No abstract

No abstract

A cosecant-squared radiation pattern synthesis for a planar antenna array by using the genetic algorithm (GA) is presented. GA makes array synthesis flexible to achieve two desired features, namely, low peak side lobe level (PSLL) and small deviation (ripples) in the shaped beam region. In order to obtain a desired csc 2 pattern with the PSLL constrained, GA optimizes both the excitation amplitude and phase weights of the array elements. Dynamic range ratio (DRR) of the excitation amplitudes is improved by eliminating the weakly excited array elements from the optimized array without distorting the obtained pattern. To illustrate the effectiveness and advantages of GA, the beam pattern with specified characteristics is obtained for the same array by using particle swarm optimization (PSO). Results show that the performances of GA and PSO are comparable when dealing with small-tomoderate planar antenna arrays. However, GA significantly outperforms PSO on large arrays. Moreover, numerical results reveal that GA is superior to PSO in terms of cost function evaluation and statistical tests.

Achieving robust and fast two-dimensional adaptive beamforming of phased array antennas is a challenging problem due to its high-computational complexity. To address this problem, a deep-learning-based beamforming method is presented in this paper. In particular, the optimum weight vector is computed by modeling the problem as a convolutional neural network (CNN), which is trained with I/O pairs obtained from the optimum Wiener solution. In order to exhibit the robustness of the new technique, it is applied on an 8 × 8 phased array antenna and compared with a shallow (non-deep) neural network namely, radial basis function neural network. The results reveal that the CNN leads to nearly optimal Wiener weights even in the presence of array imperfections.