.png){kind=icon}
.png){kind=icon}
Return
New Algorithms for Enabling High-Dimensional Aperiodic Planar Antenna Array Optimization
.png){kind=icon}
OA
AI
DOI：10.1109/OJAP.2025.3637531.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
As 6G and beyond technologies are stepping into the spotlight, phased arrays are becoming increasingly promising candidates for agile antenna systems. In particular, aperiodic phased arrays provide many desirable traits for such technologies, such as reduced mutual coupling, sidelobe suppression, and grating lobe elimination. However, the design of such systems creates a complex, high-dimensional optimization space due to the competing traits. Traditional optimization methods often fail to scale well when applied to such high-dimensional problems. To this end, three promising algorithms that claim to perform well for such complex and large problems have been identified: Trust region Bayesian optimization (TuRBO), hybrid remora crayfish optimization algorithm (HRCOA), and the zeroth-order optimization toolbox (ZOOpt) featuring a sequential randomized coordinate shrinking classification algorithm (RACOS). The study allows maximal freedom in optimization through using free-floating elements and suggested minimum spacing through weights in the cost function. Two cases of 256 and 625 element arrays, 512 and 1250 variables, respectively, are studied using these three algorithms and compared against the covariance matrix adaptation evolutionary strategy (CMA-ES). The convergence characteristics of each algorithm are analyzed, and their robustness for application in high-dimensional antenna array optimization problems is examined.
Keywords:
Aperiodic
classification algorithms
optimization
particle swarm optimization
phased arrays
stochastic processes
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
I
IF:
3.6
Papers:
803
Citations:
1.6K