An intelligent metasurface assisted multi-target positioning system

Abstract

The application discloses a kind of intelligent metasurface assisted multi-target positioning system, belong to wireless communication field.The system includes: intelligent metasurface;Receiver;Multiple wireless access points, can periodically emit probe signal, and the probe signal can reach intelligent metasurface by multiple targets, reach receiver after being reflected by intelligent metasurface;Received signal model construction module is used to obtain the expression of the received signal of intelligent metasurface and receiver under different time block time slot;Virtual multidimensional signal construction module calculates the multidimensional signal according to the received signal of receiver, obtains the virtual steering vector received by receiver;And, spatial information positioning module is used to obtain the spatial spectrum of multiple signal classification algorithm;The steering vector corresponding to multiple peak values in spatial spectrum is obtained by searching the spatial spectrum, and the angle and distance of multiple targets are obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wireless communication, in particular to a multi-target positioning system assisted by intelligent metasurface. BACKGROUND

[0002] The prior art "Intelligent metasurface assisted passive target positioning method based on MUSIC and maximum parallelism (CN117075095A)" proposes a RIS assisted single target positioning system. On the one hand, the phase reconfigurable characteristics of RIS are used to construct a one-dimensional pseudo-MUSIC spatial spectrum based on the relationship between the characteristic matrix and the signal space and the noise space, so as to estimate the azimuth angle of the target and the base station. On the other hand, the time-varying characteristics of the intelligent metasurface are used to construct a channel cross-term codebook, and the parallelism operation is performed with the actual channel cross-term to obtain a parallelism spatial spectrum, so as to estimate the azimuth angle of the target and the intelligent metasurface, and then estimate the position of the single target, which has high precision and low computational complexity. The prior art "Multi-target positioning method and system (CN110376548B)" provides a multi-target positioning method. A plurality of anchor nodes are deployed inside a monitoring area. When N targets enter the monitoring area, the signals received by the targets from the anchor nodes in the monitoring area are directly collected. The anchor nodes corresponding to the strongest N signals are used as cluster centers, and a grid clustering algorithm is used to divide the monitoring area into N grids. In each grid, the positions of the multiple targets can be calculated by different methods according to the anchor node signals received by the targets.

[0003] The defects of the prior art mainly lie in: 1) The existing channel model is still based on the far-field plane wave assumption, and the distance domain information contained in the near-field spherical wave propagation characteristics has not been fully tapped. In the 6G high frequency band and super large array scene, the communication distance within tens of meters may be in the near-field region, and the traditional method is limited by the far-field assumption, resulting in insufficient distance dimension information positioning; 2) The existing multi-target positioning system mainly uses multi-antenna receivers to calculate the angle of arrival of the signal by using the phase difference or time difference of the signal in space. Single-antenna receiving systems cannot directly obtain spatial phase difference, making it difficult to realize angle resolution. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a multi-target positioning system assisted by intelligent metasurface.

[0005] The object of the present application can be achieved by the following technical solutions:

[0006] In a first aspect, the present application relates to a multi-target positioning system assisted by intelligent metasurface, comprising:

[0007] Intelligent metasurface, arranged in the near-field region of the target;

[0008] Receiver, located in the far-field region of the intelligent metasurface;

[0009] a plurality of wireless access points, capable of periodically transmitting probe signals, and the probe signals can reach the intelligent metasurface through a plurality of targets, and after being reflected by the intelligent metasurface, reach the receiver;

[0010] a received signal model construction module, configured to divide a probe signal period into a plurality of time blocks, each time block including a plurality of time slots, the reflection pattern of the intelligent metasurface being the same on each time block, and the reflection pattern being different between different time slots of the same time block, and obtaining an expression of the received signal of the intelligent metasurface and the receiver in different time block time slots;

[0011] a virtual multi-dimensional signal construction module, configured to accumulate the received signal of the receiver in each time block composed of a plurality of continuous time slots to form a multi-dimensional signal, and calculate the multi-dimensional signal according to the received signal of the receiver to obtain a virtual steering vector received by the receiver;

[0012] and a spatial information positioning module, configured to realize joint estimation of angle and distance parameters by orthogonality of a signal subspace and a noise subspace of the received signal of the receiver through a multiple signal classification algorithm, obtain a spatial spectrum of the multiple signal classification algorithm, and search the spatial spectrum to obtain steering vectors corresponding to a plurality of peaks in the spatial spectrum, and obtain angles and distances of a plurality of targets.

