Bendable resin having patterned elements for improving penetration of wireless communication signals

A multilayer structure attachable to a substrate and configured to enhance penetration of an incidental radio wave from a first region through the substrate to a second region by forming one or more communication signal beams in the second region, the multilayer structure comprising:

• a base layer of a first material;
• a transparent film attached to the base layer for interfacing the substrate and lowering dielectric constant of the substrate;
• and
• one or more patterned elements each formed by providing a meta-pattern of a second material on the base layer or the transparent film;
• wherein:
• the first and second materials are different and selected from the group consisting of a dielectric material and a metallic material;
• and
• each individual patterned element is configured to tilt the incidental radio wave to form the one or more communication signal beams, wherein each individual communication signal beam is beam-focused at a predetermined focal point or a predetermined focal area in the second region.

A method for generating a meta-pattern for forming a patterned element on a bendable resin, the bendable resin attachable to a substrate and configured to enhance penetration of an incidental radio wave from a first region through the substrate to a second region by forming one or more communication signal beams in the second region, the method comprising the steps of:

• obtaining a set of specifications associated with a predetermined focal point or a predetermined focal area in the second region;
• determining a real image based on actual data or transmission results;
• randomly setting input parameters and generating, by a pattern generator, a plurality of feature patterns arranged in rows and columns repeatedly to form an emulation image;
• determining, by a discriminator, a discriminator loss and a generator loss based on a discrimination in a signal strength between the real image and the emulation image;
• iterating the input parameters to generate other novel emulation images for determining the discrimination in the signal strength;
• and
• determining the meta-pattern for the set of specifications with smallest values in the discriminator loss and the generator loss.