A multi-intelligent super-surface omnidirectional broadcasting signal processing method and system

Abstract

The application relates to the field of information and communication technology, and discloses a multi-intelligent-surface omnidirectional sensing integrated broadcast signal processing method and system, which comprises the following steps: a base station transmits a positioning pilot signal; a plurality of intelligent surfaces utilize beamforming technology to reflect the positioning pilot signal to a broadcast area; while constructing a virtual line-of-sight link, the data information is modulated onto the phase of the reflected wave by utilizing reflection modulation technology; a user terminal receives the reflected signals of the intelligent surfaces, constructs a factor graph model containing position information and data information, and updates the position information and the data information through iterative transmission of messages between variable nodes and factor nodes; and based on the iteratively converged messages, the joint estimation and detection results of the user position information and the data information are output. The base station only needs to transmit the positioning pilot signal, mutual interference between the positioning signal and the communication signal is avoided, the hardware architecture of the base station is simplified, and the positioning accuracy and the data detection reliability are improved.

Description

Technical Field

[0001] This application relates to the field of information and communication technology, and more specifically to a method and system for integrated sensing broadcast signal processing using a multi-intelligent metasurface. Background Technology

[0002] With the rapid development of mobile communication technology, the integration of communication and sensing has become one of the key enabling technologies for sixth-generation mobile communication systems. Smart metasurfaces, as two-dimensional metamaterial arrays composed of numerous low-cost passive reflective elements, can reflect incident electromagnetic waves to a specified direction by adjusting the phase of each reflective element. Because they can reconstruct the wireless propagation environment in a low-cost and low-power manner, they are considered an important auxiliary technology in integrated communication and sensing systems. A typical approach is to maximize the receiving gain of the target user by jointly optimizing the beamformers of the base station and the smart metasurface when the base station cannot establish a direct line-of-sight link with the user, or to divide the smart metasurface into two functional parts to reflect communication signals and sensing signals respectively.

[0003] However, the existing solutions mentioned above have the following technical problems: the base station needs to transmit communication signals and sensing signals simultaneously, resulting in complex hardware architecture and high energy consumption, which weakens the original hardware sharing advantages of integrated sensing; the smart metasurface only acts as a passive reflector for passive forwarding and does not participate in the active carrying and modulation of data information, thus limiting the system's spectral efficiency; the user end processes position estimation and data demodulation as independent tasks, ignoring the inherent coupling relationship between position information and data information in the physical layer signal model, which limits the sensing accuracy and data detection reliability, and the two signals interfere with each other at the smart metasurface. To solve at least one of the above problems, this application proposes a method and system for integrated sensing broadcast signal processing using multiple smart metasurfaces. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this application is to provide a method and system for integrated sensing broadcast signal processing using multiple intelligent metasurfaces, which can effectively solve the problems in the background technology. The specific technical solution of this application is as follows:

[0005] A method for integrated sensing broadcast signal processing using a multi-intelligent metasurface includes:

[0006] The base station sends a positioning pilot signal, and multiple smart metasurfaces use beamforming technology to reflect the positioning pilot signal to the broadcast area. While constructing a virtual line-of-sight link, reflection modulation technology is used to modulate the data information onto the phase of the reflected wave.

[0007] The user terminal receives the reflection signals from each smart metasurface, constructs a factor graph model containing user location information and data information, and updates the location information and data information based on the factor graph model through iterative message transmission between variable nodes and factor nodes.

[0008] Based on the messages after iterative convergence, the joint estimation and detection results of user location information and data information are output, completing the integrated broadcast signal processing of communication and positioning.

[0009] Specifically, the intelligent metasurface modulates data information onto the phase of the reflected wave using reflection modulation technology while performing beamforming, including:

[0010] Based on the incident wave direction angle and the desired reflected wave direction angle of the smart metasurface, calculate two phase gradients along the row and column directions of the metasurface;

[0011] Based on the data to be broadcast, the corresponding reflection modulation phase symbol is selected from the preset phase symbol set according to the preset bit mapping relationship;

[0012] For each reflection unit, the adjustable phase of the reflection unit is calculated by adding the product of the row index and the row direction phase gradient, the product of the column index and the column direction phase gradient, and the reflection modulation phase symbol.

[0013] Based on the calculated adjustable phase of each reflection unit, the phase state of each reflection unit is independently adjusted so that the incident positioning pilot signal, after reflection, points its beam towards the broadcast area, and the phase of the reflected wave carries the data information.

[0014] Specifically, the two phase gradients are calculated as follows:

[0015] Acquire the pitch and azimuth angles of the incident wave when it arrives at the smart metasurface, and the pitch and azimuth angles of the expected reflected wave when it leaves the smart metasurface;

[0016] The phase gradient along the metasurface direction is equal to the product of the sine of the elevation angle of the reflected wave and the cosine of the azimuth angle of the reflected wave, plus the product of the sine of the elevation angle of the incident wave and the cosine of the azimuth angle of the incident wave.

[0017] The phase gradient along the metasurface column direction is equal to the product of the sine of the elevation angle of the reflected wave and the sine of the azimuth angle of the reflected wave, plus the product of the sine of the elevation angle of the incident wave and the sine of the azimuth angle of the incident wave.

[0018] Specifically, the construction of the factor graph model, which includes user location information and data information, includes:

[0019] Using the reflection modulation data symbols of each smart metasurface in each OFDM symbol as elements, a data symbol vector is constructed, the dimension of which is equal to the total number of smart metasurfaces;

[0020] Define the three-dimensional position coordinates of the user terminal in the m-th OFDM symbol as the position variable to be estimated;

[0021] Based on the spatial geometric relationship between the signal arrival angle and the user position and the position of the smart metasurface, an angle-position probability constraint is established.

[0022] Based on the physical layer propagation relationship between channel gain and user location, smart metasurface location, and reflection-modulated data symbols, a joint probability constraint of channel-location and data is established.

[0023] Based on the data symbol vector, the location variable, the angle-location probability constraint, and the channel-location and data joint probability constraint, a factor graph model is constructed. The factor graph model includes location variable nodes, angle variable nodes, channel variable nodes, data variable nodes, and factor nodes.

