Metasurface design and placement
A method of designing a passive metasurface system within an environment includes receiving a three-dimensional model of the environment including one or more transmitter locations and one or more target locations, and determining one or more metasurface designs and placements to achieve a given objective. The method may also include computing a received signal strength at the one or more target locations based on the three-dimensional model of the environment, computing the received signal strength at the one or more target locations based on the three-dimensional model of the environment and the one or more metasurface designs and placements, and/or calculating ray traces and/or using machine learning to determine the received signal strength at the one or more target locations on the three-dimensional model of the environment and the one or more metasurface designs and placements.
1 . A method, implemented by a computing device, of designing a passive metasurface system for deployment within an environment, the method comprising:
receiving, by the computing device, a three-dimensional model of the environment including one or more transmitter locations and one or more target locations; and
determining, using the three-dimensional model, at least one structural design of one or more passive metasurfaces of the passive metasurface system and at least one placement of the one or more passive metasurfaces within the environment to achieve a given objective during the deployment, wherein the objective is to optimize the designs of low-cost passive metasurfaces.
2 . The method of claim 1 , further comprises:
computing a received signal strength at the one or more target locations based on the three-dimensional model of the environment.
3 . The method of claim 2 , further comprises:
computing the received signal strength at the one or more target locations based on the three-dimensional model of the environment and the at least one structural design and the at least one placement.
4 . The method of claim 3 , further comprises:
simulating rays that propagate from the one or more transmitter locations to locations within the environment, wherein at least some of the rays are simulated to propagate through the one or more passive metasurfaces within the environment and to yield attenuation and phase of each ray propagating through the one or more passive metasurfaces; and
computing the received signal strength at the one or more target locations for each ray arriving at each target location within the environment based on the attenuation and the phase of each ray propagating through the one or more passive metasurfaces.
5 . The method of claim 4 , further comprising:
measuring the received signal strength at one or more locations in the environment, wherein the environment includes one or more objects; and
starting from initial propagation coefficients, tuning a propagation coefficient for each object in the three-dimensional model to minimize error between the received signal strength measured at each object in the environment and the received signal strength simulated at each object for at least some of the rays.
6 . The method of claim 5 , wherein the propagation coefficient for each object is initialized based on a material type detected on an exterior surface of the object using computer vision.
7 . The method of claim 3 further comprising:
using machine learning to determine the received signal strength at the one or more target locations on the three-dimensional model of the environment and the at least one structural design and the at least one placement.
8 . The method of claim 1 , wherein structural design configurations of the one or more passive metasurfaces include metallic pattern distributions within the one or more passive metasurfaces, and further comprising:
tuning positions of the one or more passive metasurfaces placed within the environment, and the structural design configurations of the one or more passive metasurfaces placed within the environment to achieve the given objective based on placement of an access point within the environment and on a codebook of transmission signal properties of signals transmitted by the access point.
9 . The method of claim 1 , wherein structural design configurations of the one or more passive metasurfaces include metallic pattern distributions within the one or more passive metasurfaces, and further comprising:
tuning positions and the structural design configurations of the one or more passive metasurfaces to achieve the given objective.
10 . The method of claim 9 , wherein the structural design configurations of the one or more passive metasurfaces include metallic pattern distributions within the one or more passive metasurfaces, and further comprising:
tuning positions and the structural design configurations of the one or more passive metasurfaces placed within the environment and codebooks at a transmitter and/or a receiver to achieve the given objective.
11 . The method of claim 9 , wherein the structural design configurations of the one or more passive metasurfaces include metallic pattern distributions within the one or more passive metasurfaces, and further comprising:
tuning positions and the structural design configurations of the one or more passive metasurfaces placed within the environment and codebooks and positions of a transmitter and/or a receiver to achieve the given objective.
12 . The method of claim 9 , wherein tuning includes a search based on simulated annealing, Bayesian optimization, reinforcement learning, or a genetic algorithm.
13 . The method of claim 1 , further comprising:
tuning, based on placement of an access point within the environment and on a codebook of transmission signal properties of signals transmitted by the access point, structural design configurations of the passive metasurface system within the environment to identify a particular design configuration to achieve the given objective.
14 . The method of claim 1 , wherein determining the at least one structural design includes selecting metallic pattern distributions of unit cells of the one or more passive metasurfaces.
15 . A method, implemented by a computing device, of designing a passive metasurface system for deployment within an environment, the method comprising:
receiving, by the computing device, a three-dimensional model of the environment including a location of an access point and a location of a target point;
calculating, using the three-dimensional model, rays that propagate from the access point to locations within the environment, wherein at least some of the rays are simulated to propagate through one or more passive metasurfaces within the environment and to yield attenuation and phase of each ray propagating through the one or more passive metasurfaces;
computing, using the three-dimensional model, a received signal strength at the location of the target point for each ray arriving at the target point based on the attenuation and phase of each ray propagating through the one or more passive metasurfaces; and
determining structural design configurations of the one or more passive metasurfaces within the environment based on the received signal strength computed at the location of the target point for each ray.
16 . The method of claim 15 , wherein the three-dimensional model of the environment further includes one or more objects including the one or more passive metasurfaces within the environment, and further comprising:
measuring a received signal strength at each passive metasurface of the one or more passive metasurfaces within the environment;
computing a received signal strength at each object of the one or more objects included in the three-dimensional model; and
tuning a propagation coefficient for each passive metasurface of the one or more passive metasurfaces to minimize error between the received signal strength measured at each passive metasurface and the received signal strength computed at each object of the one or more objects.
17 . The method of claim 16 , wherein the structural design configurations of the one or more passive metasurfaces include includes metallic pattern distributions within the one or more passive metasurfaces, and tuning of the propagation coefficient for the one or more passive metasurfaces comprises:
tuning a number of the one or more passive metasurfaces placed within the environment, positions of the one or more passive metasurfaces placed within the environment, and the structural design configurations of the one or more passive metasurfaces placed within the environment to maximize the received signal strength at the target point in the environment.
18 . The method of claim 16 , further comprising:
tuning the access point with respect to the one or more passive metasurfaces to maximize the received signal strength at the target point in the environment based on a codebook of transmission signal properties of signals transmitted by the access point.
19 . The method of claim 16 , further comprising:
tuning the access point with respect to the one or more passive metasurfaces to maximize the received signal strength at the target point in the environment based on placement of the access point within the environment.
20 . The method of claim 16 , wherein tuning includes a search based on simulated annealing, Bayesian optimization, a gradient descent algorithm, or a genetic algorithm.
21 . The method of claim 16 , wherein tuning includes maximizing signal-to-noise ratio of signals received at the target point or a capacity of a communication link between the access point and the target point.
22 . The method of claim 16 , wherein metallic pattern distributions within the one or more passive metasurfaces are modeled on an equivalent circuit for each passive metasurface including metallic patterns of the metallic pattern distributions and a substrate with a predefined thickness supporting the metallic patterns.