Measurement, characterization and modelling of vegetation barriers to reduce electromagnetic pollution in sensitive areas

Resumo

The fear to radiofrequency has grown among people in the last few decades with the deployment of a huge amount of mobile telephony antennas. The power levels emitted at base stations are reasonably low, and so, it seems that the risks for health in such systems have been oversized. Thus, many countries have included new legislation to control the radioelectric exposure of the population and, some of them, to protect the most sensitive sectors of the population, like elderly or children. So, some zones classified as "sensitive areas" like kindergartens, geriatrics or even hospitals are expected to maintain their electromagnetic pollution to minimum levels around their installations. A typical way to reduce the radioelectric pollution is the use of electromagnetic shielding. Several electromagnetic shielding techniques have been already studied and their effectiveness has been demonstrated, like wide concrete walls, selective absorber surfaces, or even metallic netting in the walls. However, all these methods are very expensive, very difficult to implement and, in most cases, cause a big strong visual impact. This Thesis presents a more economic and ecological solution, with a soft visual impact, that is the using of the excess attenuation induced by vegetation elements to set up fences that isolate the areas under interest. Three huge measurement campaigns have been carried out to be able to develop the theories and methods explained in this PhD, two inside an anechoic chamber and one developed outdoors, at free space conditions. All the samples have been taken for the frequency bands of mobile telephony, both GSM (900 and 1800 MHz) and UMTS (2100 MHz), at both horizontal and vertical polarisations. The first step to analyze is the attenuation provided by isolated vegetation specimens under radiofrequency radiation, both in calm and windy conditions. This study provides information about the attenuation of a single element obstructing the radio path. Seven different species were considered during the experiment. The analysis of the results leads in obtaining specific attenuation constants for the vegetation specimens under study, as well as an empirical equation to predict the attenuation provided at several other frequencies. As the measurements include different wind speeds, an analysis of the scattering effects at long-term and short-term is also included. Once the vegetation species have been parameterized, the second step should be locate them in regular disposals to study the behaviour of different vegetation barrier configurations. Therefore, a second extensive measurement campaign shows the attenuation results for several barrier configurations and different vegetation species. Results remark that current ITU-R and CCIR methods to predict attenuation due to vegetation in radio paths are not accurate enough, and they provide estimated attenuations with errors up to 10 dB in some cases. The huge error in the estimation is because the method do not have into account neither the disposal of the barrier elements, nor the characteristics of the vegetation specie itself, as humidity, massive trunks or branches, foliage density, water content, etc. Two simple and new mathematic approaches have been obtained from all the empirical data that estimate the attenuation because of vegetation with an error lower than 2 dB, which is accurate enough to be used in radio planning. The first method computes the attenuation based on the specific attenuation of the specie under study, the frequency and the density of the barrier. However, obtaining the specific attenuation of vegetation specie involves a large measurement campaign, so a more general method is presented. The second method can be applied for all kind of vegetation specimens, as its results depend only in the frequency, the density of the barrier and a parameter depending on the type of the vegetation specie itself. This parameter is the one taking into account those characteristics as water content, foliage type, etc., and it implies a very simple but previous classification of the vegetation specie to be analysed. Results obtained in this section seem to be very accurate, and vegetation barriers work as expected, providing attenuations up to 9 dB in almost all cases of species measured. To improve even more the behaviour of the vegetation fences, and to reduce the variability due to the wind, an innovative idea has been implemented: the use of lattice supporting elements combined with vegetation specimens. Lattice structures increase even more the excess attenuation while reducing the spreading of the receiver power around median values. This effect is based in the fact that lattices are rigid elements that do not sway with the wind, and so, attenuation becomes much more stable and easily predictable. Different lattice materials and hole patterns have been measured using commercial structures for gardens and yards, for several barrier configurations, frequencies, polarizations and wind speeds. The configurations with better results were those combining mixed lattice structures and shrubs at one or both sides of the lattice. Due to the complex configuration of this third measurement campaign, performed within an anechoic chamber, only one vegetation specie was tested with the lattice. The specie chosen was the Ficus elastica, because it provides good attenuation values along all frequency band and because it is a shrub commonly seen both in outdoor environments as indoor ones. Results show the good behaviour of the lattice element as an additional attenuator, and the influence of the material in the attenuation provided by the barrier, confirming the aim of this Thesis. In summary, this PhD Thesis demonstrates that the use of vegetation barriers can be a non expensive, soft, and ecological way to reduce electromagnetic pollution levels in our sensitive areas.