[0013] Optionally, the wireless access points and the intelligent metasurface are both configured with a uniform linear array, and the reflection units of the intelligent metasurface can control the phase of the incident signal.

[0014] Optionally, a near-field channel between the intelligent metasurface and the target can be modeled using a uniform spherical wave model, and the modeling method includes the following steps:

[0015] Let r k and θ k represent the distance and the angle of arrival of the kth (k = 1, 2, …, K) target to the center reflection unit of the RIS, respectively, and d R represent the distance between two adjacent reflection units of the RIS. The steering vector of the position of the target k to the RIS is:

[0016]

[0017] wherein, represents a complex domain, a m (r k , θ k ), m = -M, …, M represents the mth element of the steering vector; let r k,m represent the distance from the kth target to the mth reflection unit of the RIS, and the phase difference between the kth target and the mth reflection unit of the RIS and the center reflection unit of the RIS is calculated as:

[0018]

[0019] r k,m Using second-order Taylor expansion, we have:

[0020]

[0021] Based on Fresnel approximation, higher-order terms are ignored, so the phase difference can be approximated as:

[0022]

[0023] Considering all targets are in the far-field region of the AP, we assume d A denotes the spacing between adjacent elements of the AP antenna array, denotes the angle of departure of the AP antenna array to the kth target, then the steering vector of the AP to the kth target is:

[0024]

[0025] Let the distance from the AP to the kth target be d AT,k , the distance from the RIS to Rx be d RR , the path loss constant from the AP to the kth target and then to the RIS be γ ATR,k , and the path loss constant from the RIS to Rx be γ RR ; the channel fading from the AP to the kth target and then to the RIS is:

[0026]

[0027] where d0 is the reference distance. Therefore, the channel from the AP to the kth target and then to the RIS is modeled as:

[0028]

[0029] where ε k denotes the radar cross section of the kth target.

[0030] Alternatively, the steering vector of the smart metasurface to the receiver is:

[0031]

[0032] γ denotes the angle of departure from the RIS to Rx;

[0033] Therefore, the channel from the smart metasurface to the receiver is modeled as:

[0034] h RR = δb(γ)

[0035] where representing the channel fading of the RIS-to-Rx path.

[0036] In a second aspect, the present application relates to a method for smart surface assisted multi-target localization, comprising the following steps:

[0037] controlling a wireless access point to periodically transmit a probe signal, the probe signal being capable of reaching a smart surface through a plurality of targets, and reaching a receiver after being reflected by the smart surface;

[0038] the period comprises a plurality of time blocks, each time block comprises a plurality of time slots, the reflection pattern of the smart surface is the same in each time block, the reflection pattern is different between different time slots in the same time block, and an expression of the signal received by the smart surface and the receiver in different time block time slots is obtained;

[0039] accumulating the signal received by the receiver in each time block composed of a plurality of continuous time slots to form a multi-dimensional signal, and calculating the multi-dimensional signal according to the expression of the signal received by the smart surface to obtain a virtual steering vector;

[0040] performing joint estimation of angle and distance parameters by the orthogonality of the signal subspace and the noise subspace of the signal received by the receiver through a multiple signal classification algorithm, obtaining a spatial spectrum of the multiple signal classification algorithm, and obtaining the angles and distances of a plurality of targets by searching the spatial spectrum to obtain a plurality of peak values corresponding to the steering vectors in the spatial spectrum.

