A new FDTD parallel algorithm for wave propagation and absorption on big urban areas

Using FDTD to study wave propagation in big urban areas requires a huge amount of memory and processing time. This research deals with the problem of memory using a special formulation of 2-D FDTD to save memory. A system of linear equations is presented as a mathematical equivalent of Yee´s equations. A recursive algorithm using Gaussian back-substitution is presented. It was adapted to parallel processing in a network of workstations and used to solve this system. Results that validate this approach are presented and the advantages and future improvements to this method are discussed.