Using Automated Regularization deconvolution to extract input parameters of the phase function from measurement

In recent years theory of Radiative Energy Transfer (RET) has been adapted as a reliable tool to predict radio wave propagation through and near vegetation. 4 independent input parameters are demanded for the developed RET model. Those parameters so far had to be established from one measurement only, thus limiting their accuracy. An independent measurement, which is termed the phase function, can readily yield 2 of the 4 input parameters independently, which significantly increases the accuracy of these parameters. However one major factor influencing the phase function measurement is the radiation pattern of the receiving antenna. The actually measured pattern will be the convolution production of the antenna radiation pattern and the phase function. The measurement therefore needs to undergo an "inverse process", i.e. deconvolution, before any RET input parameters can be derived from it. This paper presents a deconvolution technique developed by using the Automated Regularization method. Improved results are demonstrated in terms of the RMS errors. The measurements were conducted in a vegetation setup inside an anechoic chamber.