外部記憶装置を効率的に用いるFD-TD法の計算アルゴリズムに関する提案

抄録

In this report a new calculation algorithm of FD-TD is proposed. The principle of FD-TD depends on only the calculus of finite difference. It can be applied to the complex structure of a waveguide without any mathematical principle. The disadvantage of FD-TD is simple. If the simulation need high calculation accuracy, the fundamental mesh size of the difference equation must be short. It means huge memory capacity is required to make a virtual waveguide on the computer. Our method uses the outer storage unit with high efficiency to simulate the longer waveguide size with the real memory capacity. Then new algorithm can offer the high speed FD-TD calculation for the longer waveguide without the collision happening on the waveguide end boundary due to the improved data moving system.