Optimization of Module Partition Based on Weighted Clustering Method for Wideband Irregular Subarrayed Array

• Published in: IEEE transactions on antennas and propagation 2023-07, Vol.71 (7), p.1-1
• Main Authors: Chen, Ji-yuan, Xu, Zhen-hai, Dong, Wei, Xiao, Shun-ping
• Format: Article
• Language: English
• Subjects: Arrays; Broadband; Cluster analysis; Clustering; Complexity; minimum module quantity; module partition; Modules; Optimization; Optimization models; Pattern matching; Vector quantization; weighted clustering method; Wideband irregular subarrayed array
• ISSN: 0018-926X; 1558-2221
• DOI: 10.1109/TAP.2023.3270380

In wideband beam scanning with module-level time delay and subarray-level phase shifting methods in irregular subarrayed planar arrays, obtaining the optimal module partition is a critical challenge. In this paper, we propose a module partition optimization method based on the weighted clustering method. Firstly, we demonstrate that pattern matching is equivalent to the subarray-level excitation matching principle. Then we transform the complex optimization model into a real clustering optimization framework based on the second-order Taylor approximation, which reduces the computational complexity. Next the weighted clustering method is proposed to reduce the matching error by modifying the membership rule and cluster center of K-means clustering, ultimately obtaining the optimized module partition scheme. Additionally, we propose a minimum module quantity estimation method to explore the relationship between array bandwidth and module quantity. The proposed methods are applicable to both uniform and non-uniform weighting cases. The effectiveness and potential of the proposed method are verified by different numerical experiments and comparison of state-of art method.