[0041] Optionally, the method for calculating the expression of the signal received by the smart surface comprises the following steps:

[0042] each time block comprises L time slots, and the total number of time blocks required for estimating the target (r k ,θ k ) is Q; in the same time block q, q = 1, …, Q, the wireless access point repeatedly transmits the same sensing signal in each time slot l, l = 1, …, L, which is denoted as which satisfies n = 1, …, 2N + 1, CN(0, 1) represents a complex Gaussian distribution with mean 0 and variance 1,

[0043] Let be the transmit beamforming matrix of the sensing signal, and let W be a unitary matrix, then the signal transmitted by the wireless access point in the lth time slot of the qth time block is represented as:

[0044]

[0045] the signal received by the smart surface in the lth time slot of the qth time block is:

[0046]

[0047] The smart metasurface uses the same reflection pattern on different time blocks, and the reflection pattern is different between different slots in the same time block, that is, the reflection phase shift matrix satisfies:

[0048] Φ ((q-1)L+1) ≠Φ (( q -1)L+2) ≠…≠Φ (qL)

[0049]

[0050] where l=1,…,L, q=1,…,Q, is the constant smart metasurface reflection pattern used in the lth slot of all time blocks;

[0051] Optionally, the calculation method of the expression of the signal received by the receiver comprises the following steps: the signal received by the receiver in the lth slot of the qth time block is represented as:

[0052]

[0053] where n (( q-1)L+l ) ~CN(0,σ 2 ) represents additive white Gaussian noise, σ 2 represents the noise power;

[0054] The above formula can be further written as:

[0055]

[0056] Optionally, the signal received by the receiver satisfies the condition of using a multiple signal classification algorithm, that is:

[0057] ①L>K

[0058] ②

[0059] ③ does not change over time

[0060] where is the covariance matrix of

[0061] Optionally, in the multiple signal classification algorithm, the joint estimation of the angle and distance parameters is realized by the orthogonality of the signal subspace and the noise subspace, and first the covariance matrix of the signals received in the Q time blocks is obtained as:

[0062]

[0063] Then Eigenvalue decomposition yields:

[0064]

[0065] where Λ=diag(λ1,λ2,…,λ L The diagonal elements of matrix Λ are the eigenvalues ​​of R, and U = [u1, ..., u2]. L The vectors are composed of the corresponding eigenvectors. Without loss of generality, we assume that λ1≥λ2≥…≥λ L ,definition Since the space is a noise subspace, the spatial spectrum used in multi-signal classification algorithms can be defined as:

[0066]

[0067] Therefore, the estimated value By performing a two-dimensional search, K peaks of the aforementioned spatial spectrum can be obtained, thus yielding K estimated distances. and angle

[0068] A third aspect of the present invention relates to a computer-readable storage medium storing instructions that, when executed, enable the above-described intelligent metasurface-assisted multi-target positioning method.

[0069] A fourth aspect of the present invention relates to a spatial positioning device, comprising the aforementioned computer-readable storage medium or the aforementioned intelligent metasurface-assisted multi-target positioning system.

[0070] The beneficial effects of this invention are:

[0071] (1) Compared with traditional far-field multi-target localization systems, the multi-target localization system based on RIS proposed in this invention can not only accurately estimate the distance-angle information of different targets when multiple targets are in different positions, but also estimate the position information of different targets when multiple targets are in the same angle.

[0072] (2) Compared with traditional multi-antenna receiving systems, this invention proposes a joint parameter estimation method by rationally designing the time-varying RIS phase shift matrix. The created time-domain multidimensional signal can be equivalent to multi-channel received signals, simulating the spatial phase difference characteristics of a multi-antenna system. Based on this, 2D-

[0073] The MUSIC algorithm can accurately estimate the range-angle information of multiple targets in a RIS-assisted target localization system. The proposed scheme enables the system to achieve localization accuracy comparable to traditional multi-antenna systems with a single receiver antenna architecture, while significantly reducing receiver complexity and cost. Attached Figure Description

[0074] The application will be further described below with reference to the drawings.

