Using artificial neural networks to scale and infer vegetation media phase functions

Abstract Accurate vegetation models usually rely on experimental data obtained by means of measurement campaigns. Nowadays, RET and dRET models provide a realistic characterization of vegetation volumes, including not only in-excess attenuation, but also scattering, diffraction and depolarization. Nevertheless, both approaches imply the characterization of the forest media by means of a range of parameters, and thus, the construction of a simple parameter extraction method based on propagation measurements is required. Moreover, when dealing with experimental data, two common problems must be usually overcome: the scaling of the vegetation mass parameters into different dimensions, and the scarce number of frequencies available within the experimental data set. This paper proposes the use of Artificial Neural Networks as accurate and reliable tools able to scale vegetation parameters for varying physical dimensions and to predict them for new frequencies. This proposal provides a RMS error lower than 1 dB when compared to unbiased measured data, leading to an accurate parameter extracting method, while being simple enough for not to increase the computational cost of the model.

[1]  Thuy Le Toan,et al.  Radiative transfer modeling of cross-polarized backscatter from a pine forest using the discrete ordinate and eigenvalue method , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[2]  Telmo R. Fernandes,et al.  Radiative Energy Transfer Based Model for Radiowave Propagation in Inhomogeneous Forests , 2006, IEEE Vehicular Technology Conference.

[3]  Rafael F. S. Caldeirinha,et al.  Retrieving Vegetation Reradiation Patterns by Means of Artificial Neural Networks , 2016, IEEE Antennas and Wireless Propagation Letters.

[4]  M. Al-Nuaimi,et al.  Measurements and prediction model optimisation for signal attenuation in vegetation media at centimetre wave frequencies , 1998 .

[5]  K. Sarabandi,et al.  A Physics-Based Statistical Model for Wave Propagation Through Foliage , 2007, IEEE Transactions on Antennas and Propagation.

[6]  Akira Ishimaru,et al.  Wave propagation and scattering in random media , 1997 .

[7]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[8]  I. Cuiñas,et al.  Modeling vegetation attenuation patterns: A comparison between polynomial regressions and artificial neural networks , 2016, 2016 IEEE International Symposium on Antennas and Propagation (APSURSI).

[9]  A. Seville,et al.  A generic narrowband model for radiowave propagation through vegetation , 2005, 2005 IEEE 61st Vehicular Technology Conference.

[10]  Werner Wiesbeck,et al.  Millimeter-wave scattering and penetration in isolated vegetation structures , 2000, IEEE Trans. Geosci. Remote. Sens..

[11]  J. A. Gay-Fernández,et al.  Analysis of the performance of vegetation barriers to reduce electromagnetic pollution , 2011 .

[12]  Constantine A. Balanis,et al.  Antenna Theory: Analysis and Design , 1982 .

[13]  C. Oestges,et al.  A Ray Based Method to Evaluate Scattering by Vegetation Elements , 2012, IEEE Transactions on Antennas and Propagation.

[14]  Telmo R. Fernandes,et al.  A Discrete RET Model for Millimeter-Wave Propagation in Isolated Tree Formations , 2005, IEICE Trans. Commun..

[15]  Benjamin Naumann,et al.  Learning And Soft Computing Support Vector Machines Neural Networks And Fuzzy Logic Models , 2016 .

[16]  Fawwaz T. Ulaby,et al.  Millimeter-wave bistatic scattering from ground and vegetation targets , 1988 .

[17]  R. Caldeirinha,et al.  RET input parameter estimation for a generic model of propagation through vegetation using excess attenuation and phase function measurements , 2003 .