[0024] Specifically, updating the location information and the data information through iterative message passing between variable nodes and factor nodes includes:

[0025] Based on the sum-product algorithm, the messages transmitted on each edge are updated sequentially in the order from variable node to factor node and then from factor node to variable node.

[0026] In each iteration, message updates are executed sequentially in the following eight steps: message updates from angle variable nodes to angle-position factor nodes; message updates from channel variable nodes to channel-position and data factor nodes; message updates from angle-position factor nodes to position variable nodes; message updates from channel-position and data factor nodes to position variable nodes; message updates from position variable nodes to angle-position factor nodes; message updates from position variable nodes to channel-position and data factor nodes; message updates from angle-position factor nodes to angle variable nodes; message updates from channel-position and data factor nodes to channel variable nodes.

[0027] Specifically, the message update is performed according to the following rules:

[0028] The messages passed from the angle variable node to the angle-position factor node are modeled using the von Mises distribution, which is used to describe the probabilistic characteristics of the periodic angle variable.

[0029] The messages transmitted from the channel variable node to the channel-location and data factor node are modeled using a complex Gaussian distribution, which is used to describe the probabilistic characteristics of the complex channel gain.

[0030] Specifically, the location variable node receives messages from two types of factor nodes and updates its confidence score, including:

[0031] The angle-position factor node transmits a first constraint message to the position variable node based on the von Mises distribution message of the angle variable node and the spatial geometric constraints determined by the user antenna direction vector and the smart metasurface position vector. The first constraint message is used to constrain the user position to be located in the spatial direction determined by the reflection angle.

[0032] The channel-location and data factor node transmits a second constraint message to the location variable node based on the complex Gaussian distribution message of the channel variable node and the prior distribution of the data symbol vector. The second constraint message is a weighted sum of multiple Gaussian distributions, where each Gaussian component corresponds to a possible value in the phase symbol set, and the weights of each Gaussian component are equal and are the reciprocal of the total number of phase symbols.

[0033] The location variable node multiplies its prior probability density function, the first constraint message, and the second constraint message to obtain the updated location confidence.

[0034] Specifically, the message updates transmitted from the location variable node to the angle-position factor node, and the message updates transmitted from the location variable node to the channel-position and data factor node, include:

[0035] The location variable node transmits messages to the angle-location factor node and the channel-location and data factor node respectively based on the currently updated location confidence. The messages are modeled using a multidimensional Gaussian distribution, and the mean vector and covariance matrix of the multidimensional Gaussian distribution are obtained by moment matching on the location confidence.

[0036] Specifically, the message update passed from the angle-position factor node to the angle variable node includes:

[0037] The angle-position factor node transmits a message to the angle variable node based on the Gaussian distribution message transmitted by the position variable node, combined with the spatial geometric constraints determined by the user-end antenna direction vector and the smart metasurface position vector. The message is modeled using a von Mises distribution.

[0038] Specifically, the message update transmitted from the channel-location and data factor node to the channel variable node includes:

[0039] The channel-location and data factor nodes transmit messages to the channel variable nodes based on the Gaussian distributed messages transmitted by the location variable nodes and the prior distribution of the data symbol vectors. The messages are modeled using a complex Gaussian distribution, and their covariance matrix is ​​the product of the identity matrix and the number of subcarriers.

[0040] Specifically, the estimated result of the output user location information includes:

[0041] Obtain the confidence level of the updated location variable node, wherein the confidence level is determined by the product of the prior probability density function of the location variable and the messages passed from all factor nodes to the location variable node;

[0042] The location variable is weighted and summed according to the confidence level over all possible values ​​to obtain the estimated user location.

[0043] Specifically, the detection results of the output data information include:

[0044] Using the estimated user location information, calculate the reference signal of the positioning pilot signal after it is reflected from the base station through each smart metasurface to the user terminal;

[0045] The actual received signal at the user terminal is correlated with the reference signal to eliminate the phase offset introduced by channel propagation;

[0046] The phase components of the relevant calculation results are extracted, and the phase components are mapped to corresponding bit data according to a preset bit mapping relationship, which serves as the detection result of the data information broadcast by the smart metasurface.

[0047] A multi-intelligent metasurface inductive integrated broadcast signal processing system, specifically comprising:

[0048] Base stations are used to transmit positioning pilot signals;

[0049] Multiple smart metasurfaces are used to receive the positioning pilot signal and reflect the positioning pilot signal to the broadcast area using beamforming technology. While constructing a virtual line-of-sight link, reflection modulation technology is used to modulate the data information onto the phase of the reflected wave.

[0050] The user terminal receives reflection signals from various smart metasurfaces, constructs a factor graph model containing user location information and data information, updates the location information and data information based on the factor graph model through iterative message transmission between variable nodes and factor nodes, and outputs the joint estimation and detection results of user location information and data information based on the message after iterative convergence.

[0051] The beneficial effects of this application are as follows: The base station only needs to transmit positioning pilot signals. While constructing a virtual line-of-sight link, the intelligent metasurface uses reflection modulation to load data information onto the phase of the reflected wave, so that the reflected wave carries both positioning information and data information. The base station does not need to transmit communication signals independently, avoiding mutual interference between positioning signals and communication signals and simplifying the base station hardware architecture. By constructing a factor graph model to model the location variables and data variables as a joint probability distribution, the joint estimation and detection of location information and data information is realized by using the message iterative transmission between variable nodes and factor nodes. This fully utilizes the inherent coupling relationship between the two in the physical layer signal model, avoids the error propagation problem caused by separate processing methods, and improves positioning accuracy and data detection reliability. Multiple intelligent metasurfaces simultaneously reflect signals from different spatial angles, providing multi-angle observation information for the factor graph model to accelerate iterative convergence. The factor graph model then uses joint inference to accurately decouple user location and data information from multi-angle signals. The combination of the two makes spatial diversity gain and joint inference capability complementary, improving the performance of the integrated sensing system in non-line-of-sight scenarios. By designing the phase gradient, the beam control problem of the two-dimensional reflective phased array is transformed into a linear combination of two one-dimensional phase gradients, reducing the complexity of reflection phase modulation. The factor graph message passing uses this low-dimensional parameterization structure to decompose the high-dimensional joint estimation problem into iterative updates of multiple low-dimensional messages, reducing the algorithm complexity from exponential to polynomial level, and realizing online real-time processing of the integrated sensing system. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a multi-intelligent metasurface inductive integrated broadcast signal processing method according to an embodiment of this application;

[0053] Figure 2 The figures show the simulation results of user location information and data information under different signal-to-noise ratio conditions in the embodiments of this application. In the figures, (a) is a comparison of positioning accuracy and (b) is a comparison of data detection reliability.

