Simulation for a Crosstalk Canceling Algorithm in Lossless Multiconductor Transmission Lines

摘要

This letter presents an algorithm for canceling crosstalk in multiconductor transmission lines (MTL). At present, to reduce the crosstalk, multiconductor systems usually use differential mode transport, in which the number of transmitted signals is half the number of conductors. A method is proposed to cancel crosstalk in MTL, through the singular value decomposition (SVD) method in matrix theory, in which the transmission line matrix is converted into an identity matrix. That is, the off-diagonal elements in the transmission matrix are zero, which corresponds to crosstalk cancellation. In this letter, this method is applied to the previous literature models to verify its effectiveness firstly. Then, the method is applied to a triple-twisted strand. In the strand, different rotation degrees correspond to different per unit length (p.u.l) parameter matrix. A set of parameters are extracted, the rest of are predicted by the beetle antennae search-back propagation neural network (BAS-BPNN) proposed in a previous work. The simulation results show that the near and far end crosstalk can be canceled effectively.