Sparse and irregular sampling in array processing

Rule-based thinning and element placement methods are first discussed. This includes random arrays, binned random arrays, periodic thinning, element shadowing properties, and fractal arrays. The algorithmic optimization of layouts of 1-D and 2-D sparse arrays is presented and it is empirically derived that the peak sidelobe level for an array with K active elements is proportional to 1/K for a 1-D array and 2/K for a 2-D array. The constant of proportionality depends strongly on the width of the mainlobe relative to that of the full array. With an increase of 50%, the constant is 1.5