A mathematical theory of antenna arrays with randomly spaced elements

Various probabilistic properties of a large antenna array with randomly spaced elements have been studied. It is found that for almost all cases of practical interest the required number of elements is closely related to the desired sidelobe level and is almost independent of the aperture dimension, the resolution (or the beamwidth) depends mainly on the aperture dimension, and the directive gain is proportional to the number of elements used if the average spacing is large. As a consequence the number of elements required is considerably less than that with uniform spacings. Starting with a given number of elements and a given aperture size, it is possible to improve the resolution by a factor of ten, a hundred, or more by spreading these elements over a larger aperture with little risk in obtaining a much higher sidelobe level and a lower directive gain. In fact, this method offers a solution which is optimum in a certain statistical sense, i.e., all sidelobes are of equal level with equal probability. In addition, this analysis also gives a simple estimate of the sidelobe level of most nonuniformly spaced antenna arrays. In a number of such arrays studied by various investigators with high speed computers, the agreement found is remarkable.