Design of Negative Curvature Hollow Core Fiber Based on Reinforcement Learning

In negative curvature hollow core fibers (NCHCFs), light guidance is based on the capillary structure in the cladding. To achieve desirable fiber propagation properties, various designs of the capillary structure have been proposed in literature. However, the design process so far depends more or less on experience. In this article, we propose a reinforcement learning (RL) based method of systematically optimizing the capillary structure to achieve low average confinement loss (CL) for a given operating wavelength range and core radius. We use a recurrent neural network (RNN) to interactively study the properties of different capillary structures. The wavelength averaged CLs of the resulting designs are more than one order of magnitude lower than the lowest average CL of prior designs in literature. The same approach can be applied to search for optimum capillary structures in terms of other fiber propagation properties such as bending loss (BL), higher order modes extinction ratio (HOMER), overlap of the optical mode with the capillary structure, or a trade-off among these properties.