CAREER: EleXIR: A Top-Down Approach to Research and Learning in Efficient Hierarchical Electromagnetic Simulation Methods for Complex Structures and Systems

职业：EleXIR：复杂结构和系统的高效分层电磁仿真方法的自上而下的研究和学习方法

• 批准号: 0093102
• 负责人: Vikram Jandhyala
• 金额: --
• 依托单位: University of Washington
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-06-15 至 2008-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0093102&HistoricalAwards=false
• 关键词: CAREER; EleXIR; Top; Down; Approach

0093102JandhyalaComputational electromagnetics (CEM) is a diverse field that sees several new applications each year. Owing to advances in computational algorithms and hardware, it has become feasible to analyze progressively more complex electromagnetic (EM) problems. Indeed, the development of future generations of computers will itself be heavily dependent on the availability of efficient, integrated CEM tools for chips and interconnects.With the enhanced global emphasis on high-speed systems of unparalleled complexity, it is evident that EM effects are becoming increasingly dominant. Wireless networks are expanding at a dramatic rate in the United States and in several parts of the world. High-speed systems-on-chip are now the focal points in myriad commercial, defense, and research areas. Several critical current and future domains with unprecedented potential including the next-generation internet, battlefield communications and intelligence, and large-scale distributed computing do or will rely on complex wireless networks and systems-on-chip. These areas will have a massive impact on the increasingly information-based economy and defense of the United States, and on the global high-technology economy of the future.The design and development needs of these complex systems and structures have led to a critical requirement for accurate, hierarchical EM analyses. These simulations are necessitated in order to develop new designs and paradigms, and to save on the exorbitant costs and time requirements of iterated physical prototyping and testing. The sheer complexity of the simulations is such that the possibility of achieving even moderately accurate results with reasonable computing resources and time did not exist until recently.While evolutionary advances in CEM are omnipresent, the Electromagnetic-Simulation: Instruction and Research (EleXIR) approach in this proposal presents paradigm-shifting, physics-based, revolutionary CEM methodologies that exhibit the potential to render feasible, over the next few years, the accurate, efficient, and integrated EM simulation of large-scale structures and systems. These approaches will lead to the development of advanced tools with the ability to perform rigorous and automated three-dimensional full-wave EM simulation with the ease and transparency of present-day circuit-level modeling. The resulting advances will have a strong potential to affect several simulation aspects of present and future EM applications, including wireless-based network and system-on-chip issues, such as: Propagation through random media and scattering from rough surfaces Signal integrity analysis for high-speed radio-frequency circuits and systems-on-chip Computation of radar cross sections of complex targets Design and analysis of novel antenna arrays Analysis of modem electromechanical, optical, and quantum devicesThese applications, while not being exhaustive, are extensive and varied. To address these, a unified class-oriented approach is proposed that relies on physics-based, redundancy-extracting approaches to hierarchical integral-equation modeling of EM problems. A combination of efficient integral equation formulations, fast algorithms tuned to the physics of specific classes of problems, and hierarchical and reduced-order modeling schemes, is the paradigm presented here for the development of revolutionary new CEM tools for several system-on-chip and wireless-related EM applications over the next few years. Along with the advances in such techniques comes the necessity and challenge of creating a pool of talent with the critical mass necessary for sustainable research and development in these high-technology areas, which will be addressed through new integrated curricula incorporating new CEM methods in a top-down manner in areas of interest i.e. classical EM, high-speed circuits, wireless communications, and devices. These crucial requirements in research, instruction, and technology-transfer that are of strategic importance to the United States and to global economics form the focus of the EleXIR approach proposed here.The general goals of EleXIR include the development, teaching, and technology transfer of seamless topdown approaches to modem CEM in high-technology areas where it is necessitated, or will be necessitated in the near future, including wireless and high-speed circuit applications.***

0093102Jandhyala计算电磁学（CEM）是一个多样化的领域，每年都有几个新的应用。 由于计算算法和硬件的进步，它已成为可行的分析越来越复杂的电磁（EM）问题。 事实上，未来几代计算机的发展本身将严重依赖于芯片和互连的高效集成CEM工具的可用性。随着全球对无与伦比的复杂性的高速系统的重视，EM效应显然正变得越来越占主导地位。 无线网络在美国和世界上的几个地方正以惊人的速度扩展。 高速片上系统现在是无数商业、国防和研究领域的焦点。 当前和未来的几个关键领域具有前所未有的潜力，包括下一代互联网、战场通信和情报以及大规模分布式计算，这些领域都依赖于或将依赖于复杂的无线网络和片上系统。 这些领域将对美国日益信息化的经济和国防以及未来的全球高科技经济产生巨大影响。这些复杂系统和结构的设计和开发需求导致了对精确、分层电磁分析的关键要求。 这些模拟是必要的，以开发新的设计和范例，并节省过高的成本和时间要求的迭代物理原型和测试。 模拟的复杂性是如此之大，以至于直到最近才有可能利用合理的计算资源和时间获得中等精度的结果。虽然CEM的进化进步无处不在，但电磁模拟：教学和研究（EleXIR）的方法在这个建议提出了范式转变，基于物理，革命性的CEM方法，表现出潜在的渲染可行，在接下来的几年里，精确，高效，大型结构和系统的综合电磁仿真。 这些方法将导致先进的工具的开发，能够执行严格的和自动化的三维全波电磁仿真与当今电路级建模的方便性和透明度。 由此产生的进步将有很大的潜力影响当前和未来EM应用的几个仿真方面，包括基于无线的网络和片上系统问题，例如： 通过随机介质的传播和粗糙表面的散射高速射频电路和片上系统的信号完整性分析复杂目标的雷达散射截面计算新型天线阵列的设计和分析现代机电、光学和量子器件的分析这些应用虽然不是详尽无遗的，但范围广泛且多种多样。 为了解决这些问题，提出了一个统一的面向类的方法，依赖于基于物理的，冗余提取方法分层积分方程建模的EM问题。 一个有效的积分方程公式，快速算法调整到特定类别的问题的物理，分层和降阶建模方案的组合，是在这里提出的几个系统芯片和无线相关的EM应用程序在未来几年的革命性的新的CEM工具的发展的范例。 随着这些技术的进步，沿着的是创造一个人才库的必要性和挑战，在这些高科技领域的可持续研究和发展所需的临界质量，这将通过新的综合课程，包括新的CEM方法，在感兴趣的领域，即经典EM，高速电路，无线通信和设备，以自上而下的方式解决。 这些研究，教学和技术转让的关键要求，是具有战略意义的美国和全球经济形成的EleXIR方法的重点在这里提出。EleXIR的总体目标包括开发，教学和技术转让的无缝自上而下的方法，以现代CEM在高科技领域，它是必要的，或将在不久的将来，包括无线和高速电路应用。