A Parallel Implementation of Acoustic Propagation Model on Cloud Platform

Three dimensional wave propagation model of parabolic approximation type is widely used in exploring the ocean. For this application, FOR3D model is one of the mostly used models, for it takes the azimuthal coupling into consideration and the result gains a better precision. When the precision requirement is high, large scale computation task would be faced, which cannot be solved only in one computer or constrained time. In this paper, we propose a parallel method to decompose the computation task, which divide the original computation into small size. Then each processor get one piece of task and run the FOR3D independently. Our method was implemented on Windows Azure and the result shows that we almost gain a linear speed-up.