| Summary: | For modern electronic systems, one of the important issues in electromagnetic compatibility (EMC) is the characterisation of electromagnetic emissions from a complex device which has essentially characteristics of random noise. This thesis presents the study of a development and challenges in realising near field correlation measurement for efficient characterisation of stochastic electromagnetic fields. The main contribution of this work is presented as developing efficient field-field correlation measurement strategies in enabling efficient numerical methods for modelling the propagation of radiated transient, noisy electromagnetic field ofmultifunctional electronic devices. The new numerical propagation methods used are based on ideas fromwave chaos theory using Phase-Space representation, the Wigner Distribution Function (WDF). The propagation techniques make use of the connection between the wave correlation functions and phase space densities which will be tested using the input data obtained from measurement. Providing relevant experimental procedures are an essential prerequisite to investigate the efficiency and accuracy of the propagation technique.
The state of the art research facilities at the George Green Institute for Electromagnetics Research (GGIEMR) will be used to develop the one-probe and twoprobe near-field scan for the measurement of field-field correlation for electronic equipment with many uncorrelated sources that essentially have a stochastic field distribution. The one-probe scan is a frequency-domain measurement while the twoprobe scan is a time-domain measurement. The phase/time reference issues and II other challenges in performing the measurements are also addressed in this thesis. As the use of two probe scanning is very time consuming and requires large computer resources, methods for reducing the complexity of measurement of stochastic fields using principal component analysis is explored. Also, an approach to analyse nonstationary emissions will be investigated in order to efficiently characterise emissions from a complex and time dependent system.
|
|---|