[0054] Figure 3 The figures show the simulation results of user location information and data information under different numbers of intelligent metasurfaces (RIS) in the embodiments of this application. In the figures, (a) shows the change in positioning accuracy with the number of RIS, and (b) shows the change in data detection reliability with the number of RIS. Detailed Implementation

[0055] The present application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0056] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0057] Hereinafter, the terms "first," "second," and other generic terms are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0058] The method diagram is as follows Figure 1 As shown, assume the base station has Each user is equipped with one transmitting antenna and communicates with users within a designated area via K smart metasurfaces. Each receiving antenna consists of a smart metasurface. It consists of several reflective units, among which and These represent the number of columns and rows, respectively. Both the base station and user antennas use linear antenna arrays. The spacing between the antennas and reflector elements is set to... ,in The wavelength is [wavelength]. Furthermore, since both the base station and the smart metasurface are fixed deployments, users can obtain this location information in advance. Consider [further details needed]. The user receives a signal from one subcarrier:

[0059]

[0060] in:

[0061] User at the The received signal matrix on all subcarriers within an OFDM symbol.

[0062] : No. During an OFDM symbol period, the matrix formed by stitching together the receive steering vectors of all smart metasurfaces arriving at the user end is:

[0063]

[0064] : No. During the OFDM symbol period, from the 1st The user terminal antenna array steering vector corresponding to the signal arriving at the user terminal from the smart metasurface.

[0065] : No. During the OFDM symbol period, the user terminal receives data from the first OFDM symbol. Arrival angle parameters of a smart metasurface reflected signal.

[0066] User at the The overall channel gain matrix under OFDM symbols, i.e.:

[0067]

[0068] : No. During the OFDM symbol period, the user and the first Channel gain between smart metasurfaces.

[0069] : No. The user's received noise matrix during one OFDM symbol period.

[0070] : Transpose of a matrix or vector.

[0071] Dimensions are The complex matrix space.

[0072] The Joint Detection and Estimation (JLIR) method using user location information and data information employed in this invention includes the following steps:

[0073] S1. Set up beamforming on the intelligent metasurface and use reflection modulation to modulate the data information. The first OFDM symbol The first intelligent metasurface The adjustable phase of each unit is expressed as:

[0074]

[0075] in, The phase symbol after data information reflection modulation. It is the size of the phase sign set; and These represent the phase gradients along the x-axis and y-axis of the metasurface, respectively, and are calculated as follows:

[0076] ,

[0077] in, These are the elevation and azimuth angles of the incident wave and the elevation and azimuth angles of the reflected wave, respectively.

[0078] S2, Definition Construct a factor graph model;

[0079] S3, based on the sum-product algorithm, in the... During each iteration, the angle variable node is calculated. Passed to the angle-position factor node The messages follow a von Mises distribution:

[0080] ,

[0081] in, Represents the von Mises distribution. Includes user location information, and These are the messages passed between the two parties. The mean and variance;

[0082] S4. Calculate the channel variable node Transmitted to channel-location and data factor nodes The messages follow a complex Gaussian distribution:

[0083]

[0084] in, Indicates a complex Gaussian distribution. It is an N×N identity matrix, where N is the number of subcarriers. and These are the messages passed between the two parties. The mean and variance of.

[0085] S5. Calculate the angle-position factor node. Passed to the position variable node The message is:

[0086] ,

[0087] in, Let be the unit direction vector of the user-end antenna. Let k be the position vector of the k-th smart metasurface. This represents the natural exponential function.

[0088] S6. Calculate the channel-location and data factor nodes. Passed to the position variable node The message is a weighted sum of multiple Gaussian distributions:

[0089] ,

[0090] in, Indicates a Gaussian distribution. and These are the messages passed between the two parties. The mean and variance of , where V is the size of the phase sign set.

[0091] S7, Calculate the position variable node. Passed to the angle-position factor node The messages follow a Gaussian distribution:

[0092] ,

[0093] in, and These are the messages passed between the two parties. The mean and variance of.

[0094] S8, Calculate from the position variable node Transmitted to channel-location and data factor nodes The messages follow a Gaussian distribution:

[0095]

[0096] in, and These are the messages passed between the two parties. The mean and variance of.

[0097] S9. Calculate the angle-position factor node. Passed to the angle variable node The messages follow a von Mises distribution:

[0098] ,

[0099] in, Represents the von Mises distribution. and These are the messages passed between the two parties. The mean and variance of.

[0100] S10, Calculate the channel-location and data factor nodes. Passed to the channel variable node The messages follow a complex Gaussian distribution:

[0101] ,

[0102] in, and These are the messages passed between the two parties. The mean and variance of.

[0103] like If the maximum number of iterations is exceeded, the iteration ends; otherwise, Return to S3.

[0104] S11, Estimated location information and data information They are respectively

[0105] ,

[0106]

[0107] in, for The prior probability density function, The symbol is a multiplication sign, where K is the total number of intelligent metasurfaces. To obtain the phase of a complex number, for The posterior probability density function, For the beamforming gain associated with the k-th smart metasurface, For subcarrier spacing, The number of reflective units on the intelligent metasurface. Let be the distance between the base station and the k-th smart metasurface. For wavelength, At the speed of light, Pilot signals sent by the base station Indicates the user's estimated location With the A smart metasurface The Euclidean distance between them.

