Reducing the number of elements in a pattern reconfigurable antenna array by the multi‐task learning

Abstract

The synthesis of pattern reconfigurable antenna array with as few elements as possible can find wide applications in radar tracking, biomedical imaging, satellite and ground communications, and remote sensing applications. In this study, an efficient method for multi-task learning (MTL) is exploited to the design of sparse pattern reconfigurable antenna array. Toward this end, the design of sparse and pattern reconfigurable antenna array is reformulated as an equivalent problem of multi-matrices linear regression and iterative shrinkage threshold method for MTL is utilised to obtain jointly optimal design of positions and the excitations of the radiating elements for multi-pattern synthesis, in which compromise between the array sparseness and pattern matching is achieved. Some numerical simulations are presented to assess the efficiency of the proposed method and the synthesis performance comparisons with matrix pencil method and genetic algorithm are also performed to demonstrate the superiority of the proposed algorithm over traditional algorithm in this literature.