The Research Of Boundary Effect And Operational Costs In Fscs

Adaptive Random Testing (ART) is a kind of black box testing method which developed in recent years. Fixed Size Candidates Set Version of ART(FSCS) is one of the most classic and most widely accepted algorithm to implement the ART algorithm. The ability of the FSCS to find bugs increased significantly compared with the Random Testing(RT). However, for FSCS, due to the limitation of the input field has led to the phenomenon of boundary effect. In the choice of the ART algorithm also generates new test case it need a lot of distance calculation, especially when the failure rate of small and large dimension of the case, The huge cost of running the algorithm has reached a very difficult intolerable level, which seriously affect the efficiency of the testing.Based on the FSCS in-depth analysis, and access to a large number of documents, this paper put forward the Circled FSCS(referred to as C-FSCS),which transfer the input field with boundary to a boundary-free circle.Thus, solves a fundamental problem of boundary effect algorithms. To solve the FSCS's problem of huge cost, this paper proposed an Improved C-FSCS(referred to as the IC-FSCS), the algorithm directly cut out most of the unnecessary computational overhead, which greatly FSCS increased level of operating efficiencyIn addition, this paper applies this method to Restricted Random Testing(RRT), proposed the Circled RRT(C-RRT) algorithm and Improved C-RRT algorithm which solve the problem of huge cost and the boundary effect which also exits in RRT algorithm.In order to verify the validity of the proposed algorithm, we made a large number of experiments. The experimental results show that this improvement is very effective.