Dynamic control of unmanned aerial vehicles using reconfigurable intelligent surfaces

Abstract

To provide a method, a system, and a computer readable medium for controlling an unmanned aerial vehicle (UAV) using a reconfigurable intelligent surface (RIS).SOLUTION: A method for establishing direct communication using an unmanned aerial vehicle (UAV) with a reconfigurable intelligent surface (RIS) includes a step of configuring an RIS parameter on the basis of compensating for undesired vibrations in the position and the orientation associated with the UAV. Signal returns associated with the beam signal are steered to the target area by the UAV's RIS on the basis of the RIS parameters. A signal beam is from a transmitter.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] Cross-reference of prior applications Priority is claimed by U.S. Provisional Application No. 63 / 167,131, filed on 29 March 2021, the entire contents of which Provisional Application are incorporated herein by reference.

[0002] The present invention relates to a method, system, and computer-readable medium for controlling an unmanned aerial vehicle (UAV) using a reconfigurable intelligent surface (RIS). [Background technology]

[0003] RIS is a key technology for next-generation wireless networks that can provide complete control over the propagation environment. Specifically, RIS is an artificial passive surface of electromagnetic material that is electronically controlled using integrated electronics technology and possesses unique wireless communication capabilities. One of the main advantages of RIS is its ability to transform the inherently probabilistic nature of the wireless environment into a programmable channel that passively plays an active role in how wireless signals propagate. This advantage enables the construction of advanced applications such as ultra-high-precision localization systems or ultra-capacity focused areas while keeping operational and installation costs low. [Overview of the project] [Means for solving the problem]

[0004] In one embodiment, the present invention provides a method for establishing direct communication using an unmanned aerial vehicle (UAV) having a reconfiguration intelligent surface (RIS). The method includes the steps of: configuring RIS parameters based on compensating for undesirable positional and orientation oscillations associated with the UAV; and steering a signal reflection associated with a signal beam to a target area by the UAV's RIS based on the RIS parameters, wherein the signal beam is from a transmitter.

[0005] Embodiments of the present invention will be described in more detail below with reference to exemplary drawings. The present invention is not limited to exemplary embodiments. All features described and / or illustrated herein may be used alone or in different combinations in embodiments of the present invention. Features and advantages of various embodiments of the present invention will become apparent by reading the following detailed description with reference to the accompanying drawings shown below. [Brief explanation of the drawing]

[0006] [Figure 1] This figure schematically shows a system layout having a transmitter, RIS, and receiving area applicable to embodiments of the present invention. [Figure 2] This diagram schematically shows a UAV equipped with a RIS (Responsive Sound Indicator) that compensates for undesirable vibrations, according to one embodiment of the present invention. [Figure 3] This figure schematically illustrates a method and system for determining a RIS configuration using a RIS configuration processor, according to one embodiment of the present invention. [Figure 4] This figure shows, graphically, the gain obtained in terms of the received signal-to-noise ratio (SNR) and fairness between target locations when compensating for undesirable vibrations on a UAV using one embodiment of the present invention. [Figure 5]This figure shows an exemplary UAV with a lightweight RIS (Remote Access System) to cover initial response teams and victims in an emergency scenario, according to one embodiment of the present invention. [Figure 6] This figure schematically shows another system layout having a transmitter, RIS, and receiving area, applicable to embodiments of the present invention. [Figure 7a] This figure schematically shows the radiation patterns in the RIS along the azimuth and elevation directions, obtained using multiple methods according to embodiments of the present invention. [Figure 7b] This figure schematically shows the radiation patterns in the RIS along the azimuth and elevation directions, obtained using multiple methods according to embodiments of the present invention. [Figure 8a] This figure schematically illustrates the robustness of a drone against motion in an RIS (RiFe: RIS to compensate flight effects) method and an agnostic method for compensating for flight effects, according to embodiments of the present invention. [Figure 8b] This figure schematically illustrates the robustness of the RIS(RiFe) method and the agnostic method for compensating for flight effects against drone motion, according to embodiments of the present invention. [Figure 9a] This figure schematically illustrates the trade-off between performance improvement and the magnitude of oscillation in the RiFe method and the agnostic method according to embodiments of the present invention. [Figure 9b] This figure schematically illustrates the trade-off between performance improvement and the magnitude of oscillation in the RiFe method and the agnostic method according to embodiments of the present invention. [Figure 10] This figure schematically illustrates a bar graph showing the robustness of a drone against motion in the RIS(RiFe) method and an agnostic method for compensating for flight effects according to embodiments of the present invention. [Figure 11a]This figure schematically shows the mean minimum signal-to-noise ratio (SNR) obtained using the RiFe method, the agnostic method, and 3D beam flattening according to embodiments of the present invention. [Figure 11b] This figure schematically shows the mean minimum signal-to-noise ratio (SNR) obtained using the RiFe method, the agnostic method, and 3D beam planarization according to embodiments of the present invention. [Figure 12a] This diagram schematically illustrates the trade-off between improved performance and the magnitude of vibration according to an embodiment of the present invention. [Figure 12b] This diagram schematically illustrates the trade-off between improved performance and the magnitude of vibration according to an embodiment of the present invention. [Figure 13] This figure schematically shows the cumulative signal-to-noise ratio (CDF) of the received signal-to-noise ratio obtained using the RiFe method, the agnostic method, the Fair RiFe method, and the Fair-agnostic method according to embodiments of the present invention. [Figure 14a] This figure schematically illustrates a comparison of the computation times of the RiFe method and the Fair-RiFe method according to embodiments of the present invention. [Figure 14b] This figure schematically illustrates a comparison of the computation times of the RiFe method and the Fair-RiFe method according to embodiments of the present invention. [Modes for carrying out the invention]

[0007] Embodiments of the present invention provide a passive surface on a UAV to counteract undesirable UAV vibrations while pursuing the maximization of overall communication performance, particularly in emergency situations.

[0008] According to one embodiment, the present invention provides a method for dynamically and automatically compensating for unexpected changes in the movement and orientation of a UAV while appropriately configuring the RIS to provide connectivity toward a selected area. For example, the propagation settings of the RIS, such as the angle of reflection or absorption, may be dynamically and conveniently changed in pursuit of a given key performance indicator (KPI). In emergency scenarios such as when a disaster occurs, it may be necessary to extend network capabilities with UAVs that can fly over hard-to-reach locations in a relatively short amount of time. UAVs may support first-response teams by providing ultra-band connectivity. However, UAVs equipped with portable base stations (BS) or active antennas suffer from limited battery life (see Alfattani, S. et al., "Link budget analysis for reconfigurable smart surface in aerial platform", arXiv:2008.12334, pp. 2-7 (August 27, 2020) and Wu, Q. et al., "5G-and-Beyond Networks with UAVs: From Communications to Sensing and Intelligence", arXiv:2010.09317, pp. 1-29 (October 19, 2020). Each of these is incorporated herein by reference).

[0009] In embodiments of the present invention, the RIS is mounted on a UAV as a passive, lightweight device that may extend signal propagation and coverage in a controlled manner. In contrast to other methods of using RIS, embodiments of the present invention provide an efficient control solution that compensates for undesirable vibrations of the UAV (in terms of orientation, tilt angle, and position) while continuing to steer reflected signals toward a specific area (for example, to provide ultraband connectivity to a first-response team).

[0010] According to embodiments of the present invention, the RIS is used to minimize the impact of battery consumption on an emergency UAV equipped with portable networking equipment. Furthermore, embodiments of the present invention provide a robust solution that takes into account unexpected changes in the position and orientation of the UAV while optimally configuring the RIS to cover a selected area.

[0011] Embodiments of the present invention may be particularly advantageously applied to system layouts such as the one shown in Figure 1, where, for example, in an emergency scenario, the receiver may not be directly reachable from the transmitter due to debris or obstacles between the transmitter and receiver. In this context, a UAV may fly and hover over a specific area to provide connectivity to users on the ground in the area by a portable (3G, LTE, or 5G) base station. To address the technical problem that this technique may rapidly deplete the battery power of the UAV depending on the radiated power set at the UAV base station, embodiments of the present invention utilize lightweight, passive elements such as RISs that are installed on the UAV and used to control the reflection of propagated signals to reach specific users on the ground (or a wider area on the ground).

