Adaptive generation algorithm for executable state identification sequences in EFSM model

Abstract

In the protocol conformance testing system specified in extended finite state machine (EFSM) model,an algorithm for adaptive generating executable state identification sequences (ESIS) is introduced to improve efficiency in the Computation of state identification sequences (SIS).The executability of the ESIS generated in the novel algorithm can be ensured by an executability analysis method based on the executability analysis tree (EAT).Transition distinguish degrees and node convergence degrees are introduced to compute the weight of the current node for the EAT,which can be used to evaluate the correctness of the current search direction.Using a search strategy based on the node-weight function for the EAT,the node which is to be the root of the new sub-spanning tree is adaptively selected based on the weight of the latest explored node.In this way,the problem of automatic ESIS generation is formulated as adaptive exploration of the maximum weight node from unexplored nodes in EAT.Experimental results show that the proposed method needs to explore less state configuration than the executability analysis algorithm based on bread-first-search.