Abstract

This paper introduces the stochastic Gabor function, an excitation waveform that can be used for finite-difference time-domain S-parameter estimation. The stochastic Gabor function is a Gaussian function modulated by uniformly distributed noise; it has wide frequency spectrum representation regardless of the stimuli pulse length. The stochastic Gabor function was studied in the time-frequency domain and was compared to Gaussian and Gabor stimuli functions with the same length. As shown by frequency concentration measurements, the stochastic Gabor function is least compact in the sample frequency phase plane. Numerical results obtained by using a multilayer stripline indicate that the stochastic Gabor function provides convergence and stability similar to those provided by the Gabor and Gaussian functions, but produces a much wider frequency band response when used as a pointwise hard voltage source stimulus