Abstract
The performance of silicon photonic devices is highly sensitive to fabrication-induced deviations in waveguide core dimensions from their design values. This motivates the development of optimization algorithms that explicitly account for such uncertainties and can identify robust designs, that is, designs whose performance is insensitive to fabrication-induced variations. In this work, a novel scheme based on robust Bayesian optimization (RBO) is proposed, which leverages well-established robustness measures and aims to minimize deviations from nominal performance under input uncertainty. The study focuses on an S-bend directional coupler (DC), but the proposed methodology is general and can be applied to a wide range of passive and active photonic components. Related Research Topics
Related Projects
|
|