[0012] In embodiments of the present invention, a ground transmitter steers a signal beam toward a RIS flying with a UAV at a given altitude. The RIS is configured to control the reflection angle of the incoming signal in order to steer the signal toward a specific ground user (if the user's location is known) or a wider area (if only the user's spatial statistics are known). Based on the known location of the RIS, the corresponding configuration may be readily derived based on geometric considerations. However, based on the fact that the location of the RIS is affected by uncertainty (e.g., due to weather conditions and / or other environmental factors), the corresponding RIS configuration may lead to a degradation of service if not adequately addressed.

[0013] In one embodiment, the present invention provides a method for establishing direct communication using an unmanned aerial vehicle (UAV) having a reconfigurable intelligent surface (RIS). The method includes the steps of: configuring RIS parameters based on compensating for undesirable positional and orientational vibrations associated with the UAV; and steering a signal reflection associated with a signal beam to a target area by the RIS of the UAV based on the RIS parameters, wherein the signal beam is from a transmitter.

[0014] In one embodiment, the method further includes the step of configuring beamforming for steering a signal beam toward a UAV by a transmitter, wherein the transmitter is a base station.

[0015] In one embodiment, the RIS parameter is an input voltage to one or more RIS elements that determines one or more phase shifts induced by a signal beam.

[0016] In one embodiment, the step of configuring the RIS parameters includes determining the RIS parameters using an optimization algorithm that compensates for undesirable vibrations based on second-order statistics of motion associated with the UAV.

[0017] In one embodiment, the optimization algorithm is based on a mathematical optimization tool and / or an artificial intelligence algorithm.

[0018] In one embodiment, the method further includes the step of using the RIS of a UAV to maintain direct communication between a target area and a transmitter, based on continuously configuring new RIS parameters to compensate for undesirable vibrations of the UAV and using the new RIS parameters to continuously steer signal reflections associated with the signal beam.

[0019] In one embodiment, the method further includes the steps of: obtaining sensor measurements from one or more sensors of a UAV, the sensor measurements indicating one or more coordinates and orientations of the UAV; and steering a signal reflection associated with a signal beam to a target area by the RIS of the UAV based on RIS parameters, the signal beam being from a transmitter.

[0020] In one embodiment, determining UAV statistics related to one or more coordinates and orientations of the UAV is based on constructing a scaled histogram of sensor measurements.

[0021] In one embodiment, the method further includes the steps of obtaining the service area from the transmitter and performing area sampling based on the service area, wherein the step of configuring the RIS parameters is based on performing area sampling.

[0022] In one embodiment, the step of performing area sampling includes generating sample points belonging to an area as equally spaced points based on the fact that information regarding the location of users within the area being serviced is unavailable, and generating sample points belonging to an area according to a probability density function (pdf) based on the fact that a pdf is available regarding the location of users within the area being serviced.

[0023] In one embodiment, the transmitter is a base station, and the method further includes the step of obtaining the coordinates of the base station from the base station, and the step of configuring the RIS parameters is based on the coordinates of the base station.

[0024] In one embodiment, the step of configuring the RIS parameters is based on using closed-form expressions.

[0025] In one embodiment, the step of configuring RIS parameters is based on using semidefinite programming (SDP).

[0026] In another embodiment, the present invention provides a system for establishing direct communication using an unmanned aerial vehicle (UAV) having a reconfigurable intelligent surface (RIS). The system includes a transmitter configured to transmit a signal beam to the UAV, and the UAV having the RIS, the UAV having RIS parameters configured on the basis of compensating for undesirable positional and orientational vibrations related to the UAV, and the RIS of the UAV being configured to steer signal reflections related to the signal beam to a target area based on the RIS parameters.

[0027] In a further embodiment, a tangible non-temporary computer-readable medium having instructions, wherein when the instructions are executed individually or in combination by one or more processors, it provides the execution of a method according to any embodiment of the present invention.

[0028] Figure 1 schematically shows a system layout having a transmitter, RIS, and receiving area applicable to embodiments of the present invention. For example, Figure 1 shows a coordinate plane spanning the x-axis (x), y-axis (y), and z-axis (z). The center of the RIS is a position q = [q] which is affected by uncertainty. x , q y , q z It is located at ]. The target position w is also shown. Furthermore, the variables x', y', and z' represent a relative reference frame with the origin at the lower right corner of the RIS, and the (x', y') plane lies on the surface of the RIS. The variable k indicates the direction of propagation in the RIS of an incoming plane wave originating from a base station (BS). Also, k' is the projection of k onto the (x', y') plane lying on the surface of the RIS. Embodiments of the present invention take into account unexpected vibrations that affect the position and orientation of the RIS, and the azimuth angle φ R and elevation angle θ RThe goal is to optimally configure the RIS parameters so that the signal coming from the transmitter reaches the target area (A). As shown, φ T and θ T are the transpositions of the azimuth angle φ R and the elevation angle θ R respectively. The control of the RIS by the UAV in FIGS. 1 will be described in more detail below.

[0029] In the approach of Wu, Q. et al. and in the approach of Zhang, Q. et al., "Reflections in the Sky: Millimeter wave communication with UAV-carried intelligent reflectors", 2019 IEEE Global Communications Conference (GLOBECOM), pp. 1-6 (December 2019), which is incorporated herein by reference, the usefulness of some network is maximized by intentionally changing the position of the UAV without providing a solution to this technical problem without considering the undesirable vibrations caused by jitter. In particular, Wu, Q. et al. only mention the need to develop a robust optimization algorithm for the undesirable misalignment caused by the jitter of the UAV, without proposing any solution. Zhang, Q. et al. explain that the position of the UAV is intentionally optimized to ensure a line-of-sight link with the user and assume that this position is fully controllable by the network. However, in a real scenario, the (nominal) position of the UAV is affected by the uncertainties caused by possible vibrations in the environment that can have an adverse impact on communication performance if not taken into account. This technical problem is not even recognized by Zhang, Q. et al.

[0030] Figure 2 schematically shows a UAV with a RIS that compensates for undesirable vibrations according to one embodiment of the present invention. As shown, the environment 200 includes a first computing device 202 (e.g., a transmitter or 5G base station) that transmits a signal 204 to a second computing device 206 (e.g., a computing device including a receiver). However, the first computing device 202 may not be able to transmit the signal 204 directly to the second computing device 206 due to an obstacle 208 (e.g., debris and / or other obstacles that may interfere with direct communication between the transmitter and receiver). For example, an emergency scenario may occur requiring network connectivity, but the first and second computing devices 202 and 206 may not be able to communicate with each other due to the obstacle 208. Therefore, a UAV 210 with a RIS may be used to direct the signal 204 to the second computing device 206. Furthermore, although not shown, the second device 206 may also include a transmitter that enables communication back to the first computing device (e.g., a base station). Icon 212 indicates that there may be weather conditions such as rain, wind, and / or other environmental factors that could affect the UAV 210 having RIS. Therefore, as described below, the UAV 210 having RIS may be used to determine an optimal location and position itself so that the UAV 210 can provide connectivity between the first computing device 202 and the second computing device 206 even through environmental factors.

[0031] Referring to Figure 2, in one embodiment, the present invention provides a stationary UAV 210 equipped with a RIS. The size of the RIS is preferably selected based on the altitude of the UAV 210. Particularly in emergency scenarios, but also in other scenarios, the UAV 210 may be equipped with a RIS located at the bottom of the UAV 210 to control signal propagation coming from a ground-placed transmitter while steering the signal toward a selected target area (or ground user), such as area (A) shown in Figure 1. The RIS is passive and lightweight. Control of RIS parameters is performed by an independent control channel which may rely on a proprietary protocol operating at a frequency such as 2.4 gigahertz (GHz). Advantageously, such a control channel requires very little communication capacity and may be performed by existing communication technology.

[0032] When UAV 210 is hovering over a specific stationary position, adverse atmospheric conditions such as wind or rain (indicated by icon 212) may affect the position of UAV 210 by causing unexpected oscillations in the UAV's position or orientation. While such changes are typically mechanically compensated for by the rotor of UAV 210 based on feedback from the UAV 210's sensors, these changes can lead to erroneous instantaneous configurations of RIS parameters, further impairing overall communication between the transmitter (e.g., first computing device 202) and the RIS, and between the RIS and the receiver (e.g., second computing device 206).