[0108] The calculation is as follows:

[0109] ,

[0110] in, The direction vector at the base station end. Let the transmission angle of the k-th base station be . Indicates the row number of the reflection unit. Indicates the number of columns of the reflecting unit. These represent the elevation and azimuth angles of the incident wave from the intelligent metasurface, respectively. This indicates the conjugate transpose.

[0111] In this embodiment, the base station transmits a positioning pilot signal via a transmitting antenna array. This signal is a known reference signal sequence covering N subcarriers in the frequency domain. Multiple smart metasurfaces are deployed at different spatial locations within the base station's coverage area. The position coordinates of each smart metasurface are pre-measured and stored in the system. The position coordinates of the base station are also known information. When the positioning pilot signal reaches each smart metasurface, since the relative position between the base station and the smart metasurface is fixed, the direction angle of the incident wave on each smart metasurface can be calculated from the base station coordinates and the smart metasurface coordinates. Each smart metasurface determines the desired reflected wave direction angle according to the pre-configured broadcast area direction of the system, so that the reflected wave can cover all users within the target broadcast area.

[0112] The modulation order V determines the number of bits carried by each reflection modulation phase symbol. Specifically, V=4 carries 2 bits of data per symbol, and V=2 carries 1 bit of data per symbol. V=4 corresponds to quadrature phase shift keying (QPSK) modulation, with a phase symbol set of {0, π / 2, π, 3π / 2} and a bit mapping relationship of 00→0, 01→π / 2, 10→π, 11→3π / 2. V=2 corresponds to binary phase shift keying (BPSK) modulation, with a phase symbol set of {0, π} and a bit mapping relationship of 0→0, 1→π. The modulation order V is selected based on the trade-off between data rate requirements and demodulation reliability in actual communication scenarios: when the user terminal is in a high signal-to-noise ratio (SNR) environment or requires a high data rate, V=4 is chosen to improve spectral efficiency; when the user terminal is in a low SNR environment or needs to prioritize communication reliability, V=2 is chosen to increase the Euclidean distance between phase symbols and reduce the demodulation bit error rate. The system can pre-configure the value of V based on the broadcast area coverage, typical signal-to-noise ratio at the user end, and data transmission rate requirements, and maintain this value fixed during system operation without requiring dynamic adjustment due to instantaneous channel changes. The intelligent metasurface controller determines the corresponding reflection modulation phase symbol according to the aforementioned mapping table based on the bit data to be broadcast. .

[0113] To further illustrate the specific calculation process of the phase gradient, a set of numerical examples is given below. Assume that the incident wave elevation angle of a certain smart metasurface is... =30°, incident wave azimuth angle is =45°, the desired elevation angle of the reflected wave is =60°, azimuth angle of reflected wave is =-30°, substitute the above angle into the phase gradient calculation formula:

[0114] =sin(60°)·cos( +sin(30°)·cos(45°)

[0115] =0.8660×0.8660+0.5×0.7071

[0116] =0.7500+0.3536

[0117] =1.1036

[0118] =sin(60°)·sin +sin(30°)·sin(45°)

[0119] =0.8660× +0.5×0.7071

[0120] =-0.4330+0.3536

[0121] =-0.0794

[0122] The phase gradient of the intelligent metasurface was calculated to be: =1.1036、 =-0.0794. (This likely refers to a specific value or value, but without further context, it's difficult to translate accurately.) , and reflection modulation phase symbol Substitution This allows us to obtain the adjustable phase value of the (i,j)th reflective unit on the intelligent metasurface. Taking the (5,3)th reflective unit as an example, when... When =π / 2, =π×4×1.1036+π×2× +π / 2 = 4.4144π - 0.1588π + 0.5π = 4.7556π. After normalizing to the interval [0, 2π), this can be used as the control phase of the reflecting unit. The angle values ​​described in this example are only for demonstrating the calculation process. In actual deployment, the incident wave direction angle and reflected wave direction angle of each smart metasurface are determined by its spatial location and broadcast area.

[0123] Furthermore, while performing beamforming to construct a virtual line-of-sight link, the intelligent metasurface utilizes reflection modulation technology to modulate data information onto the phase of the reflected wave, including:

[0124] In this embodiment, the adjustable phase of the (i,j)th reflecting unit on the kth smart metasurface in the mth OFDM symbol time is expressed as the phase expression described above. The reflection modulation phase symbol of the data information from the kth smart metasurface in the mth OFDM symbol time is selected from a preset set of phase symbols, the size of which determines the modulation order of the data information. The value of the phase gradient determines the main lobe direction of the reflected wave, and its calculation method is as described above, determined jointly by the incident wave direction angle and the desired reflected wave direction angle. Specifically, the incident wave direction angle is determined by the direction of the line connecting the base station coordinates and the smart metasurface coordinates, and the desired reflected wave direction angle is determined by the direction of the line connecting the smart metasurface coordinates and the center coordinates of the broadcast area. Substituting the incident wave direction angle and the desired reflected wave direction angle into the aforementioned formula, the two directional phase gradients of the smart metasurface in the mth OFDM symbol time can be calculated. With the above configuration, the adjustable phase of each reflection unit is decomposed into three components: a component determined along the row direction by the row index and the row direction phase gradient, a component determined along the column direction by the column index and the column direction phase gradient, and a reflection modulation phase symbol determined by the data information. The first two components together achieve beamforming, while the third component achieves data modulation. The two components are linearly superimposed in the phase domain and do not interfere with each other.

[0125] Furthermore, the user terminal receives the reflection signals from each smart metasurface, constructs a factor graph model containing user location information and data information, and updates the location information and data information based on the factor graph model through iterative message passing between variable nodes and factor nodes, including:

[0126] S301. The user terminal receives the positioning pilot signals reflected by each smart metasurface through the receiving antenna array, demodulates the received signals to obtain the frequency domain received signals of each reflection link; after compensating the known channel components of the link from the base station to the smart metasurface, the frequency domain received signals of each reflection link are used as observation data and input into the factor graph model for joint inference.

[0127] S302. For any OFDM symbol, define the data symbol vector of each smart metasurface in the OFDM symbol time. The dimension of the data symbol vector is equal to the total number of smart metasurfaces, and each element is the reflection modulation data symbol of the corresponding smart metasurface in the current OFDM symbol time. Based on the data symbol vector, the user location information and the probability dependency relationship of the received signal on the OFDM symbol, construct a factor graph model.

