Computational Electromagnetics

Catalog description

Development of modern technologies such as 5G wireless networks, terabit silicon photonic optical interconnects, and optical computing relies on a deep understanding of the underlying electromagnetic principles governing their operation. Engineers must rely on numerical simulations to predict and model their behaviors when designing these systems. This class will give a graduate-level introduction to computational methods for solving partial differential equations describing physical phenoma that commonly arise in the real world. Primarily finite difference methods, in both the time and frequency domains, will be covered, although integral equation-based approaches and finite element methods will be introduced well. The course will also introduce modern inverse design approaches for automating the design of new electromagnetic structures, including gradient-based methods and the adjoint method, as well as global search strategies. Numerous examples drawing from practical applications, primarily in electromagnetics, will be presented for solving relevant real-world problems, including radiating antennas for wireless communication, dielectric waveguides for nanophotonic integrated circuits, as well as electromagnetic scattering from arbitrary dielectric objects for applications in radar scattering and remote sensing. Open to PhD, MS, and advanced undergraduate students.