Numerical computation of scattering from a perfectly conducting random surface
Abstract
An angular resolution (and therefore stability) criterion is derived in terms of the width of a window function, the incidence angle, the wavelength, and the angular variations of the average reradiated power. The effect of the actual width of a sample surface on computing a power of field estimate is also discussed. The total number of the sample surfaces required is established by applying the formula for the confidence interval or a chisquare variable. The onedimensional random surface used was generated by first generating a sequence of Gaussian variates. A technique described by Naylor et al. (1966) was employed to smooth these variates repeatedly to form a random surface. A description is presented of the scattered field computation. The parameters affecting the scattered field computation are considered in detail.
 Publication:

IEEE Transactions on Antennas and Propagation
 Pub Date:
 May 1978
 DOI:
 10.1109/TAP.1978.1141871
 Bibcode:
 1978ITAP...26..482A
 Keywords:

 Digital Simulation;
 Electric Conductors;
 Electromagnetic Scattering;
 Random Processes;
 Surface Roughness Effects;
 Angular Resolution;
 Magnetic Induction;
 Monte Carlo Method;
 Numerical Stability;
 Plane Waves;
 Radiation Distribution;
 Statistical Distributions;
 Communications and Radar