ABSTRACT

This paper presents an algorithm for predicting crosstalk in multiconductor transmission lines (MTL). Currently, crosstalk has an increasing impact on signal integrity. In the twisted pair cable (TPC), different rotation degrees correspond to different RLCG parameter matrices, these matrices are predicted by the beetle antennae search-back propagation neural network (BAS-BPNN) algorithm, and use implicit wendroff-finite difference time domain (IWFDTD) method to solve near-end and far-end crosstalk voltages. The CST simulation results show that near end and far end crosstalk can be predicted effectively. This proposed method is expected widely used in electronic applications for fast data transfer.

摘要

本文提出了一种预测多导体传输线（MTL）中串扰的算法。当前，串扰对信号完整性的影响越来越大。在双绞线电缆（TPC）中，不同的旋转度对应于不同的RLCG参数矩阵，这些矩阵是通过甲虫天线搜索回传神经网络（BAS-BPNN）算法预测的，并使用隐式Wendroff时差有限域（ IWFDTD）方法来解决近端和远端串扰电压。CST仿真结果表明，可以有效地预测近端和远端串扰。预期该提出的方法将广泛用于电子应用中以进行快速数据传输。