Unconventional and Irregular Clustered Arrays

Abstract

In this paper, two different structures based on fully and partially irregular clustered elements are presented to optimize the radiation patterns of the large arrays. In the Fully Irregular Clustered Elements (FICE) structure, all the elements of an ordinarily large linear array are divided into multiple irregular and unequal-size clusters, while in the Partially Irregular Clustered Elements (PICE) structure, only the elements those located at the edges of the ordinarily linear array are grouped into unequal-size clusters. In this structure, the central elements are left without clustering (i.e., excited individually). The PICE structure has several advantages over the FICE structure. Since the quantized amplitudes (i.e., discrete distribution) are used with the proposed clustered structures, the corresponding clustered array patterns are usually associated with the undesirable high periodic sidelobes. In order to overcome this problem, the elements in the clusters are distributed a periodically. Simulation results demonstrate the ability of the proposed two structures, FICE and PICE, to significantly reduce the high periodic sidelobe level to -35dB and -40dB respectively for an array with a number of elements N=100 and unequal-size clusters.

Beam Pattern Optimization Via Unequal Ascending Clusters

“…Each cluster has only one controller to feed the clusters in terms of amplitude and phase, or both. Although the desired beamwidth and directivity are still achieved, the use of the cluster architecture offers limited operating bandwidth and low reconfiguration capabilities due to the quantized distribution of the excitation across the aperture array [3]. For example, if a linear array is portioned into equal clusters, the radiation pattern will suffer from the appearance of grating lobes, and the number of these lobes increases with the growing number of elements in the cluster.…”

“…For example, if a linear array is portioned into equal clusters, the radiation pattern will suffer from the appearance of grating lobes, and the number of these lobes increases with the growing number of elements in the cluster. To bypass this problem, arrays made of irregularly arranged or polyomino-shaped clusters are used in large planar arrays [3], [4].…”

“…In the past years, several efficient methods have been proposed to address the problems of cluster synthesis. As far as we know, two types of methods based on evolutionary algorithms are used, such as genetic algorithm [2], [3] and particle swarm [5] algorithm. Mathematical analysis strategies, such as compressed sensing [6]- [8] are relied upon as well.…”

“…This leads to a lower reduction in the system's complexity. In [3], two structures based on the genetic algorithm are proposed as well. In the first structure, the radiating elements in the main array are divided into several irregular clusters, and in the second structure, only the elements located at the sides are divided into irregular clusters.…”

“…In this paper, the methods described in [2] and [3] have been extended and simplified by proposing fully and partially clustered architectures that are arranged in ascending order in terms of the number of elements in each cluster needed to build the desired radiation pattern under the applicable constraints. In the fully clustered array, the original linear array is divided into a group of clusters, with each one of them having an unequal number of elements.…”