[0128] S303, Based on the sum-product algorithm, in the... During each iteration, the message update loop between variable nodes and factor nodes is executed according to the preset message passing order;

[0129] S304. Determine the iteration termination condition. If the maximum number of iterations is exceeded, the iteration ends and the position and data information is output; otherwise, let... = +1, continue to the next iteration.

[0130] In this embodiment, the user terminal receives reflected signals from each smart metasurface via a receiving antenna array. Since the positions of the base station and each smart metasurface are fixed, the propagation paths of the reflected signals to the user terminal are different, and each reflected signal contains coupling information about the user's location. The user terminal directly inputs the received signals into a factor graph model for joint inference—in the factor graph, after compensating for the known channel components of the link from the base station to the smart metasurface, the frequency domain received signals of each reflected link are used as observation data. Through iterative message passing between variable nodes and factor nodes, the user's location, signal angle of arrival, channel gain, and the confidence level of the smart metasurface reflected modulation data symbols are updated simultaneously. Ultimately, joint estimation and detection of user location information and data information are achieved without needing to separately extract the angle of arrival or channel gain parameters before constructing the factor graph.

[0131] The user terminal uses its own three-dimensional position coordinates at the m-th OFDM symbol time as the position variable to be estimated. In probabilistic modeling, the user terminal constructs the following probabilistic dependencies: the signal arrival angle is determined by the spatial geometric relationship between the user position and the smart metasurface position, and there is a definite functional mapping between the two; the channel gain is jointly determined by the user position, the smart metasurface position, and the reflected modulated data symbols, where the user position affects the signal propagation distance, and the reflected modulated data symbols affect the phase rotation of the reflected signal; the likelihood function of the received signal under the above parameter conditions is determined by the statistical characteristics of the noise. Based on the above probabilistic dependencies, the user terminal constructs a factor graph model containing position variable nodes, angle variable nodes, channel variable nodes, data variable nodes, and various factor nodes.

[0132] The angle-position factor node The specific form of the probability function is determined by the spatial geometric relationship between the user's position and the position of the smart metasurface, and is a deterministic mapping in the form of a Dirac function. The arrival angle parameter is used to receive the reflection signal from the k-th smart metasurface at the user end. Let be the user's three-dimensional position vector at the m-th OFDM symbol time. This factor node characterizes the angle of arrival of the signal from the k-th smart metasurface to the user terminal, given the user's position, which is uniquely determined by the geometric relationship between the user's position and the smart metasurface position. = ,in Let be the unit direction vector of the user-end antenna. The message output by this factor node in the factor graph message passing is in the form of exp(κ⋅cos(θ−μ)), which reflects the spatial direction constraint of the arrival angle on the user position, where θ is the spatial angle determined by the user position and the metasurface position, and μ and κ are the mean and concentration of the message, respectively.

[0133] The channel-location and data factor nodes The specific form of the probability function is determined by the physical characteristics of the wireless propagation channel, and is a deterministic mapping of the Dirac function form. The channel gain between the user terminal and the k-th smart metasurface. Let be the reflection modulation phase symbol of the k-th smart metasurface in the m-th OFDM symbol time. This factor node characterizes how, given the user location and the reflection modulation phase symbol, the channel gain is uniquely determined by two parts: path loss and reflection modulation phase. The path loss part is determined by the user location. With metasurface position The distance between them is determined by the free-space path loss formula; the reflection modulation phase part is determined by the data symbols. The determination is manifested as a phase rotation of the channel gain. In factor graph messaging, the output message of this factor node follows a complex Gaussian distribution, reflecting the joint constraint of channel gain on user location and data symbols.

[0134] The parameters required for each factor node are obtained as follows. Noise variance is estimated by sampling the signal-free intervals in the received signal at the user end, or pre-determined during the calibration process at the system initialization stage. These parameters are pre-configured during system initialization and are slowly updated on a minute-level timescale according to environmental changes during system operation. The slow update is triggered when the average likelihood function of the received signals for multiple consecutive OFDM symbols falls below a preset threshold; at this point, the system re-estimates the parameters and updates the parameter configuration of the factor nodes.

[0135] Before executing message iterative passing, the location variable nodes in the factor graph are initialized. The prior distribution of the location variable nodes is provided by the user's GPS. During the initialization phase, the coarse location information output by the GPS is used as the initial prior information of the location variable nodes, and this coarse location information is modeled as a Gaussian distribution. Angle variable nodes and channel variable nodes do not need to be initialized; their initial confidence levels are naturally determined by the message passing rules of the factor graph model in the first iteration. Each variable node uses its initial prior distribution or initial confidence level as the basis for passing messages outward in the first iteration, thereby initiating the message passing loop of the factor graph.

[0136] Furthermore, the user end uses a sum-product algorithm for message passing on the factor graph model. In each iteration, messages are updated sequentially in the following order: angle variable node to angle-position factor node, channel variable node to channel-position and data factor node, angle-position factor node to position variable node, channel-position and data factor node to position variable node, position variable node to angle-position factor node, position variable node to channel-position and data factor node, angle-position factor node to angle variable node, and channel-position and data factor node to channel variable node. Each iteration completes all eight message update steps without interruption; only after all eight factor nodes have completed message output is the confidence distribution stored in the corresponding variable node updated synchronously. This message passing sequence forms a complete closed loop, ensuring that each variable node can integrate information from different factor nodes in each iteration. Specifically, the angle-position factor node outputs constraint messages about the user's position, which reflect the spatial geometric information inferred from the arrival angle parameter. The channel-location and data factor nodes output another constraint message regarding the user's location, reflecting propagation loss information inferred from channel gain and reflection modulation data. The location variable nodes fuse the received messages from the two types of factor nodes with their prior information, updating their probabilistic understanding of the user's location. This updated information is then passed back to each factor node to help correct the estimates of angle and data variables. After multiple iterations, the confidence levels of the various variable nodes gradually converge.

