Multitasking Path Coverage Model And Multifactorial Optimization Framework Of Test Case Generation For Path Coverage

Automated test case generation for path coverage(ATCG-PC)is an important task of automated software testing.ATCG-PC problem aims to achieve higher path coverage with less testing cost.Due to its complex nonlinear structure,many scholars in recent years have modeled the problem as single-objective optimization models or a multi-objective optimization models by using functions based on path structure(F-PS).There is certain similarity among these F-PSs.However,existing optimization models do not consider exploiting the similarities between F-PSs to facilitate the optimization of multiple objectives.If the similarity among these F-PSs is fully utilized in optimization,the solution quality and efficiency of ATCG-PC problem may be further improved.Based on the above ideas,the main research works of this paper are as follows:1.A multitasking path coverage(MtPC)model for solving ATCG-PC problem is proposed.Based on the similarity among F-PSs,the MtPC model is proposed in this paper.This is the first attempt to apply multitask optimization to solve ATCGPC problem.Compared with the multitask optimization model,the MtPC model aims to solve ATCG-PC problem through knowledge transfer and joint cooperation between two tasks,while multitask optimization solves multiple optimization problems simultaneously through knowledge transfer.2.A MfO-PC framework inspired by multifactorial optimization(MFO)framework is designed for solving MtPC model.The framework contains two tasks which are optimized by a designed multifactorial optimization algorithm based on MfO-PC framework.And the two tasks can jointly generate the desired test cases through the automatic assignment strategy of MfO-PC framework.Compared to MFO,MfO-PC framework is a designed solver for solving MtPC model,while MFO is a paradigm for solving multiple optimization problems simultaneously.3.Four multifactorial optimization algorithms based on MfO-PC framework is designed.By applying differential evolution algorithm(DE)to MfO-PC framework,a multifactorial differential evolution algorithm based on MfO-PC framework(MFPCDE)is designed to verify the effectiveness of MtPC model.By applying immune genetic algorithm(IGA),adaptive particle swarm optimization algorithm(APSO)and backtracking search optimization algorithm(BSA)are respectively applied to MfO-PC framework,multifactorial immune genetic algorithm based on MfO-PC framework(MFPCIGA),multifactorial adaptive particle swarm optimization algorithm based on MfO-PC framework(MFPC-APSO)and multifactorial backtracking search optimization algorithm based on MfO-PC framework(MFPC-BSA)are designed to verify the effectiveness of MfO-PC framework.In order to verify the effectiveness of MtPC model and MfO-PC framework,the paper first combines several F-PSs that are common in ATCG-PC problem into four multitasking path coverage problems,and then according to different multitasking path coverage problems,the performance of the designed algorithms MFPC-DE,MFPC-IGA,MFPC-APSO and MFPC-BSA are tested on six classic fog computing test programs and six classic natural language processing test programs.The experimental results show that compared with the original algorithms,the algorithms MFPC-DE,MFPC-IGA,MFPCAPSO and MFPC-BSA designed in this paper can obtain higher path coverage with fewer test cases on most test programs.The results show that MtPC model and MfO-PC framework designed in this paper using the similarity between F-PSs can effectively improve the efficiency and quality for solving ATCG-PC problem,and reduce test cost.The research in this paper provides a new vision for solving ATCG-PC problem,and has certain theoretical significance and application value.