[0033] Figure 3 schematically illustrates a method and system for determining a RIS configuration by a RIS configuration processor according to one embodiment of the present invention. In particular, Figure 3 shows a block diagram of a method and system according to one embodiment of the present invention for controlling a UAV 210. It is shown that a network block 304 and a UAV sensor 306 provide information to the RIS configuration processor 302. The RIS configuration processor 302 may be any kind of hardware and / or software logic, such as a central processing unit (CPU), module, controller, and / or logic, that executes computer-executable instructions for performing the functions, processes, and / or methods described herein. The RIS configuration processor 302 includes an optimized feature derivation block 308 and a maximization block 310 for the minimum expected signal-to-noise ratio (SNR) across area samples. The optimized feature derivation block 308 includes a UAV position and vibration statistics extraction block 312 and an area sampling block 314.

[0034] In other words, block 308 uses data received from UAV sensor block 306 (e.g., output traces from the global position sensor (GPS), compass, gyroscope, and accelerometer) as well as data received from network block 304 (e.g., the covered area and the location of the serving bus stop) to derive (e.g., determine) the necessary RIS optimization parameters (optimization features) at the input of the RIS configuration optimization processor 302.

[0035] Block 312 uses the output traces of the GPS, compass, gyroscope, and accelerometer that the UAV is equipped with to derive a statistical representation (e.g., first and second moments) of the RIS's position and orientation in space.

[0036] Block 314 samples the area of ​​interest. Depending on the available information about the area to be covered, different techniques may be used, such as uniform sampling or Gaussian sampling based on the known distribution of initial responders within the area.

[0037] Block 310 optimizes the RIS configuration so that the minimum SNR across the entire area sample is maximized. The optimization takes into account the geometric arrangement of the problem (locations of the BS, UAV, and area sample) and statistics on the UAV's position and orientation.

[0038] During operation, on the network side (e.g., network block 304), the area to be served is defined and the location of at least one serving base station (BS) (e.g., a first computing device 202 such as a receiver) is given. The serving BS is selected to provide connectivity to the area by reflection on the RIS mounted on a UAV 210, for example, a drone. In other words, the UAV 210 may receive information from the network (e.g., the first computing device 202 which is a BS) indicating the area to be served (e.g., (A) from Figure 2) and the location of one or more BSs (e.g., coordinates / location of the first computing device 202). The RIS configuration processor 302 may receive this information and / or additional information from the network block 304.

[0039] The area to be covered is approximated by sampling of a set of points. Different sampling methods may be used for this purpose. For example, the area may be sampled according to a regular sampling grid, or techniques such as Monte Carlo sampling may be used if the probability density function of the location of the serviced user within the area is known. Area sampling techniques allow the RIS configuration processor 304 to take into account the uncertainty of the user's location. In other words, the RIS configuration processor 302 receives the serviced area from the network block 304. Using the serviced area, the RIS configuration processor 302 uses the area sampling block 314 to determine an area sample, and the area sample is provided to the maximization block 310 for the area sample as a whole with the minimum expected SNR.

[0040] In other words, the area sampling block 314 may take the coordinates of a target area as input in terms of the limits of the x, y, and z coordinates being served (e.g., x_1 < x < x_2 and similarly for y and z), and output T sample points (e.g., T triplets (x,y,z) of points belonging to the area). The sample points may be generated in several different ways. For example, based on the fact that information about the statistics of the user's position is not available, the points may be generated at equal intervals within the area (e.g., according to a regular grid). Based on the fact that the probability density function (pdf) of the user's position is known, the coordinates of the T points are generated using the likelihood defined by the pdf of that distribution. In operation, this may be done by subtracting a uniform number between 0 and 1 and giving that uniform number to an inverse cumulative distribution function which may be obtained from the pdf.

[0041] Furthermore, for flight control and stabilization, the UAV 210 is natively equipped with sensors 306 (e.g., gyroscope, compass, accelerometer, GPS sensor, etc.). For example, sensors 306 may determine coordinate and orientation measurements of the drone (UVA 210) and provide those measurements to the optimization feature derivation block 308. Embodiments of the present invention utilize the information from those sensors 306 (e.g., coordinate and orientation measurements) to determine (e.g., estimate) the position and vibration distribution of the RIS and obtain a statistical representation of the uncertainty of the position and orientation of the reflective surface that allows the drone to be equipped. This statistical representation allows embodiments of the present invention to take into account the effects of undesirable surface movement (e.g., movement caused by unpredictable weather such as wind and rain). In other words, the RIS configuration processor 302 obtains sensor measurements (e.g., coordinate and / or orientation measurements) from the UAV sensors 306. Based on the sensor measurements, the RIS configuration processor 302 uses block 312 to determine statistical representations such as the coordinate and orientation statistics of the UAV and provides those statistical representations to block 310.

[0042] In other words, the UAV sensor in block 306 outputs measurements of the drone's coordinates and orientation over time. Once a sufficient number of measurements are available, block 312 may extract coordinate and orientation statistics. This can be done in several ways, such as by constructing a scaled histogram of the measurements. Once such statistics are estimated, they indicate the degree of variation in the UAV's position and orientation relative to the nominal position of the UAV, which may be assumed to be known. Thus, the variation is due to undesirable movements caused by swaying.

[0043] Statistical representations of the drone's positioning and orientation are provided to an RIS configuration optimization block (e.g., a maximization block 310 across area samples) along with the position of the serving BS and a sample of the input area, where the RIS configuration is optimized with the goal of maximizing the expected minimum SNR across the sample of the area being served. Maximizing the minimum expected value takes into account statistics for both positional and orientational fluctuations, thus obtaining an RIS configuration that is robust to positional and orientational fluctuations of the metasurface. Block 310 may optimize the RIS configuration using mathematical optimization tools and / or suitably designed artificial intelligence techniques. The position of the BS may be used to derive a suitable RIS configuration. The RIS may, in some way, behave as an adjustable mirror, and therefore, in order to properly configure the RIS to reflect signals coming from the BS in a desired direction, it may be necessary to know the angle of arrival of the incoming signal, which may be derived based on knowing the position of the BS. In some cases, the RIS configuration processor 302 may use statistical representations of the UAV's coordinates and orientation, area samples, and / or the coordinates of the BS to optimize the RIS configuration. For example, the phase shift and reflection coefficient of each RIS element may be optimized. These parameters allow control over the reflection of the signal. Thus, the RIS configuration is a set of phase shift and reflection coefficients applied to the RIS elements. Block 310 is described in more detail below (for example, algorithms 1 and 2 described below).

[0044] In some cases, the RIS configuration processor 302 may be implemented by a transmitter (for example, transmitter 202). In other cases, the RIS configuration processor 302 may be implemented by a UAV 210.

[0045] Embodiments of the present invention improve overall communication performance by bringing undesirable vibrations under control and providing an optimal RIS configuration. Figure 4 graphs the gain obtained in terms of received signal-to-noise ratio (SNR) and fairness between target locations when compensating for undesirable vibrations on a UAV with one embodiment of the present invention. In particular, the obtained gains are experimentally shown in Figure 4, where embodiments of the present invention (labeled "robust") are compared with an agnostic solution (labeled "agnostic") in which vibrations are not taken into account. For evaluation, the height of the UAV was fixed at 20m and 1000m 2 A circular target area of ​​size was used. For simplicity of presentation, it was assumed that the dispersion of the RIS surface position fluctuations along the x, y, and z axes was the same and denoted as σ. On the left, the cumulative distribution function (cdf) of the received SNR is shown in dB at the target position with respect to different values ​​of σ expressed in frequency. It can be seen that embodiments of the present invention result in a significant increase in the worst received SNR. In fact, the obtained cdf is shifted toward higher values ​​of SNR compared to the agnostic solution and shows a much steeper slope, indicating that embodiments of the present invention obtain higher fairness between target positions. Thus, the agnostic solution may obtain high SNR at a few locations but still result in low SNR at many other locations. Such behavior is shown on the right side of Figure 4, which shows Jain's fairness index for the two methods with respect to the dispersion of the RIS surface position fluctuations σ. One reason that embodiments of the present invention provide these improved results is that, when the UAV is subjected to motion, embodiments of the present invention account for the extra shift in the beamforming design that reflects as much of the power radiated at the beamstation (BS) toward the target area as possible. In contrast, agnostic solutions reflect the beam in a fixed direction at the beam rigging (RIS) regardless of motion, and as a result, the power radiated at the BS is often directed away from the target area and ultimately wasted.