[0137] The iteration process terminates after reaching the preset maximum number of iterations. At this point, the confidence function of the location variable node integrates the angle and channel information provided by all smart metasurfaces, reflecting the user's actual location with high accuracy; the confidence function of each data variable node integrates the location estimation result and the channel estimation result, accurately recovering the data symbols modulated by each smart metasurface. Through the above iterative process, the user terminal can simultaneously complete its own positioning and data demodulation using only the signal reflected by the smart metasurface, without requiring the base station to transmit additional communication signals.

[0138] Iteration termination employs a dual-determination rule: iteration terminates when either of the following two conditions is met. The first condition is hard-determination: the maximum number of iterations is set to 10; if the current iteration count reaches 10, the iteration terminates. The second condition is soft-determination: calculate the norm difference between the distribution of all messages in the current iteration and the previous iteration; if this difference is less than... If the factor graph message passing is converged, the iteration terminates. If either condition is met, the iteration stops immediately, and the loop exits to estimate the output. If neither of the above two conditions is met, the next iteration continues.

[0139] Furthermore, in the message passing process, the specific forms of various messages include, but are not limited to: messages passed from angle variable nodes to angle-position factor nodes are modeled using a von Mises distribution, which is suitable for describing the probabilistic characteristics of periodic angle quantities; messages passed from channel variable nodes to channel-position and data factor nodes are modeled using a complex Gaussian distribution, which is suitable for describing the probabilistic characteristics of complex channel gains; messages passed from angle-position factor nodes to position variable nodes have an exponential function form, with its kernel function being an exponential form of the cosine function; messages passed from channel-position and data factor nodes to position variable nodes are modeled using a Gaussian mixture model, which is a weighted sum of multiple Gaussian distributions; messages passed from position variable nodes to each factor node are modeled using a multidimensional Gaussian distribution, with the mean vector and covariance matrix obtained by moment matching with the current confidence level of the position variable; messages passed from each factor node to angle variable nodes and channel variable nodes are modeled using a von Mises distribution and a complex Gaussian distribution, respectively. These various types of messages are mutually passed and updated during the iteration process, jointly driving the convergence of the factor graph model.

[0140] For the weighting rule of Gaussian mixture messages, the message transmitted from the channel-location and data factor nodes to the location variable nodes is modeled using a Gaussian mixture model. The weight of the mixture term is fixed at 1 / V, where V is the modulation order (i.e. the size of the phase symbol set). All phase symbols are assumed to be summed with equal probability, and no additional prior weighting coefficients are required.

[0141] Furthermore, based on the messages after iterative convergence, the joint estimation and detection results of user location information and data information are output, including:

[0142] S401. Calculate the estimated user location based on the confidence function of the location variable nodes after iterative convergence.

[0143] S402. Based on the confidence function of each data variable node after iterative convergence, detect the data information broadcast by each smart metasurface;

[0144] S403. Output the joint estimation and detection results of user location information and data information.

[0145] The user terminal ultimately outputs the user's three-dimensional spatial coordinates and the data sequence broadcast by each smart metasurface. The location information can be used for location-based services or as initial values ​​for subsequent positioning and tracking, while the data is reconstructed into user-readable broadcast content. Since the location estimation and data detection are jointly performed within the same factor graph framework, they fully utilize the inherent coupling relationships in the physical layer signal model, resulting in higher accuracy and reliability compared to processing them independently.

[0146] In this embodiment, by Figure 1 The illustrated method shows that the base station only needs to transmit a positioning pilot signal. The intelligent metasurface performs reflection modulation of the data information while simultaneously beamforming. The user terminal achieves joint estimation of location and data through factor graph iteration. Throughout the process, the base station does not need to configure a dedicated communication signal transmission link, avoiding mutual interference between communication and positioning signals and reducing system hardware complexity. The spatial diversity configuration of multiple intelligent metasurfaces provides sufficient multi-angle observation information for the factor graph model, enabling the user terminal to obtain high-precision positioning and communication services through a virtual line-of-sight link even in non-line-of-sight environments.

[0147] Furthermore, to verify the effectiveness of the method proposed in this application, a simulation experiment is provided in this embodiment. Specific parameter settings are shown in the table below:

[0148] parameter symbol Value Subcarrier spacing 15 kHz Number of subcarriers 20 Number of transmitting antennas 64 Number of receiving antennas 16 Number of intelligent metasurfaces 4 Number of reflective units L Smart Metasurface Position (0m,18m,12m),(15m,20m,10m), (0m,6m,15m),(20m,15m,10m),(5m,10m,17m),(18m,12m,5m) User direction vector (0,1,0) User location (8m, 8m, 0m)

[0149] like Figure 2 As shown, simulation results of user location information and data information under different signal-to-noise ratios are presented. Figure 2 The horizontal axis represents the signal-to-noise ratio, in dB. Figure 2 In (a), the vertical axis represents the root mean square error of the positioning. Figure 2 In (b), the vertical axis represents the symbol error rate; a logarithmic scale is used. In the figure, ● (solid circular line) represents the Joint Localization and Information Recovery (JLIR) method proposed in this application, ■ (dashed square line) represents the Multi-Variational Bayesian Line Spectral Estimation (MVALSE) algorithm, and ▲ (dashed triangle line) represents the Multi-Newtonized Orthogonal Matching Pursuit (MNOMP) algorithm. From Figure 2 As can be seen, with the increase of signal-to-noise ratio, the root mean square error of localization of the three algorithms gradually decreases, and the bit error rate gradually decreases. The localization error and bit error rate of the method proposed in this application are lower than those of the existing MVALSE and MNOMP algorithms. The performance improvement is due to the fact that the factor graph iterative method designed in this application uses a message loop iterative approach, which can more accurately obtain the prior information of the estimated parameters and avoid the error propagation problem caused by the separate processing method.

