Human-Guided Antenna Design
Optimization-based approaches to antenna design have enjoyed limited success. The task is often computationally intractable and it is often difficult to capture all relevant design issues and trade-offs in a single mathematical objective function. Therefore, human experts typically specify and refine antenna designs by hand, using computers only to evaluate their candidate designs by simulation. In this project we propose a middle ground between this traditional approach and fully automatic optimization - a human-guided interactive system.
Background & Objective: The idea of using computer-based optimization for design tasks has been applied to many problems, including antenna design. However, this idea does not always work well: the optimization problems are often intractable and it is often impossible to consider all relevant design criteria in the optimization process. In this project we propose that the computer be used differently, leaving the task of choosing a final design from the computer-generated sampling to the human user, who can apply experience and judgment to recognize and then refine the most useful antenna design. Thus the "generation" of the candidate set, and "visualizing" the set are separated into two tasks.
Technical Discussion: FY05 efforts built on our FY04 results, again addressing the issue of phased-antenna arrays with non-uniform spacing. The FY04 results provided a statistical background and framework for such design. In FY05 we used machine learning techniques to explore the space of antennas with the goal of rapid optimization. While in previous approaches we used exhaustive searches, which carry a large computational burden, we have now developed a new approach that results in significantly faster and "smarter" search, and also low computational complexity for fieldable applications. The trade-off for this new approach is the inability to guarantee that our results are truly optimal. However, using the statistics obtained from our FY04 project, we have shown that our new search results are only slightly sub-optimal. Such sub-optimal results are acceptable in physically realizable applications given the increased computational efficiency of the new methods.
Contact: Kent Wittenburg
| Technical Reports: | |
| Sidelobe Minimizaton of Uniformly-Excited Sparse Linear Arrays using Exhaustive Search and Visual Browsing | |
| QueryLines: Approximate Query for Visual Browsing | |
| Exhaustive Generation and Visual Browsing for Radiation Patterns of Linear Array Antennas | |
| Semi-Automatic Antenna Design Via Sampling and Visualization | |
Technology Area: Sensor and Data Systems
Modification Date: June 12, 2008
