Efficient techniques for near-field computation and subreflector analysis

A technique for near field computation is proposed. Integral representations of the radiating near field are obtained by making Fresnel approximations in the phase, magnitude, and direction of the exact field expression. Advantages of this method over other existing methods are shown, and details for implementing the method for both planar aperture and reflector antennas are given. The problem of reflection from smooth convex surfaces is considered. A generalization of the conventional geometrical optics is presented and closed form expressions are derived. The derivation of this generalized geometrical optics formula is based on the Fresnel transformation between two planes, one of which contains the specular point, while the other passes through the observation point. An efficient technique for computing the reflected field from a smooth numerically specified surface is developed. Both surface interpolation and search for the specular point are bypassed. The method starts by computing the reflected rays off the surface at the numerically specified points. The observation point is associated with the pencil composed of the reflected rays nearest to it.