Maxim Konovalyuk

Near-field propagation of cyclostationary stochastic electromagnetic fields

We have extended the method for modeling the stochastic EM near-field which has already been described in [5, 13] for stationary stochastic fields to the case of cyclostationary fields. Areas of application are the modeling of the electromagnetic interference radiated by digital circuitry inside the system and also into the environment, where the period of the cyclostationary EMI is given by the clock frequency of the digital circuits.

Planar stochastic sources localization algorithm in EMC problems

A two dimensional localization algorithm for stochastic EMI sources based on time domain measurements of electromagnetic near-field tangential components is presented. The autocorrelation and cross-correlation functions obtained by using the two-point planar scanning system are used for the characterization of stochastic field distribution. The simulation and experimental data are processed for a spatial localization of equivalent point sources. The estimated field distributions and patterns are presented.

Parametric reconstruction of radar image based on Multi-point Scattering Model

The wideband coherent-pulse radar provides high-resolution image of the target. The model of this image is a complex envelope superposition corresponding to signals diffracted by the point scatterers. The values of complex envelopes are distributed over the radar image coordinate plane in accordance with the point scatterer positions and their reflection coefficients. The radar image model consists of range and Doppler profiles. The parameters of the target point scatterers were defined by processing of two-dimensional (2D) data extracted from the complex 2D discrete Fourier transforms of the radar image. The proposed parametric system identification method performs the estimation of the model parameters for a short dwell time and the extrapolation of the radar data image beyond this time. The modified procedure of inverse synthesis aperture radar imaging applied to actual data showed a reduction of the Doppler smearing and some improvements of image resolution.

The joint inverse filtering and parametric identification for complex radar image

The paper suggests combining the inverse filtering and the parametric identification in the post-processing sequence of the complex radar image. The image can be modeled assuming the superposition of the identical partial responses from the effective target point-scatterers. Their positions correspond to the geometrical profile of target in the image coordinate plane. The proposed algorithm of radar image post-processing consists of the parametric and the non-parametric procedures which are used to improve the image resolution and to identify geometrical form of the target. The presented simulated results illustrate the main steps of radar image post-processing.

Localization of cyclostationary EMI sources based on near-field measurements

The near-field measurements of the stochastic fields radiating by digital electronic devices can be used for the localization of the stochastic electromagnetic interference (EMI) sources on the surface of device under test. The characterization of the space distribution for the wide sense cyclostationary (WSCS) stochastic field in the observation plane can be accomplished by a cross correlation between the reference and all remaining knots. The proposed Tikhonov regularization algorithm in conjunction with L-curve method provides a stable inverse reconstruction of the multi-dipole model distribution in the object plane. Parametric identification procedure based on the model order selection and two-dimensional Matrix Pencil Algorithm with subsequent fitting by Minimum Least Square technique gives the geometrical configuration of the WSCS sources. Implemented simulations and measurement experiments verified the choice of the proposed signal processing EMI sources localization algorithm.

Stochastic EMI sources localization based on time domain near-field data

The paper presents a planar stochastic EMI sources localization approach based on two-point time-domain data of electromagnetic near-field tangential components. The radiating structure is considered as a set of simple electrical dipoles arranged on a grid in the object plane. The autocorrelation and cross-correlation spectra of eradiated signals in two different space points are used to determine the distribution of dipole moments and to identify the parameters of the effective radiation sources. The modeling data was processed with the proposed localization procedure.

Network oriented modeling of EMI sources

Summary form only given. The near-field measurement is one of the widely used methods of complex device electromagnetic compatibility (EMC) analysis. The device radiation pattern can be calculated directly using the near-field measurement or after the reconstruction of equivalent dipole model in the object plane [1]. In accordance with equivalence principle the dipole model parameters completely determine the current distribution in the object surface. In the case of a planar structure such as Printed Circuit Board (PCB), these currents characterize the physical trace of signal transfer and conversion. Thus, the resulting dipole distribution can be used not only for the calculation of electromagnetic field at any space point, but also for the synthesizing of a simplified network oriented model of EMI sources inside the device under test (DUT). One of the possible modeling approach is the reconstruction of the structure external characteristics into the equivalent electrical circuit consisting of lumped elements (resistors, inductors and capacitors) and coupling elements, for example ideal transformers or delay lines [2]. This approach allows determining the conversion of a known input signal in the structure elements as well as its radiation characteristics. The presented model also makes it possible to find an equivalent circuit of the correction scheme that can be used for the reduction of transmission losses and minimization of spurious emissions. In this paper the EMI sources equivalent circuit synthesis based on the electrical dipole model is discussed. The generalized Brune method [3] is used to determine the multiport equivalent network. To estimate the parameters of system functions the Vector Fitting Method [4] is applied. The equivalent electrical circuit synthesis procedure based on the full-wave modeling of the electromagnetic structure is presented.