Oral: In-situ Wave Control in Complex Systems Based on Cavity-shaped Inverse-Design Methods

We introduce an algorithm that identifies optimal cavity and wavefront parameter configurations in-vivo which lead to extreme control of wave propagation in complex scattering environments. The method allows us to establish a variety of specific modalities like targeted mode transmission, perfect absorption, invisibility of the scattering domain, etc. Our approach is inspired by computational optimization techniques from inverse-design. Specifically, it utilizes an adjoint method that adjusts device and injected wavefront parameters through large-scale, gradient-based optimization with many degrees of freedom, requiring only two full-field measurements of a forward and backward wave propagation. This technique enables rapid gradient computation without the need for neural-network training, unlike machine learning algorithms or brute-force optimization methods like Bayesian or surrogate algorithms. Our results establish this protocol as a versatile scheme for creating reconfigurable hot and cold spots in complex enclosures (e.g., buildings or vessels), with applications in next-generation telecommunications, long-range wireless power transfer, and electromagnetic warfare.