Cognition-Driven Modeling and Optimization Technology for Multi-Disciplinary Design of High-Frequency Components and Communication Subsystems

我们拟议的研究将致力于认知驱动的建模和设计RF/微波和太赫兹组件和多千兆位通信电路的新兴技术所需的规模和临界质量。 基于我们在神经网络和空间映射建模和优化方面的高度认可的工作，所提出的认知驱动方法将通过正式涉及思维方面（人类直觉，决策等），利用新的机会来提高自动建模和设计的有效性，更好地利用经验模型，等效电路，半解析方程，内部电磁结构信息等专业知识和知识，以及多物理场设计和仿真。 将开发新技术，通过详细的电磁场和多物理场模拟将认知驱动算法形式化，以实现快速准确的建模和优化解决方案。 该技术将进一步推广，以满足多学科建模和设计需求，如电气/电磁/热/应力/EMC设计。 我们的目标是创造一个新的临界质量，用于开发一种综合技术，作为下一代高频电子设备、组件和子系统的建模和设计优化的引擎，以帮助大幅提高无线和有线通信和信息处理系统设计的生产力。

Our proposed research will be dedicated to the scale and critical mass necessary for the emerging technology of cognition-driven modeling and design of RF/microwave and terahertz components and multi-giga-bit communication circuits. Built-on our highly recognized work on neural network and space mapping modeling and optimization, the proposed cognition-driven approach will elevate the state-of-the-art of high-frequency electronic modeling and design into a new level by formally involving aspects of thinking (human intuition, decision-making etc.), with new opportunities to increase the effectiveness of automated modeling and design, to better harness expertise and knowledge such as empirical models, equivalent circuits, semi-analytical equations, internal electromagnetic structural information, together with multiphysics design and simulation. New techniques formalizing the cognition-driven algorithms with detailed electromagnetic and multiphysics simulations will be developed to achieve fast and accurate modeling and optimization solutions. The technology will be further generalized to address multi-disciplinary modeling and design needs, such as, electrical/electromagnetic/ thermal/stress/EMC design. Our goal is to create a new critical mass for developing a comprehensive technology to serve as the engine for modeling and design optimization of next generation of high-frequency electronic devices, components and subsystems to help dramatically increase the productivity of wireless and wireline communications and information processing systems design.