[0046] Embodiments of the present invention provide the following improvements. 1. UAV 210 equipped with passive equipment for proactively and dynamically and optimally controlling the reflection angle of incoming signals toward a desired target area. 2. Passive surfaces, particularly RIS, are optimally and continuously configured to address undesirable UAV vibrations of both positional and oriental motion by taking into account collected statistics of both positional and oriental motion.

[0047] In one embodiment, the present invention provides a method for establishing direct communication between a ground-based transmitter, an aerial object (particularly a UAV with a configurable passive surface in the form of a RIS), and a target area, the method comprising: 1. Configure beamforming in the transmitter (e.g., the first computing device 202) to steer a signal beam toward a UAV in flight (e.g., UAV 210). For example, the transmitter may be configured to steer the signal toward the center of the RIS. Since the coordinates of the latter are known with some uncertainty, the transmitter may steer the signal in a manner robust to such uncertainty. This is in contrast to existing methods that assume the position of the UAV (and by extension, the position of the center of the RIS) is fully known. Therefore, if not considered, uncertainty in the position of the UAV can result in misalignment of the signal being steered in the transmitter and thus a loss in terms of overall performance. 2. By taking into account undesirable oscillations of the UAV's position and orientation, the RIS parameters, including the input voltages to the RIS elements that determine the phase shift induced by the incoming signal, are appropriately set. For example, the RIS is generally constructed as an array of passive antennas (e.g., PIN diodes) that apply a controllable phase shift to the incoming signal. The desired phase shift is obtained by suitably controlling the input voltage to each RIS element (e.g., passive antenna) (which generally requires a very small amount of power). Thus, optimizing the RIS configuration suggests suitably selecting the phase shift to be applied, and in turn, the applied phase shift can be linked to the input voltage required at each RIS element. The RIS parameters may refer to a set of selected phase shifts and associated input voltages. Unlike existing methods that assume that the UAV's position and orientation can be fully controlled by the network, the present invention uses one or more algorithms for configuring RIS parameters that compensate for undesirable oscillations in the UAV. The present invention may design robust RIS parameters based on quadratic statistics of oscillation. The optimization algorithm may be based on mathematical optimization tools and / or suitably designed artificial intelligence techniques. 3. The signal reflections from the RIS mounted on the UAV are steered toward the target area, and this connection is kept stable. In contrast, in existing methods, when oscillations occur, the connection may be interrupted because these oscillations are taken into account in the present invention, whereas they are not taken into account in existing solutions.

[0048] A significant improvement offered by installing the RIS on the UAV 210 is the potential for the RIS to steer signals toward hard-to-reach locations without the battery consumption burden required by an active antenna. Compared to typical land-based solutions where the RIS is installed in front of a building, such a solution leverages the availability of in-line-of-sight links with a high probability. This characteristic has been proven beneficial to the average received SNR. Wu, Q. et al. examine several advantages of RIS-assisted UAV communications, but also demonstrate a way to maximize some network usefulness by intentionally changing the UAV's position without considering undesirable vibrations caused by turbulence. In this case, overall communication performance can be negatively affected when fluctuations (minor fluctuations) occur in the orientation and / or position of the RIS (e.g., due to weather). Indeed, while UAVs have sensor systems that stabilize the drone while hovering and keep the drone's position fixed, when faced with bad weather conditions, the automatically performed drone movement counteraction may momentarily result in an incorrect RIS configuration. In Alfattani, S. et al., “Aerial platforms with reconfigurable smart surfaces for 5G and beyond,” IEEE Communications Magazine, 59(1), pp. 96-102 (February 2021), incorporated herein by reference, the configuration of the RIS on a UAV is optimized to maximize the capacity for a set of users whose locations are known. In this case, too, it is assumed that the UAV with the RIS remains stationary throughout the data transmission time. Similarly, Zhang, Q. et al. describe that augmentation techniques are used to find the correct (nominal) location of a UAV in order to maximize downlink capacity. In such a scenario, when swaying occurs, if the swaying is not taken into account in the optimization procedure, it will result in an incorrect configuration of the RIS.

[0049] Therefore, existing methods, for the sake of simplification, assume that the UAV flies in perfect stability when optimizing the RIS configuration. In contrast, embodiments of the present invention can optimize in real-world scenarios where the UAV is subjected to motion (e.g., due to weather) and provide a system that takes undesirable motion into account while continuously configuring the onboard RIS. This procedure aims to maximize overall performance in terms of fairness within a target area and the minimum experienced SNR. Furthermore, embodiments of the present invention will first provide an on-UAV RIS for emergency scenarios. According to embodiments of the present invention, the RIS can be optimally deployed on the UAV even when encountering adverse weather conditions.

[0050] Embodiments of the present invention, including those described above, are described in more detail below in Figures 1 to 4. In particular, next-generation mobile networks need to expand into uncharted territory to enable the digital transformation of society. In this context, aerial devices such as unmanned aerial vehicles (UAVs) are expected to address this gap in hard-to-reach places. However, limited battery life is an obstacle to the smooth adoption of such solutions. Reconfigurable intelligent surfaces (RISs) represent a promising solution to address this challenge, as onboard passive, lightweight, and controllable devices can efficiently reflect signal propagation from ground-based BSs toward specific target areas.

[0051] Embodiments of the present invention focus on air-to-ground networks over which UAVs equipped with RIS may fly over selected areas to provide connectivity. In particular, the present invention describes a method of using RIS (RiFe) to optimally compensate for flight effects and for compensating for flight effects, and Fair-RiFe, a practical implementation of RiFe that automatically configures RIS parameters to account for undesirable UAV vibrations caused by poor atmospheric conditions. The results show that both algorithms / techniques provide robustness and reliability while outperforming state-of-the-art solutions under several conditions investigated.

[0052] Unmanned aerial vehicles (UAVs), also known as drones in this specification, are increasingly becoming a part of people's lives by improving the jobs people do, such as delivering packages, how they have fun, and how they enhance the safety and security of society. Lithium-ion (Li-ion) batteries have already reached maturity, enabling devices of limited size to fly to hard-to-reach places in very short amounts of time. In recent years, the telecommunications industry and academia have made considerable efforts to bring flexibility and agility to advanced wireless networks that rely on flying access points, i.e., air-to-ground networks. UAVs have demonstrated their ability to easily establish line-of-sight (LoS) links to ground users, thereby proving suitable for avoiding obstacles that would impair overall communication quality. Dense urban scenarios further exacerbate the obstacle problem, making UAVs a practical solution for building reliable networks on demand.

[0053] Due to their rapid deployment characteristics, UAVs are recognized as a primary technology for addressing emergency situations, such as accelerating rescue operations and assisting first-response teams in areas where connectivity may be unavailable or when network infrastructure is temporarily damaged or unavailable. For example, UAVs may be used to bring backup connectivity into such areas and / or to locate missing persons by leveraging advanced sensing and localization technologies that utilize cellular protocol stacks. However, since these UAVs are envisioned as flying mobile base stations (BS) with one or more active antennas, a significant increase in total power consumption and, consequently, battery drain issues are anticipated due to i) the weight of the active elements, ii) the power radiated to reach ground targets, and iii) the additional power burden required to establish backhaul links and process incoming packets.

[0054] To overcome the aforementioned problems, lightweight, low-energy devices are sometimes required within an aircraft. In this context, reconfigurable intelligent surfaces (RISs) are currently attracting considerable attention thanks to their ability to control the propagation environment by altering the reflection, absorption, and amplitude properties of the materials to which signals bounce. This introduces a new and powerful tool that enables the efficient alteration of the direction of signal propagation with very limited power expenditure; for example, a varactor diode may apply a phase shift and / or absorb an incoming signal in a real-time, reconfigurable manner. Passive, flexible, and configurable elements are suited to a wide range of anticipated applications, spanning from communications with increased electromagnetic field exposure efficiency (EMFEE) to ultra-high-precision localization mechanisms and user-centric, quality-of-service connectivity.

