Sidelobe reduction of random arrays by element position and frequency diversity

The high sidelobes of random, thinned arrays can be reduced through the use of diversity techniques. Element position diversity and frequency diversity are considered in this paper. Image artifacts due to the high sidelobes change their locations from image to image when the element positions are altered or the operating frequency is changed. Superimposing or averaging images tends to build up stable, correct portions of an image while reducing, by smoothing, the image artifacts. In principle, all the sidelobe crests can be reduced to the average background level and all the troughs in the side radiation pattern will rise to this level. The theory, supported by simulation experiments, indicates that dual position diversity reduces the sidelobe level by 2 to 2 1/2 dB, depending upon array size. Higher order position diversity reduces the sidelobe level several dB further. Under frequency diversity it is found that when
Q
, the reciprocal of the fractional bandwidth, is less than the array size
L/\lambda
(measured in units of wavelength), the peak sidelobe power level is approximately
N^{-1}\lnQ
, where
N
is the number of array elements. For larger values of
Q
, the level asymptotes approximately to
N^{-1}\ln(L/lambda)
.