspects:①Because of fault running sequence and one success running sequence for fault may cause not to find error point, first analyzed the sequence features of the software running, describes the concept of running sequence and biological sequence alignment in the corresponding relationship, alignment will run into the problem of biological sequence alignment; And on this basis, propose a multi-running series of consensus sequence generation algorithm, the algorithm considers a number of success running sequences impact of the fault location, by reference biological sequence alignment in principle, the sequence number of the nearest success running sequences carried out extraction, was a consensus sequence, and a detailed description of the algorithm.②On previous paper basis, research on software fault location based on biological consensus sequence:according to "Nearest Neighbor" thinking, used to compare differences in consensus sequence and fault running sequence, get stable set of sequence segment, then removed stable set of sequence segment from the contents of fault running sequence, solve the suspicious sequence segment, contains the point of failure to find a "basic block", to achieve fault location; Tests show that this method of fault location during procedures to exclude the more part of the fault and not to narrow the source code review and improve the efficiency of fault location.