Test Suite Generation From Extended Finite State Machine By Using Search-based Approach

Extended Finite State Machines(EFSMs)are widely used to model the behavior of a system.Testing from EFSMs is expressed in term of generating transition paths and then deriving test data to trigger these paths.However,a transition path may be infeasible,thus it is impossible to generate test data to execute it.The path infeasibility problem is about generating transition paths through EFSM that are infeasible and cannot satisfy a given test coverage criterion.In an EFSM model,path can be infeasible due to the control and data part combination.The path infeasibility can also increase the fitness value of a whole test suite.Search-based approach has been proven efficient for testing from EFSM model.This thesis proposes test suite generation from EFSM model by using search-based approach.The proposed technique generates a set of transition paths from an EFSM model that satisfy the requirement-based and constrained path coverage criteria.A search-based approach is applied by using two techniques:(1)A fitness metric using dependence analysis to estimate the feasibility of a transition path;and(2)A total fitness function for the whole test suite fitness estimation.Two simple case studies are reported in this thesis to evaluate the effectiveness of the proposed technique.The results demonstrate the efficiency of the proposed technique.