[0055] Figure 5 shows an exemplary UAV with a lightweight RIS for covering first responder teams and victims in an emergency scenario, according to one embodiment of the present invention. As shown, a system based on a standard UAV may be combined with a moderately sized RIS to effectively reflect incoming signals to hard-to-reach locations while keeping power consumption very low. Modeling a proper RIS assumes that the positions of the transmitter, receiver, and RIS are known, but the high mobility characteristics of the UAV introduce unprecedented random variables into such a complex analysis. In addition, when small fluctuations occur in the orientation and / or position of the RIS, the overall communication performance may be adversely affected. A sophisticated sensor system and / or GPS antenna may be built to stabilize the drone and keep its position fixed while it hovers over a selected area, meaning that actions are automatically taken to counteract the drone's movement when it encounters adverse weather conditions, but even so, orientational oscillations or positional fluctuations may momentarily result in an incorrect RIS configuration.

[0056] The present invention focuses on the above-described scenario in which a UAV has a passive RIS for supporting an initial response team within a selected target area. Embodiments of the present invention describe using RiFe to address the problem of undesirable UAV vibration while steering signal reflections in order to build a robust and reliable air-to-ground network solution.

[0057] Embodiments of the present invention describe the design of an optimization framework to address undesirable UAV turbulence—caused by weather—which worsens when higher altitudes are considered. This is in stark contrast to conventional studies, which typically assume perfectly stable flight. Since UAV flight turbulence is always present and unavoidable in real-world situations, the present invention (e.g., RiFe and / or Fair-RiFe) aims to maximize the worst-case SNR averaged across undesirable turbulence on the UAV in a target area where a large number of receivers are present. RiFe may be based on quadratic statistics of turbulence and receiver position. Thus, the advantage of such a method is the reduction of overhead required to acquire instantaneous channel information.

[0058] In addition to the optimization framework, RiFe has been extended to account for practical considerations such as the need to update RIS parameters due to rapidly changing channel statistics, UAV mobility, and complexity issues, and may be renamed Fair-RiFe. Numerical results show, as expected, that the present invention outperforms modern solutions that assume perfectly stable flight. In fact, the present invention results in a considerably higher received SNR at the target location thanks to more robust passive beamforming in the RIS. Therefore, among the improvements, four improvements of the present invention are summarized below. C1: A novel mathematical framework for maximizing the worst-case SNR within a given target area by taking into account undesirable rotations of the RIS surface position caused by flight turbulence. RiFe is a relatively simple method that addresses the aforementioned problem by designing both pre-coding vectors and appropriate RIS configurations in C2:BS. Fair-RiFe is an easy-to-develop solution that takes into account practical considerations such as time and / or complexity constraints for C3:RIS optimization routines and reconfiguration. C4: Comprehensive numerical results showing a significant improvement in average received SNR at the target location compared to a standard solution.

[0059] Below, we describe this system, formulate the problem for proposing RiFe to address the optimization of key physical quantities of interest and metrics of consideration, take into account practical considerations such as timing constraints and UAV mobility, define the proposed algorithm, method, and numerical results for evaluating the performance of the system, highlight the differences compared to conventional methods and the present invention when integrating RIS into UAV-based communications, and define the conclusions.

[0060] As used below, italics are used to represent scalars, while vectors and matrices are represented by bold lowercase and uppercase letters, respectively.

number

number

number

number

[0061] Figure 6 schematically shows another system layout having a transmitter, RIS, and receiving area applicable to embodiments of the present invention. Figure 6 is similar to the system shown in Figure 1, but includes additional and / or alternative features / elements. For example, area (A) shown in Figure 6 is circular, while area (A) shown in Figure 1 is rectangular. Figure 6 further includes coordinates for the pitch, roll, and yaw of the UAV. Referring to Figure 6, a geometric representation is shown including the transmitter (lower left), RIS (top), and target area A (lower right). A base station (BS) or general transmitter (e.g., first computing device 202) located at the origin and having M antennas whose signals may cover target area A where initial responders, victims, and / or other persons are located. Furthermore, a UAV-mounted RIS is considered, consisting of N reflective elements that reflect signals coming from the BS toward the target area. In particular, each of the above-mentioned receivers may be reached by a signal that experiences a minimum SNR for successfully decoding subsequent packets. In some cases, the location of the initial response team may be known with some uncertainty, while only the probability distribution function (pdf) of the spatial location of the victims may be known.

[0062] Considering that the UAV is hovering, it may be subjected to motion caused by wind or other weather conditions, which can cause undesirable roll, yaw, and pitch on the RIS surface. Therefore, q = [q x , q y , q z ] T φ represents the position of the center of RIS. R and θ R However, these represent the azimuth and elevation angles of the geographical path connecting BS to RIS, respectively. Therefore, the coordinates of q are, respectively, q x = ||q||cos(θ R )cos(φ R ), q y = ||q||cos(θ R )sin(φ R ), and q z = ||q||sin(θ R ) may also be expressed as . Furthermore, r = [ψ x ,ψ y ,ψ z ] T This takes into account the possible fluctuations in the orientation of the RIS surface, where the random variable (RV)ψ x ψ y , and ψ z These represent rotations along the x, y, and z axes, respectively. The term "r" is a three-element vector that describes the orientation fluctuations of the RIS. r represents undesirable rotations along the x axis (pitch), y axis (roll), and z axis (yaw).

[0063] In some cases, such rotations are independent of each other, with zero mean and variance, respectively.

number

[0064] The distance d1 between the source node and the RIS is defined as follows:

number

number

[0065] Given the positions of the BS and RIS relative to the target area A, the communication links between the BS and RIS, and between the RIS and the target point in A, may be determined to be in line-of-sight (LoS) with a very high probability. Therefore, the channel power gain β1 from the source to the RIS is:

number

number

number

number

number

[0066] In some modified forms, the drone receives an incoming signal.

number

number

number

number

number

[0067] Similarly, the corresponding values ​​of the phase shift on the i-th element of the RIS and the antenna array response when the signal is reflected toward the target point w are, respectively, the following:

number

[0068] Random phase shift

number

[0069] The channel between RS and RIS may be defined as follows:

number

number

[0070] From equations (17) and (18), the received signal at the intended position ω may also be described as follows: y(q, w, r,Θ, v) = h H (q, w, r)ΘG(q, r)vs + n (19)

[0071] Here,

number

number

number

number

number

number

[0072] RIS may be used to compensate for flight effects. For example, the present invention may be used to ensure coverage over a given target area A by taking into account undesirable rotation of the RIS surface caused by motion on the UAV, and to ensure, for example, that each receiver obtains a sufficiently high SNR. In this regard, below, an effective optimization problem may be formulated in which the received SNR is defined at a given target position w, and taking into account possible motion, seeks to maximize the worst SNR among all possible target positions. Accordingly, the present invention provides a solution to such a problem, called RIS (RiFe) for compensating for flight effects, based on semidefinite relaxation (SDR) and Monte Carlo sampling.

[0073] The problem is formulated as follows. For example, the received SNR at position w may be given by the following:

number

[0074] The present invention may also be used to maximize the worst-case SNR in a target area by appropriately selecting both the beamforming vector and RIS configuration in the beamforming system (BS), while taking into account the oscillations specified by r, for example,

number

[0075] The standard solution method (RiFe) is described below. The numerator of the objective function in Problem 1 can be simplified as follows.

number

number

number

number

number

number

number

number

[0076] matrix

number

number

number

number

[0077] Problem 1 may also be reformulated as follows:

number

[0078] For example, equation (37) shows the equation for the expected SNR at position w, where the expected value is obtained over the entire unknown undesirable fluctuation of the UAV's position and orientation. The matrix variable Θ is simply replaced by the vector variable θ, which is the diagonal of the latter, for this simplifies the notation. Note that the optimal value of the precoder v adopted at BS does not depend on w and may be optimized by aligning it with the direction of the outgoing link between BS and RIS, while keeping the vector θ of the RIS parameters constant. Therefore, ||a BS (q)|| 2 Noting that = M, and using maximum ratio transmission (MRT), the equation may also be written as follows:

number

[0079] Therefore, this leaves us with the following optimization problem regarding the RIS parameters.

number

[0080] In Problem 2, the beamforming vector was fixed at the transmitter, as shown in equation (38). Therefore, the remaining optimization problem is a function of the RIS configuration only (e.g., vector θ). Problem 2 is also non-convex due to the maximization of a quadratic function at θ. An efficient solution may be found by using SDR, as detailed below.

number

number

[0081] Compared to Problem 2, the above is a matrix variable

number

[0082] The general probability density function f of any given receiver distribution w To address (w), the Monte Carlo sampling method may be applied, thereby increasing the number of sample points N w However, within target area A, prior statistics f w It is reduced according to (w). N w If N is sufficiently large, correct sampling of PDFs from receivers within the target area will be obtained. w It should be noted that this can be expressed as a trade-off between the complexity and accuracy of the proposed method.

number

number

[0083] In Problem 3, the non-convex rank constraint that appeared in equation (41) was relaxed. Thus, the resulting problem is convex and can potentially be solved by a standard SDP.