[0150] like Figure 3 As shown, simulation results of user location information and data information are presented under different numbers of smart metasurfaces. Figure 3 The horizontal axis represents the number of smart metasurfaces; Figure 3In (a), the vertical axis represents the root mean square error of the positioning, using a logarithmic coordinate system; Figure 3 In (b), the vertical axis represents the symbol error rate. In the figure, ● (solid circular line) represents the Joint Detection and Estimation Method (JLIR) proposed in this application, ■ (dashed square line) represents the Variational Bayesian Line Spectrum Estimation Algorithm (MVALSE), and ▲ (dotted triangle line) represents the Newtonized Orthogonal Matching Pursuit Algorithm (MNOMP). From Figure 3 It can be seen that as the number of smart metasurfaces increases, the root mean square error of positioning for all three algorithms gradually decreases, and the bit error rate also gradually decreases. This result demonstrates that the proposed method can effectively utilize the spatial diversity characteristics of multiple smart metasurfaces, achieving satisfactory synesthetic integration performance even with a limited number of smart metasurfaces deployed.

[0151] like Figure 2 and Figure 3 As shown, the performance comparison between the proposed joint detection and estimation method (labeled JLIR in the figure) and existing algorithms such as MVALSE and MNOMP is presented. The results show that the proposed algorithm outperforms existing methods in both positioning accuracy and data detection reliability, verifying the effectiveness and superiority of the technical solution presented in this application.

[0152] A multi-intelligent metasurface-based integrated sensing broadcast signal processing system, used to implement the aforementioned multi-intelligent metasurface-based integrated sensing broadcast signal processing method, includes:

[0153] Base stations are used to transmit positioning pilot signals;

[0154] Multiple smart metasurfaces are used to receive the positioning pilot signal and reflect the positioning pilot signal to the broadcast area using beamforming technology. While constructing a virtual line-of-sight link, reflection modulation technology is used to modulate the data information onto the phase of the reflected wave.

[0155] The user terminal receives reflection signals from various smart metasurfaces, constructs a factor graph model containing user location information and data information, updates the location information and data information based on the factor graph model through iterative message transmission between variable nodes and factor nodes, and outputs the joint estimation and detection results of user location information and data information based on the message after iterative convergence.

[0156] In this embodiment, the base station only needs to be equipped with a radio frequency link and a baseband processing module for transmitting positioning pilot signals, without the need for a separate communication signal transmission link. This results in a simple hardware architecture, low cost, and low power consumption. The base station periodically transmits a known sequence of positioning pilot signals on the broadcast channel according to a preset time-frequency resource mapping rule. This signal sequence is continuously transmitted in the time domain and covers N subcarriers in the frequency domain. The base station can complete the transmission of pilot signals without relying on real-time channel state information feedback, resulting in low system implementation complexity.

[0157] Multiple smart metasurfaces are deployed at different spatial locations within the base station coverage area. Each smart metasurface is equipped with a controller and a reflective array. The controller receives the data to be broadcast via a wired connection and converts the data into a reflective modulation phase symbol according to a preset modulation mapping rule. Simultaneously, the controller calculates the phase gradient required for beamforming based on the base station location, its own location, and the broadcast area direction. The controller superimposes the phase gradient with the reflective modulation phase symbol to generate an adjustable phase control voltage for each reflective unit, which is then applied to the corresponding reflective unit. Each reflective unit adjusts its electromagnetic properties according to the control voltage, ensuring that the incident positioning pilot signal, after reflection, points its beam towards the broadcast area and contains data information in its phase. Multiple smart metasurfaces simultaneously reflect signals into the broadcast area from different spatial angles, forming spatial diversity and enhancing the signal coverage strength of the broadcast area.

[0158] The user terminal is equipped with a receiving antenna array, an RF front-end, and a baseband signal processing module. The receiving antenna array receives reflected signals from various smart metasurfaces. The RF front-end performs down-conversion and analog-to-digital conversion on the signals. The baseband signal processing module completes signal detection, parameter extraction, factor graph construction, message iteration, and joint estimation. The user terminal does not need to rely on a direct line-of-sight link between the base station and the user; it can obtain high-quality reflected signals through a virtual line-of-sight link established by the smart metasurfaces. The user terminal achieves parallel processing of positioning and communication through factor graph iteration, enabling it to recover broadcast data information while obtaining its own location information, meeting the needs of integrated sensing and communication application scenarios.

[0159] The above description is merely a preferred embodiment of this application. The scope of protection of this application is not limited to the above embodiments. All technical solutions falling within the scope of this application's concept are within the scope of protection of this application. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of this application should also be considered within the scope of protection of this application.

Claims

1. A method for integrated sensing broadcast signal processing using multiple intelligent metasurfaces, characterized in that, include: The base station sends a positioning pilot signal, and multiple smart metasurfaces use beamforming technology to reflect the positioning pilot signal to the broadcast area. While constructing a virtual line-of-sight link, reflection modulation technology is used to modulate the data information onto the phase of the reflected wave. The user terminal receives the reflection signals from each smart metasurface, constructs a factor graph model containing user location information and data information, and updates the location information and data information based on the factor graph model through iterative message transmission between variable nodes and factor nodes. Based on the iteratively converged message, the joint estimation and detection results of user location information and data information are output, and the integrated broadcast signal processing of communication and positioning is completed. The intelligent metasurface, while performing beamforming, utilizes reflection modulation technology to modulate data information onto the phase of the reflected wave, including: Based on the incident wave direction angle and the desired reflected wave direction angle of the smart metasurface, calculate two phase gradients along the row and column directions of the metasurface; Based on the data to be broadcast, the corresponding reflection modulation phase symbol is selected from the preset phase symbol set according to the preset bit mapping relationship; For each reflection unit, the adjustable phase of the reflection unit is calculated by adding the product of the row index and the row direction phase gradient, the product of the column index and the column direction phase gradient, and the reflection modulation phase symbol. Based on the calculated adjustable phase of each reflection unit, the phase state of each reflection unit is independently adjusted so that the incident positioning pilot signal is reflected and its beam points to the broadcast area, and the phase of the reflected wave carries the data information. The construction of the factor graph model, which includes user location information and data information, includes: Using the reflection modulation data symbols of each smart metasurface in each OFDM symbol as elements, a data symbol vector is constructed, the dimension of which is equal to the total number of smart metasurfaces; Define the three-dimensional position coordinates of the user terminal in the m-th OFDM symbol as the position variable to be estimated; Based on the spatial geometric relationship between the signal arrival angle and the user's position and the position of the smart metasurface, an angle-position probability constraint is established. Based on the physical layer propagation relationship between channel gain and user location, smart metasurface location, and reflection-modulated data symbols, a joint probability constraint of channel-location and data is established. Based on the data symbol vector, the location variable, the angle-location probability constraint, and the channel-location and data joint probability constraint, a factor graph model is constructed. The factor graph model includes location variable nodes, angle variable nodes, channel variable nodes, data variable nodes, and factor nodes.

