An Optimized Model Checking Parallel Algorithm Based on CUDA

Abstract

With the continuous advancement of the information society, the dependence of hardware and software systems in almost all fields is increasing. Model checking is a method of automatically verifying hardware and software systems in a finite state space. If system does not satisfy the formal formula description, then the system's counterexamples are given and the system's error will be corrected. However, the typical state explosion problem increases its complexity in time and space, which severely limits its applicability. In this paper, an optimized parallel algorithm for searching strongly connected components based on CUDA(FW-BW-CUDA) is proposed, which can effectively accelerate the speed of LTL model checking algorithm. More over, experimental results based on benchmark datasets show that compared with some state-of-the-art model checking algorithms, FW-BW-CUDA performs better in terms of speed in most cases.