[0084] The proposed method is shown in Algorithm 1. In particular, Algorithm 1 shows RiFe. In step 1, the processor (for example, RIS configuration processor 302) is N w It may be initialized. In step 2, the processor f w (w) N w individual points

number

number

number

number

[0085] The complexity of RiFe is basically determined by problem 3, which is solved by a dichotomy that generally requires 10-12 iterations, and each iteration is...

number

[0086] Below, another embodiment of the invention that takes practical considerations into account is described. In particular, a more practical solution aimed at simplifying the optimization procedure described in Algorithm 1 is described. In particular, Problem 3 needs to be solved each time the statistics of r are no longer valid, which can be too time-consuming when the statistics change rapidly, given the need to use SDP. Therefore, below, a suboptimal and simple closed-form solution to Problem 3 is described.

[0087] Total number of sampling points N w Given, the objective is to set

number

number

number

number

[0088] weight λ i is a set

number

number

number

number

number

number

number

number

number

number

number

Number

Number

Number

[0089] The proposed method is shown in Algorithm 2 and is called Fair-Rife. In step 1, the processor (for example, the RIS configuration processor 302) may initialize N w In step 2, the processor may reduce the set w by N w points

Number

Number

Number

Number

Number

[0090] Figure 7 schematically shows the radiation patterns in the RIS along the azimuth and elevation directions obtained using multiple methods according to embodiments of the present invention. In particular, Figure 7 shows σ = σ ψ,x = σ ψ,y = σ ψ,z For different values ​​of UAV orientation sway, the radiation patterns in the RIS along the azimuth and elevation directions obtained by RiFe, Fair-RiFe, and the agnostic method are shown below, using the default values ​​in Table I. In Figure 7, the radiation patterns in the RIS along the azimuth and elevation directions obtained by RiFe, Fair-RiFe, and the agnostic solution that does not take UAV sway into account are compared. As expected, the agnostic method points to the center of the target area regardless of the drone sway entity, while both RiFe and Fair-RiFe tend to diffuse energy over a wider angular range as the sway increases.

[0091] A special case of deterministic UAV motion is described below. In particular, a special case of the above scenario where the UAV is moving in a given direction and velocity is considered. In this case, the statistics of r change rapidly with the drone's movement, and this is the updated matrix.

number

Number

Number

Number

Number

[0092] Therefore, both Algorithm 1 and Algorithm 2 described above may be alternatively applied without any further modification.

[0093] Hereinafter, the performance evaluation of the use of the embodiments of the present invention will be described. In particular, numerical results for evaluating the performance of the above embodiments by an ad hoc MATLAB (registered trademark) simulation will be described below. One BS having M = 16 antenna elements is placed at coordinates (0, 0, 10) m, while the RIS is installed on a drone located at coordinates (25, 25, h d ) m, where the altitude of the drone is set to h d = 50 m unless otherwise specified. The RIS has N x ×N y antenna elements, and N x ×N y = 10 unless otherwise specified. Therefore, the total number of RIS elements is N = N x N yand the mutual distance is d = λ / 2. In the simulation, a mutual distance of λ / 2 that enables ignoring the mutual coupling effect between RIS elements is considered. The target area where the first responder team and / or the victims are located is centered at the coordinates (50, 20, 0) m. The transmission power of the BS is set to P = 24 decibel-milliwatts (dBm), and the average noise power is

Number

Number

[0094] Table I (Table 1), which shows the default simulation settings, is presented below.

Table 1

[0095] The robustness against jitter is described below. First, the position of the first responder team within the area is fully known, and the case where the drone is affected by the angular jitter is considered. The robustness of the present invention against the uncertainty regarding the drone's orientation is emphasized. RiFe is compared with an agnostic procedure that optimizes the RIS parameters by ignoring such jitter, denoted as agnostic, and assuming perfectly stable flight. Ten users uniformly dropped in a circular target area with radius R = {20, 35}m are considered.

[0096] FIG. 8 schematically shows the robustness of the RIS (RiFe) method and the agnostic method for compensating the flight effect against the jitter of the drone according to an embodiment of the present invention. In particular, FIG. 8 shows the radius r of the target area and the altitude h d of the drone for different values of the average minimum SNR obtained by the RiFe and the agnostic solution. FIG. 8 compares the performance of RiFe with the agnostic method from the perspective of the minimum SNR received by the first responder team within the target area. It is worth noting that the vibration of the drone causes a decrease in the received SNR, which is mainly caused by the misalignment of the reflected beam (on the RIS) for the covered users within the target area. In particular RiFe mitigates the impact of such misalignment by optimizing the RIS parameters according to the statistics of such vibrations. In fact, the present invention tends to generate a wider beam compared to the agnostic solution, for example, spreading the radiated power better over the target area, and thus compensating for the undesirable misalignment.

[0097] In some cases, the performance improvement of RiFe becomes less pronounced as the size of the target area increases. When the points to be covered are sufficiently separated, the reflected beam on the RIS focuses on each target point. Conversely, when the receivers are close together, RiFe tends to cover all target points with a single broad beam. This results in a similar RIS configuration compared to agnostic solutions and therefore less overall performance improvement in terms of received SNR.

[0098] Furthermore, a trade-off is observed between the effectiveness of RiFe's vibration compensation mechanism and the magnitude of such vibrations. In fact, at relatively low drone altitudes or σ ψ When the magnitude of the tremor is small due to a small value, and when the drone is at a high altitude or σ ψ In the two extreme cases, when the magnitude of oscillation is large due to a large value of σ, the two considered methods produce similar improvements. The fundamental reason for this behavior is that when oscillation is limited, its corresponding effect becomes negligible, and therefore the improvement of the proposed compensation mechanism becomes unreasonable. Conversely, at high altitudes, small oscillations can result in a large deviation of the corresponding beam relative to the desired pointing direction. In this case, RiFe generates a broad beam to diffuse the radiated power over a larger angular range, thereby causing an unavoidable decrease in received signal power. Such behavior is highlighted in Figure 9, which shows the cumulative density function (CDF) of the received SNR. Figure 9 schematically illustrates the trade-off between performance improvement and the magnitude of oscillation according to embodiments of the present invention using the RiFe method and the agnostic method. As shown, the oscillation of the UAV orientation σ and the drone altitude h d The CDF of the received SNR in a target area of ​​radius r = 35m, obtained by RiFe and agnostic solutions for different values ​​of r. In other words, Figure 9 shows the cumulative density function (CDF) of the received SNR.

[0099] Additional tests were performed to account for the effect of RIS size by varying the number of elements N. Figure 10 schematically shows bar graphs illustrating the robustness of the RIS(RiFe) method and the agnostic method for compensating for flight effects against drone motion according to embodiments of the present invention. As shown, when the altitude of the UAV is h d The average minimum SNR obtained by RiFe and the agnostic solution for different settings of the number of RIS elements N, with the distance set to 50m, the target area having a radius r = 35m, and the UAV oscillation σ = 5°. In other words, Figure 10 compares the performance of RiFe with the agnostic scheme / method in terms of minimum SNR. The overall upward trend of the minimum SNR achieved with respect to the number of elements is shown. This is due to the ability to focus, which affects the directivity of the reflected beam—proportional to its magnitude. However, the more the energy is concentrated at the target point, the greater the impact of misalignment caused by the oscillation of the drone's orientation. This, in turn, gives rise to the difference between the performance of RiFe and the performance of the agnostic optimization scheme / method, and this difference increases with the number of elements available, thus highlighting the importance of a robust RIS configuration to counteract drone oscillation.

