| Since firstly presented by K.S.Yee in 1966, the Finite Difference Time Domain (FDTD) method, characterized by it's simplification and directness, was applied extensively in various electromagnetic computation area. But in order to meet the requirement of precision and Courant stability condition, the FDTD method have to generate enormous grids when simulating electric-large objects or objects of complex structures. Hence two problems have to be confronted, huge memory consumption and long execution time, which seriously obstruct the application of FDTD in the electromagnetic numerical computation field.With the rapid improvement of the personal computer and the local network, the cluster parallel computation is widely employed in many fields. Owe to its excellent expansibility and high performance versus price ratio, the Beowulf system, which consists of personal computers connected by high speed Ethernet network, becomes more and more popular in various parallel computation areas.By dividing the whole calculation space of the FDTD into some sub-spaces and assigning every sub-space to one node in a parallel computer system, we can run the FDTD in parallel. In this way, the huge memory and CPU time requirements of the FDTD can be sharply decreased because one node only deal with a sub-space instead of the whole calculation space. Therefore the parallel FDTD is effective to simulate electric-large or complex objects. Though the parallel FDTD already been widely studied, it still be difficult for the parallel FDTD to achieve high computation efficiency. Usually its efficiency will drop... |
PDF Full Text Request