Linear Array Thinning Exploiting Almost Difference Sets

This paper describes a class of linear thinned arrays with predictable and well-behaved sidelobes. The element placement is based on almost difference sets and the array power pattern is forced to pass through N uniformly-spaced values that, although neither equal nor constant as for difference sets, are a-priori known from the knowledge of the aperture size, the number of active array elements K, and the features of the correlation function. Such a property allows one to predict the bounds of the confidence range of the peak sidelobe of the admissible arrays obtainable through simple shift operations on a binary sequence. The expected peak sidelobe performances turn out to be comparable with those from difference sets, even though obtainable in a wider set of array configurations, and improved in comparison with cut-and-try random-placements.