Telecommunications and Radio Engineering

ANALYSIS OF PULSED FIELDS REFLECTED FROM A LAYERED LOSSY MEDIUM USING ARTIFICIAL NEURAL NETWORK

ABSTRACT

A solution is found to the problem in determining the thickness of a layered medium through the use of the pulsed e.m. field. The artificial neural network (ANN) is used in the time domain to treat analytically the pulsed field reflected from a layered medium whose electric characteristics are close to those of human shin. The normal incidence of a time-shaped plane wave in the Gaussian form of is discussed. The reflected e.m. field is calculated by the finite-difference time-domain (FDTD) method. Use is made of the amplitude value of a reflected-field electric component. As an example, the network is trained to determine the thickness of one of the layers. The consistency of such determination is examined in the presence of interferences, experimental errors and slight deviations of the medium's electric parameters.

REFERENCES

1. Harmut, H., Nonsinusoidal waves in radiolocation and radio communication.

2. Dranina, Yu.P., Principles of neurolike simulation of geophysical object and processes.

3. Golovachev, D.A., Umnov, A.L., and Yashnov, V.A., The use of neural networks for interpreting interference location.

4. Kalan, R., Basic concepts of neural networks.

5. Shyrokorad, D., Dumin, O., and Dumina, O., Time domain analysis of reflected impulse fields by artificial neural network.

6. Dumin, O., Dumina, O, and Shyrokorad, D., Time domain analysis of fields reflected from model of human body surface using artificial neural network.

7. Drobakhin, O. and Doronin, A., Estimation of thickness of subsurface air layer by neuron network technology application to reflected microwave signal.

8. Alexin, S., Drobakhin, O., and Tkachenko, V., Reconstruction of permittivity profile for stratified dielectric material: Gel'fand-Levitan and Newton-Kantorovich methods.