Many natural and manufactured structures can be effectively modeled as networks of one dimensional segments joined at nodes. A new algorithm for the numerical solution of various time dependent partial differential equations on some of these networks is presented. The main novelty is a network version of the Fast Fourier Transform, which provides an efficient technique for expansions with eigenfunctions of the Laplace operator.

翻译：许多自然和人造结构可以有效地建模为由节点连接的一维线段组成的网络。本文提出了一种新的算法，用于数值求解某些此类网络上的各类时间依赖偏微分方程。其主要创新点在于一种网络版本的快速傅里叶变换，该变换为利用拉普拉斯算子特征函数进行展开提供了一种高效技术。