Phys.org July 8, 2024
Large-scale and high-dimensional permutation operations are important for various applications in telecommunications and encryption. Researchers at UCLA, California, used all-optical diffractive computing to execute a set of high-dimensional permutation operations between an input and output field-of-view through layer rotations in a diffractive optical network. In this design every diffractive layer had four orientations: 0 deg, 90 deg, 180 deg and 270 deg. Each unique combination of these layers represented a distinct rotation state, tailored for a specific permutation operation. A K-layer rotatable diffractive design could perform up to independent permutation operations. The original input information could be decrypted by applying the specific inverse permutation matrix to output patterns. This design was demonstrated by approximating 256 randomly selected permutation matrices using = 4 rotatable diffractive layers. To further enhance its multiplexing capability, input polarization diversity was also utilized. The design was experimentally validated using terahertz radiation and 3D-printed diffractive layers, providing a decent match to numerical results. According to the researchers their design is particularly useful due to its mechanical reconfigurability, offering multifunctional representation through a single fabrication process… read more. Open Access TECHNICAL ARTICLE
Schematic of a reconfigurable multiplexed diffractive deep neural network … Credit: Laser and Photonics Reviews, 04 July 2024