Nanophotonic Device Design from Conventional to Mathematical Optimization-assisted Method

As photonic integrated circuits become more advanced with different multiplexing techniques, the demand for sophisticated nanophotonic devices increases. The conventional nanophotonic design method is showing limitations with this trend. Firstly, as the nanophotonic devices become more sophisticated, only highly experienced photonic designers who have a deep understanding of photonics can propose ideas to fulfill these demands. Secondly, the sizes of conventionally designed devices tend to increase to meet the multifunctional requirements. Novel mathematical optimization-assisted methods are promising tools to significantly reduce the size of silicon photonic components while maintaining their functionality and performance. These methods have been investigated intensively over the past decades. In this talk, the conventional nanophotonic device design methods are reviewed. The mathematical description of the nanophotonic device design is given. Then, the mathematical optimization-assisted design methods are discussed, including heuristic methods and the adjoint method. Finally, the possible future directions of mathematical optimization-assisted methods will be discussed, including potential limitations and problems that still need to be solved. Speaker(s): Guowu Zhang, Room: TR0070, Bldg: Trottier Engineering Building, McGill University, 3630 University Street, Montreal, Quebec, Canada, Virtual: https://events.vtools.ieee.org/m/310881