2. The method according to claim 1, characterized in that, The two phase gradients are calculated as follows: Acquire the pitch and azimuth angles of the incident wave when it arrives at the smart metasurface, and the pitch and azimuth angles of the expected reflected wave when it leaves the smart metasurface; The phase gradient along the metasurface direction is equal to the product of the sine of the elevation angle of the reflected wave and the cosine of the azimuth angle of the reflected wave, plus the product of the sine of the elevation angle of the incident wave and the cosine of the azimuth angle of the incident wave. The phase gradient along the metasurface column direction is equal to the product of the sine of the elevation angle of the reflected wave and the sine of the azimuth angle of the reflected wave, plus the product of the sine of the elevation angle of the incident wave and the sine of the azimuth angle of the incident wave.

3. The method according to claim 1, characterized in that, The step of updating the location information and the data information through iterative message passing between variable nodes and factor nodes includes: Based on the sum-product algorithm, the messages transmitted on each edge are updated sequentially in the order from variable node to factor node and then from factor node to variable node. In each iteration, message updates are executed sequentially in the following eight steps: message updates from angle variable nodes to angle-position factor nodes; message updates from channel variable nodes to channel-position and data factor nodes; message updates from angle-position factor nodes to position variable nodes; message updates from channel-position and data factor nodes to position variable nodes; message updates from position variable nodes to angle-position factor nodes; message updates from position variable nodes to channel-position and data factor nodes; message updates from angle-position factor nodes to angle variable nodes; and message updates from channel-position and data factor nodes to channel variable nodes.

4. The method according to claim 3, characterized in that, The message update is performed according to the following rules: The messages passed from the angle variable node to the angle-position factor node are modeled using the von Mises distribution, which is used to describe the probabilistic characteristics of the periodic angle variable. The messages transmitted from the channel variable node to the channel-location and data factor node are modeled using a complex Gaussian distribution, which is used to describe the probabilistic characteristics of the complex channel gain.

5. The method according to claim 3, characterized in that, The location variable node receives messages from two types of factor nodes and updates its confidence score, including: The angle-position factor node transmits a first constraint message to the position variable node based on the von Mises distribution message of the angle variable node and the spatial geometric constraints determined by the user antenna direction vector and the smart metasurface position vector. The first constraint message is used to constrain the user position to be located in the spatial direction determined by the reflection angle. The channel-location and data factor node transmits a second constraint message to the location variable node based on the complex Gaussian distribution message of the channel variable node and the prior distribution of the data symbol vector. The second constraint message is a weighted sum of multiple Gaussian distributions, where each Gaussian component corresponds to a possible value in the phase symbol set, and the weights of each Gaussian component are equal and are the reciprocal of the total number of phase symbols. The location variable node multiplies its prior probability density function, the first constraint message, and the second constraint message to obtain the updated location confidence.

6. The method according to claim 3, characterized in that, The message updates transmitted from the location variable node to the angle-location factor node, and from the location variable node to the channel-location and data factor node, include: The location variable node transmits messages to the angle-location factor node and the channel-location and data factor node respectively based on the currently updated location confidence. The messages are modeled using a multidimensional Gaussian distribution, and the mean vector and covariance matrix of the multidimensional Gaussian distribution are obtained by moment matching on the location confidence.

7. The method according to claim 3, characterized in that, The message update passed from the angle-position factor node to the angle variable node includes: The angle-position factor node transmits a message to the angle variable node based on the Gaussian distribution message transmitted by the position variable node, combined with the spatial geometric constraints determined by the user-end antenna direction vector and the smart metasurface position vector. The message is modeled using a von Mises distribution.

8. The method according to claim 3, characterized in that, The message update transmitted from the channel-location and data factor node to the channel variable node includes: The channel-location and data factor nodes transmit messages to the channel variable nodes based on the Gaussian distributed messages transmitted by the location variable nodes and the prior distribution of the data symbol vectors. The messages are modeled using a complex Gaussian distribution, and their covariance matrix is ​​the product of the identity matrix and the number of subcarriers.

9. The method according to claim 1, characterized in that, The estimated results of the output user location information include: Obtain the confidence level of the updated location variable node, wherein the confidence level is determined by the product of the prior probability density function of the location variable and the messages passed from all factor nodes to the location variable node; The location variable is weighted and summed according to the confidence level over all possible values ​​to obtain the estimated user location.

10. The method according to claim 1, characterized in that, The detection results of the output data information include: Using the estimated user location information, calculate the reference signal of the positioning pilot signal after it is reflected from the base station through each smart metasurface to the user terminal; The actual received signal at the user terminal is correlated with the reference signal to eliminate the phase offset introduced by channel propagation; The phase components of the relevant calculation results are extracted, and the phase components are mapped to corresponding bit data according to a preset bit mapping relationship, which serves as the detection result of the data information broadcast by the smart metasurface.

11. A multi-intelligent metasurface inductive integrated broadcast signal processing system, characterized in that, For implementing the method as described in any one of claims 1 to 10, comprising: Base stations are used to transmit positioning pilot signals; Multiple smart metasurfaces are used to receive the positioning pilot signal and reflect the positioning pilot signal to the broadcast area using beamforming technology. While constructing a virtual line-of-sight link, reflection modulation technology is used to modulate the data information onto the phase of the reflected wave. The user terminal receives reflection signals from various smart metasurfaces, constructs a factor graph model containing user location information and data information, updates the location information and data information based on the factor graph model through iterative message transmission between variable nodes and factor nodes, and outputs the joint estimation and detection results of user location information and data information based on the message after iterative convergence.

Images