A Benchmark Test Suite for Antenna S-Parameter Optimization

Performance evaluation of optimization-based antenna design methods is essential. It is desirable to develop an efficient test suite with characteristics of antenna structures to replace the high-computational-cost electromagnetics (EM)-based evaluation. In this paper, four typical antennas with respect to four typical objectives based on EM S-parameter simulations are systematically investigated. Based on the investigation results, we propose a performance evaluation test suite allowing efficient and versatile benchmarking for antenna S-parameter optimization algorithms. The test suite consists of eight analytical functions (four unimodal functions, two multimodal functions, and two compositional functions). Each function matches a typical landscape of the EM simulated antenna problems in the two-dimensional case. Seven of the functions are scalable to high dimension problems. The performances of six widely used global optimization algorithms are evaluated using the test suite. The performance evaluation results are compared to those obtained using the two EM-based antenna problems (a single antenna and a multi-antenna). The results using the proposed test suite for all six optimization algorithms are consistent with those of the real EM antenna problem. The time cost for performance evaluation using the test suite is only a fraction of the cost using the real EM antenna problems.