Automatic AI-Driven Design of Mutual Coupling Reducing Topologies for Frequency Reconfigurable Antenna Arrays
Jiahao Zhang, Mobayode O. Akinsolu, Bo Liu, Guy A. E. Vandenbosch
Abstract
An automatic artificial intelligence (AI)-driven design procedure for mutual coupling reduction and a novel isolator are proposed for a frequency reconfigurable antenna array. The design process is driven and expedited by the parallel surrogate model-assisted differential evolution for antenna synthesis (PSADEA) method. The reconfigurable array element can switch its operation between the 2.5 GHz ISM band and the 3.4 GHz WiMAX band. By introducing the proposed isolator, the mutual coupling in the higher and lower band is reduced by 8 and 7 dB, respectively. The reconfigurable array was prototyped, and measurements agree well with simulations, verifying the validity of the proposed concept. Although used for a specific antenna in this communication, the proposed AI-driven design strategy is generic and can easily be employed for other array topologies.