Crosstalk in stochastic cables via numerical multiseries expansions

In this paper, a numerical technique for the computation of polynomial chaos expansion coefficients pertaining to the per-unit-length parameters of cable lines with random cross-sectional parameters is implemented and employed to create stochastic and SPICE-compatible circuit models. According to the principles of the polynomial chaos-based simulation, statistical information on the cable response is obtained via a single circuit simulation, thus being much faster compared to traditional sampling-based techniques.