Computational modelling of the electromagnetic fields close to a human body in realistic exposure environments for application in personal radiofrequency dosimetry

摘要

A promising approach in the field of personal radio-frequency (RF) dosimetry is the use of a body worn electromagnetic probe (i.e. personal RF body worn dosemeter) to record personal exposure to RF fields. Using such a dosemeter mounted on the body, an individual's exposure could be recorded at the time and place that it occurs. This would fulfil an important criterion for epidemiological and other studies into the relationship between RF exposures and health effects in humans. The relationship between a dosemeter measurement and quantities that characterise human exposure to RF fields, such as the spatially averaged exposure field and the whole body average specific absorption rate (SAR), is not well understood nor been rigorously quantified. This thesis tackled the problem of determining the relationship by employing computational modelling and statistical simulations to generate distributions of the fields close to the surface of realistic human body models. These fields are proxies for measurements made with an ideal, isotropic responding, body worn personal RF dosemeter (or simply dosemeter). It focussed on exposures occurring remote (in the far-field) from microwave antennas such as those associated with mobile and wireless base-stations, TV and other typically fixed radio transmitters. Exposures from RF transmitters worn on or in close proximity to the body were not the subject of this study. Initial computational modelling using single plane waves with arbitrary incident angles and polarisation revealed the extent and nature of the distribution of electromagnetic fields close to the surface of a body. Results indicated that a dosemeter measurement could lead to an over or under-estimation of the exposure field. Other factors affecting the close fields were also investigated to quantify their impact. Additional plane wave modelling found that dosemeter measurements could be used to estimate the exposure field or the whole body averaged SAR with similar levels of uncertainty. Finally, Monte Carlo simulations were performed to provide statistical information concerning the relationship between dosemeter measurement of the field close to the body and the exposure field and whole body averaged SAR. Exposures to RF fields were modelled using statistics that describe radio wave propagation in outdoor and indoor multipath environments at 450, 900 and 2100 MHz. Correction factors and associated standard deviations derived from the output of simulations can be applied to measurements made with actual, physical dosemeters to determine estimates of both exposure field and whole body averaged SAR.