Contents I. Introduction
With the modern complex electromagnetic environment and the increasing requirements for radar detection, the development of large antenna arrays increases the complexity, cost and power consumption of antenna systems. The antenna array is often sparse and optimally arrayed to satisfy the desired array pattern with the minimum number of elements to reduce the weight and cost of the radar system. Many optimization algorithms have been proposed in the last sixty years to synthesize such arrays [1]–[5], such as follows: analytical methods, intelligent optimization methods, fast Fourier transform methods, matrix beam methods, compressed sensing methods, etc. Despite the success of these methods, most of them can only be used to synthesize a sparse array with a single-pattern. They will be invalid in the multiple-pattern case since the best element positions usually change with different patterns. Only a few techniques have been proposed for the synthesis of a non-uniform sparse array with multiple-pattern[6]–[9]. Combined with the theory of compressed sensing, this paper is aimed at proposing a new method to synthesize a multiple-pattern array with as few elements as possible.
