Adaptive Multi-Group Differential Evolution Optimization for MIMO Antenna Design

Multiple-input multiple-output (MIMO) antenna designs require expensive full-wave electromagnetic simulations, which are time-consuming. To address this problem, a multi-group differential evolution optimization method fast MIMO antenna design is investigated focusing on implementation platforms and performance comparison with competing methods on test functions in this work. Experimental results show that (1) the adaptive multi-group differential evolution is implemented by exploiting "co-simulation" with MATLAB and CST, which allows seamless interaction between CST and MATLAB in the optimization process. (2) A test function with similar landscapes as the MIMO problem is used to illustrate the performance of the adaptive multi-group differential evolution algorithm. Compared with traditional differential evolution, the adaptive multi-group differential evolution not only increases the search speed but also maintains the ability to escape the local extremum.