Propagation problems connected with special source structures

Abstract

In my preceding talk I treated phenomena the main characteristics of which were already apparent from effects produced by an infinitesimal point source. Let me now [Walker‐Ames Lecture at the University of Washington, June 25, 1979] consider cases in which, on the contrary, the structure of an extended source is essential for the problem in question. Having some experience in problems dealing with rather different applications, such as radio waves, optical and electron‐optical imagings, and plasma physics problems, it struck me that fundamental phenomena occurring in any of them are often discussed while overlooking results that have been obtained in one of the related fields. Today I would like to emphasize the existence of some common features in all of these problems, all having to do with sources of finite dimensions. This may be illustrated by considering a special parameter with the dimension of a length and connected with the size of the source, which plays a role in all of them. As a matter of fact, it is just the finiteness of the source which implies that some length (however, not necessarily just the size itself of the source) becomes most essential. This is at once clear in the case of the three‐dimensional source representing a transmitting antenna system. In the immediate vicinity of the latter the field depends mainly on the nearest parts of the system, but at larger distances the effects of all parts interfere, and we arrive in the so‐called ‘wave zone.’ The transition occurs at distances D at which the various parts of the source, with separations L perpendicular to the propagation direction, can reach the point of observation along path lengths differing by not more than about the operating wavelength. According to the relation we find the expression for this critical distance. Of course, this transition distance is rather vague, since the dominating scale length L of the antenna, here introduced, is not well defined. However, it is well known that the approximations needed in order to determine radiation diagrams are only applicable in the wave zone which is just fixed, apart from the almost trivial conditions D ≫ λ and D ≫ L by the property D > Dcr = L2/λ.