Digital signal processing of time domain field simulation results using the system identification method

Abstract

A new technique for significantly reducing the computation time of time domain field simulations using modern digital signal processing techniques is presented. The procedure leads to a simple model of the investigated structure in form of a digital filter. The method is applied to TLM simulations of eigenvalue problems and to transient analysis using the FDTD method.

An Electromagnetically Coupled Package Feed-Through Structure for Multilayer Carrier Substrates

“…In a first step, a single transition from microstrip to triplate was calculated. With this transition, a sharp resonance was observed which, however, could only be resolved using the system identification method [7] to enable a reasonable computation time of the FDTD procedure. A thorough investigation of this resonance revealed its origin from the vias between the triplate ground planes.…”

“…The 3D structures involved in this transition were calculated with a FDTD method [5], using absorbing boundary conditions according to [6], and a system identification method [7] to speed up calculations and to improve the frequency resolution. Half of the (symmetric) feed-through structure is shown in Fig.…”

A method to decrease computation time of FDTD calculations for low frequency excitations

“…Indeed, some systems have a slow response, which need a long time to be died down. Lots of methods such as the system identification method [3], the Prony's algorithm [4] or the forward-backward autoregressive (ARMA) model [5] have been used but they are efficient only with resonant structures. The singularity expansion method of C. E. Baum [6] inspired some approaches which extrapolate transient signals in the time domain.…”

Microwave structure characterization by a combination of FDTD and system identification methods

“…Recently, digital signal processing methods have been used to reduce the computational requirements of the time-domain methods [2]- [6]. For example, the Prony's method is used to estimate the time signal in terms of the previously computed values [2], [3], also a covariance based system identification (SI) algorithm has been used to reduce the computation cost of the Transmission Line and FDTD Methods by employing a stochastic ARMA model [4], [5]. In this letter, a least-squares based system identification projection algorithm for a deterministic autoregressive moving average model is applied to the FDTD algorithm [7].…”