[0100] Furthermore, the multicast rates obtained from the two different RIS optimization methods considered are compared. Multicast rate is a key performance indicator in emergency scenarios, given the need to propagate useful information to the entire first-responder team as quickly as possible. To highlight the benefits of a robust optimization strategy for improving overall system performance and effectiveness, the obtained comparison results are shown in Table II. In particular, Table II shows multicast rate performance in bits per second (BPS) / Hertz (Hz), taking into account fluctuations in different target area radii and UAV orientations. Bold values ​​and non-bold values ​​in parentheses represent the performance of RiFe and the performance of the agnostic optimization method / method, respectively. [Table 2]

[0101] An analysis of location-independent solutions is described below. When the goal of BS is to provide services to a large number of victims whose locations are unknown, the objective of the present invention may be to provide the minimum SNR within the target area to ensure adequate signal coverage. In some examples, it is assumed that the distribution of victims within the target area is known, and that the associated standard deviations along the x and y axes centered at a given point C are σ w,x = σ w,y = σ w It follows a symmetric, independent, bivariate normal distribution equal to σ. w The value of is set to {2.5, 10}m, which represents two possible scenarios: victims located at predefined collection points, and a more sparse distribution around the affected area. w The value is 80m 2 and 1200m 2 Each of these areas corresponds to a specific region, and 95% of the victims are located within that region. Furthermore, this invention is compared to 3D beam planarization.

[0102] Figure 11 schematically shows the mean minimum SNR obtained using the RiFe method, the agnostic method, and 3D beam planarization according to embodiments of the present invention. As shown, the drone altitude h d and user base σ wAgnostic solutions and 3D beam planarization for 95% of users receiving the service, for different values ​​of . In other words, Figure 11 shows the minimum SNR achieved for victims who fell into the area, for different values ​​of drone motion, drone altitude, and victim spread within the target area. After comparing the curves for low and high victim spread within the area, you will notice that the minimum received SNR tends to decrease as drone motion increases. On the other hand, when victims are more spread out, performance tends to improve under low perturbation regimes. This behavior is due to the limited number of samples used for Monte Carlo sampling of the victim distribution. In fact, when the victim location spread is small, the method of the present invention accurately represents the underlying distribution, which in turn results in effective coverage of the area. Conversely, when the victim location spread is large, the limited number of samples leads to sparser sampling of the distribution. As a result, the RIS configuration tends to concentrate reflected energy toward the available samples rather than following the actual input distribution.

[0103] It should also be noted that when the affected area is widely dispersed, the beam pattern generated by RiFe is sufficiently diffused to effectively cover the target area, resulting in a significant improvement in SNR. On the other hand, when the affected area is more concentrated, all the algorithms considered generate beams that adequately cover the target area. However, an overall improvement in SNR obtained by RiFe is observed. Figure 12 illustrates this in more detail. Figure 12 schematically shows the trade-off between performance improvement and the magnitude of sway according to an embodiment of the present invention. As shown, the spread of the affected area σ w For different values ​​of altitude h d= CDF of received SNR acquired by RiFe and agnostic solution using a 50m drone. In other words, in Figure 12, CDF experienced by victims in the target area, considering different scenarios. As shown, the CDF acquired by RiFe has a more pronounced slope compared to the CDF acquired by the agnostic solution, thanks to a larger beam pattern.

[0104] A practical evaluation is described below. The aforementioned numerical evaluation demonstrates the effectiveness of RiFe in obtaining a robust RIS configuration that can mitigate the effects of the inherent instability of a drone in flight. Finally, the performance of the RiFe algorithm is compared to the Fair RiFe solution described above in terms of the quality of coverage of the target area and the computational cost of the optimization process. Furthermore, an agnostic version of Fair RiFe in the numerical evaluation, i.e., Fair-agnostic, is also used, in which algorithm 2 is solved while ignoring the statistics of drone motion. The same scenario described above is considered.

[0105] Figure 13 schematically shows the cumulative density function (CDF) of the received SNR obtained using the RiFe method, the agnostic method, the Fair RiFe method, and the Fair-agnostic method according to embodiments of the present invention. d = Set to 50m, the extent of the affected area is σ wThe CDFs of the received SNR obtained by the RiFe, agnostic solution, Fair RiFe, and Fair-agnostic methods are shown, with the UAV oscillation σ = 5°, equal to 2.5m. In other words, Figure 13 compares the CDFs of the SNR experienced over the target area, considering the different proposed solutions. From the results, it can be noticed that the CDFs obtained by the Fair RiFe and Fair agnostic algorithms exhibit a gentler slope than their optimal counterparts. Thus, their CDFs result in a more non-uniform distribution of radiated power over the area compared to their optimal counterparts, due to the suboptimal RIS configuration. Comparing the performance of Fair RiFe and Fair agnostic, it can be noticed that for relatively low SNR values, the distributions are very similar to each other and therefore yield similar performance in terms of the minimum SNR provided. Conversely, given relatively high SNR values, it can be noticed that the Fair RiFe solution achieves a higher overall SNR. Therefore, it still leads to performance improvements by taking into account the statistics of drone motion in the optimization process.

[0106] Furthermore, the computation time shown in Figure 14 is evaluated. Figure 14 schematically shows a comparison of the computation times of the RiFe method and the Fair-RiFe method according to embodiments of the present invention. In particular, the number of points N in the target area w The total optimization time required by the RiFe and Fair RiFe methods is compared while varying the size N of the onboard RIS. In Figure 14a, it can be seen that the optimization time is dramatically reduced when considering the Fair RiFe method compared to the RiFe method. The reason behind such improvement is that Fair RiFe does not involve an optimization process to obtain a solution. Furthermore, the optimization time of Fair RiFe compared to the RiFe solution is reduced for the number of points N. wIt can be seen that it shows limited growth. Therefore, when it is necessary to sample a target area (i.e., the unknown location of victims), it is suitable for fine-grained sampling of the distribution of victims. Figure 14b gives the optimization time for the number of RIS elements N. Fair RiFe, which keeps the optimization time nearly constant, outperforms RiFe, where the optimization time increases exponentially with respect to the number of RIS elements. Therefore, Fair RiFe enables the rapid configuration of RISs, thereby enabling the utilization of large-sized RISs in real-world scenarios.

[0107] Relevant research and its shortcomings are described below. In particular, air-to-ground (A2G) communications have recently been promoted as an emerging technology for providing advanced and sophisticated services. To date, effective A2G channel models have been introduced, experimentally validated through realistic measurements. In particular, A2G can be implemented to support vehicle-based use cases or to assist with emergency communications. In either case, methods based on optimization to maximize end-user coverage during UAV flight are proposed.

[0108] Recently, the integration of RIS (Reception Signaling) into UAV-based communications has been proposed as a means of improving the performance of air-to-ground networks. Previously, it has been proposed to use RIS to increase the power of received signals in UAVs. However, this approach has the limitation that it requires both the UAV and the user to be on the same side with respect to the RIS.

[0109] Therefore, a more feasible solution is to equip the UAV with a lightweight RIS. In this regard, several studies have investigated the problem of simultaneously optimizing the UAV trajectory and RIS parameters. To date, it has been proposed that reinforcement learning methods based on Q-learning and neural networks be used to optimize both the UAV trajectory for maintaining a line-of-sight (LoS) link with a ground user and the RIS parameters for maximizing downlink transmit capacity. Previously, maximizing the secure energy efficiency of this system was addressed by simultaneously optimizing the UAV trajectory, phase shift in the RIS, user association, and transmit power using successive convex approximation.

[0110] Similar techniques are also used to maximize the achievable rate for a single ground user. To date, the problem of maximizing the worst-case SNR in a given target area covered by the BS by appropriately selecting the RIS configuration and the placement of the airborne platform has been considered. The resulting non-convex problem is solved by finding the RIS configuration by maximizing the worst-case array gain and then separating the optimization in finding the placement of the airborne platform by balancing the resulting angular range with the path loss of the downlink channel equivalent.

