Improved Fox Optimization Algorithm for the Pattern Synthesis of Sparse Linear Array Antennas

ABSTRACT This paper addresses the limitations of current swarm intelligence optimization algorithms, which often suffer from poor solution accuracy and slow convergence when applied to the challenges of sidelobe level suppression and null control for the pattern synthesis of sparse linear array antennas. To overcome these issues, we propose the fox optimization algorithm based on cubic chaotic mapping, adaptive spiral flight, and Gaussian stochastic wandering strategy (CSGFOX). First, cubic chaotic mapping is used to initialize the population, enhancing its diversity. The spiral flight strategy is then employed to adaptively update the fox's position, thereby improving the algorithm's global exploration capability. Finally, a Gaussian stochastic wandering strategy is incorporated, balancing the algorithm's local search ability with its global search capability. To validate the performance of the proposed algorithm, we evaluated it using benchmark functions. Additionally, the algorithm is applied to the optimization of sparse linear array antennas. The simulation results demonstrate that the proposed algorithm exhibits faster convergence and higher solution accuracy in suppressing the sidelobe level and controlling the nulls of the directional pattern of linear array antennas, compared to traditional optimization methods, including the fox algorithm, genetic algorithm, gray wolf algorithm, whale algorithm, and particle swarm optimization algorithm. At the same solution quality threshold, CSGFOX converged 61.5% faster than the original algorithm, achieving a 100% success rate and significantly outperforming FOX (95%) and other algorithms (10%–65%), thus confirming the proposed algorithm's effectiveness.