Statistical Simulation of Crosstalk in Random Cables

This paper analyses the problem of the crosstalk prediction in random bundles with a novel approach based on both statistical simulation and the continuous non uniform MTL models. A direct Monte Carlo method is used to generate several random bundles statistically, loaded on both the sides by known impedances. By solving numerically the corresponding non uniform MTL equations we produce, for a single frequency, a set of data describing the range of the interference magnitude. The systematic statistical manipulation of the obtained data leads us to characterise the crosstalk in terms of probability density functions. The results of the present model are compared to measurements carried out in a published case study.