[0111] However, existing studies, for simplicity, assume that the aerial platform (e.g., UAV or HAP) flies in perfect stability. Based on real-world applications, the present invention considers cases where the UAV is subjected to weather-induced motion. In particular, a mathematical framework is described that takes into account undesirable motion while continuously configuring the onboard RIS to pursue the maximization of overall performance in terms of fairness within the target area and the minimum experienced SNR.

[0112] As a result and as a conclusion, a sensitive and flexible air-to-ground network presents a new frontier for reliable communication. In the present invention, a passive device, i.e., an unmanned aerial vehicle (UAV) equipped with a reconfigurable intelligent surface (RIS) that can control the propagation characteristics of incoming signals to provide connectivity in emergency scenarios while keeping the energy burden within a reasonable range, is described.

[0113] Among the novel features, one novelty of the present invention using the RiFe solution is the robustness and reliability of the UAV communication channel by appropriately countering undesirable flight effects, such as position jitter and UAV orientation wobbling. In addition to the optimization framework, another embodiment, Fair-RiFe, is described, and also, Fair-RiFe reduces the complexity of the optimal solution. Finally, a thorough simulation campaign is conducted to verify the above framework, where the state-of-the-art solutions are greatly outperformed (e.g., by 25 decibels (dB)).

[0114] matrix

Number

Number

[0115]

Number

number

number

number

number

number

number

number

number

number

number

number

number

[0116] In each of the embodiments described, the embodiment may include one or more computer entities (e.g., a system, user interface, computing device, server, dedicated computer, smartphone, tablet, or computer configured to perform functions specified herein) including one or more processors and memory. The processor may include one or more different processors, each having one or more cores and capable of accessing memory. Each of the different processors may have the same or different structure. The processor may include one or more central processing units (CPUs), one or more graphics processing units (GPUs), circuits (e.g., application-specific integrated circuits (ASICs)), digital signal processors (DSPs), etc. The processor may be mounted on a common board or on multiple different boards. The processor is configured to perform a particular function, method, or operation (e.g., to provide the performance of a function, method, or operation) when at least one of the different processors is capable of performing an operation that embodies the function, method, or operation. A processor can perform actions that embody a function, method, or operation, for example, by executing code stored in memory (e.g., interpreting a script) and / or by exchanging data through one or more ASICs. A processor can be configured to automatically perform any and all of the functions, methods, and operations disclosed herein. Accordingly, a processor can be configured to implement any (e.g., all) of the protocols, devices, mechanisms, systems, and methods described herein. For example, where this disclosure states that a method or device performs an operation or task "X" (or task "X" is performed), such statement should be understood to disclose that a processor is configured to perform task "X".

[0117] While embodiments of the present invention have been illustrated and described in detail in the drawings and the above description, such illustrations and descriptions should be considered illustrative or descriptive, and not restrictive. It will be understood that changes and modifications may be made by those skilled in the art within the scope of the appended claims. In particular, the present invention encompasses further embodiments having any combination of features from the different embodiments described above and below. Furthermore, the descriptions made herein characterizing the present invention refer to one embodiment of the present invention and not necessarily all embodiments.

[0118] Terms used in the claims should be interpreted to have the broadest reasonable interpretation consistent with the above-mentioned explanation. For example, the use of the article "a" or "the" when introducing an element should not be interpreted as excluding multiple elements. Similarly, the statement "or" should be interpreted as inclusive, so as not to exclude "A or B" unless it is clear from the context or the above-mentioned explanation that only one of A or B is intended. Furthermore, the statement "at least one of A, B and C" should be interpreted as one or more of the group of elements consisting of A, B, and C, and should not be interpreted as requiring at least one of each of the enumerated elements A, B, and C, whether A, B, and C are related as categories or otherwise. Furthermore, the phrases "A, B, and / or C" or "at least one of A, B, or C" should be interpreted as including any single entity from the enumerated elements, such as A; any subset from the enumerated elements, such as A and B; or the entire list of elements A, B, and C. [Explanation of Symbols]

[0119] 200 Environment 202 First Computing Device 204 Signal 206 The Second Computing Device 210 UAV 212 icons 302 RIS configuration processors 304 Network Block 306 UAV Sensor 308 Optimization Feature Derivation Block 310 Maximizing the minimum expected signal-to-noise ratio (SNR) across the entire area sample. 312 UAV Position and Vibration Statistics Extraction Block 314 Area Sampling Blocks

Claims

1. A method for establishing direct communication using an unmanned aerial vehicle (UAV) having a reconfigurable intelligent surface (RIS), The steps include configuring RIS parameters based on compensating for undesirable positional and orientational vibrations associated with the UAV, A step of steering a signal reflection associated with a signal beam to a target area using the RIS of the UAV based on the RIS parameters, wherein the signal beam is from a transmitter. Methods that include...

2. The method according to claim 1, further comprising the step of configuring beamforming for steering the signal beam toward the UAV by the transmitter, wherein the transmitter is a base station.

3. The method according to claim 1, wherein the RIS parameter is an input voltage to one or more RIS elements that determines one or more phase shifts induced by the signal beam.

4. The method according to claim 1, wherein the step of configuring the RIS parameters includes determining the RIS parameters using an optimization algorithm that compensates for the undesirable vibrations based on a second-order statistic of the vibrations associated with the UAV.

5. The method according to claim 4, wherein the optimization algorithm is based on a mathematical optimization tool and / or an artificial intelligence algorithm.

6. To compensate for the undesirable vibrations of the UAV, continuously configure new RIS parameters, and Using the new RIS parameters, continuously steer the signal reflection associated with the signal beam. The method according to claim 1, further comprising the step of maintaining the direct communication between the target area and the transmitter using the RIS of the UAV based on the above.

7. A step of obtaining sensor measurements from one or more sensors of the UAV, wherein the sensor measurements indicate one or more coordinates and orientations of the UAV. A step of determining statistics for the UAV related to the one or more coordinates and orientation of the UAV. It further includes, The method according to claim 1, wherein the step of configuring the RIS parameters is further based on statistics of the UAV.

8. The method according to claim 7, wherein the step of determining statistics of the UAV relating to the one or more coordinates and orientation of the UAV is based on constructing a scaled histogram of the sensor measurements.

9. The steps include obtaining the area where the service will be provided from the transmitter, The steps include performing area sampling based on the area where the service is provided, and It further includes, The method according to claim 1, wherein the step of configuring the RIS parameters is based on performing the area sampling.

10. The step of performing the area sampling is, Based on the fact that information regarding the location of users within the area where the service is provided is unavailable, sample points belonging to the area are generated as equally spaced points, Based on the availability of a probability density function (pdf) for the user's location within the area where the service is provided, the sample points belonging to the area are generated according to the pdf. The method according to claim 9, including the method described in claim 9.

11. The transmitter is a base station, and the method is Steps to obtain the coordinates of the base station from the base station. It further includes, The method according to claim 1, wherein the step of configuring the RIS parameters is based on the coordinates of the base station.

12. The method according to claim 1, wherein the step of configuring the RIS parameters is based on using a closed-form expression.

13. The method according to claim 1, wherein the step of configuring the RIS parameters is based on using semi-definite programming (SDP).

14. A system for establishing direct communication using an unmanned aerial vehicle (UAV) with a reconfigurable intelligent surface (RIS), A transmitter configured to transmit a signal beam to the UAV, The UAV having the RIS The UAV is equipped with, Based on compensating for undesirable positional and orientational vibrations associated with the UAV, the RIS parameters are configured as follows: Based on the RIS parameters, the RIS of the UAV steers the signal reflections associated with the signal beam to the target area. A system configured in such a way.

15. A tangible, non-temporary, computer-readable recording medium having instructions, wherein the instructions are executed individually or in combination by one or more processors, Configuring Reconstruction Intelligent Surface (RIS) parameters based on compensating for undesirable positional and directional vibrations associated with an unmanned aerial vehicle (UAV), wherein the UAV includes the RIS, Based on the RIS parameters, the RIS of the UAV steers a signal reflection associated with a signal beam to a target area, wherein the signal beam is from a transmitter. A tangible, non-temporary, computer-readable recording medium that provides the execution of a method including the following.

Images