[0075] Figure 1 A system model for multi-target positioning based on intelligent metasurface of the present application;

[0076] Figure 2 A 2D-MUSIC algorithm spatial spectrum in the multi-target positioning method based on intelligent metasurface of the present application;

[0077] Figure 3 A distance spectrum obtained by the multi-target positioning method based on intelligent metasurface of the present application;

[0078] Figure 4 An angle spectrum obtained by the multi-target positioning method based on intelligent metasurface of the present application

[0079] Figure 5 A flowchart of the multi-target positioning method based on intelligent metasurface of the present application DETAILED DESCRIPTION

[0080] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0081] In some embodiments of the present application, a multi-target positioning method based on intelligent metasurface is disclosed, comprising the following steps:

[0082] Step 1: Establishing a system model

[0083] In this embodiment, a multi-target positioning method based on intelligent metasurface is provided. As shown in FIG. 1, the method comprises the following steps: Figure 1As shown, the multi-target positioning system based on the intelligent metasurface is composed of 1 wireless access point (Access Point, AP), K targets to be detected, 1 intelligent metasurface (Reconfigurable Intelligent Surface, RIS) and 1 receiver (Receiver, Rx). The sensing signal transmitted by the AP reaches the RIS after passing through multiple detection targets, and then reaches the Rx after being reflected by the RIS. The AP is equipped with 2N+1 (2N+1>K) antennas, the Rx is equipped with 1 antenna, the RIS has 2M+1 reflecting units, and the AP and the RIS are both configured with a uniform linear array (Uniform Linear Array, ULA). Considering the urban environment, the direct link between the target and the Rx is blocked by obstacles, so the RIS is deployed between the target and the Rx to provide a reflection link, and multiple targets are all in the near-field region of the RIS.

[0084] (1) RIS array model

[0085] The system deploys a passive RIS composed of 2M+1 reflecting units. Each reflecting unit can independently control the phase of the incident signal. Φ = diag[φ -M ,…,φ M ] T is the phase shift matrix of the RIS, where diag(x) represents a diagonal matrix, and the diagonal elements are all elements in the vector x, that is, the reflection coefficient, and the non-diagonal elements are all 0.

[0086] (2) AP—Targets—RIS channel model

[0087] The RIS is deployed in the near-field region of the K targets, so the near-field channel between the RIS and the target can be modeled by a uniform spherical wave (Uniform Spherical Wave, USW) model. Assuming that r k ,θ k respectively represent the distance and the angle of arrival of the kth (k = 1, 2, …, K) target to the center reflecting unit of the RIS, and d R represents the distance between two adjacent reflecting units of the RIS. Then the steering vector of the position of the target k to the RIS is:

[0088]

[0089] wherein is the complex domain, a m (r k ,θ k ), m = -M, …, M represents the mth element of the steering vector. Let r k,mLet dkmrepresent the distance from the kthtarget to the mthreflection element of the RIS, then the phase difference from the kthtarget to the mthreflection element of the RIS and the center reflection element of the RIS is calculated as:

[0090]

[0091] For r k,m Using the second-order Taylor expansion, we have:

[0092]

[0093] Based on the Fresnel approximation, the high-order terms above the second power are ignored, so the phase difference can be approximated as:

[0094]

[0095] Considering that all targets are in the far-field region of the AP, we assume d A Let d represent the angle of departure from the AP antenna array to the kthtarget, then the steering vector from the AP to the kthtarget is:

[0096]

[0097] Let d AT,k represent the distance from the AP to the kthtarget, d RR represent the distance from the RIS to the Rx, γ ATR,k represent the path loss constant from the AP to the kthtarget and then to the RIS, and γ RR represent the path loss constant from the RIS to the Rx. In the USW model, when the propagation distance is greater than a certain threshold, i.e., the uniform power distance, it can be considered that the received power can remain relatively stable, and a small change in distance will not cause a large fluctuation in power, i.e., the channel fading of the kthtarget to the 2M+1 elements of the RIS is approximately equal. Therefore, the channel fading from the AP to the kthtarget and then to the RIS is:

[0098]

[0099] where d0is the reference distance. Therefore, the channel modeling from the AP to the kthtarget and then to the RIS is:

[0100]

[0101] where ε k represents the Radar Cross Section (RCS) of the kthtarget.

[0102] (3) RIS—Rx far-field channel model

[0103] Rx is located in the far-field region of the RIS, let γ denote the angle of departure from the RIS to Rx, then the steering vector from the RIS to Rx is:

[0104]

[0105] Therefore, the channel from the RIS to Rx can be modeled as:

[0106] h RR = δb(γ)

[0107] where denotes the channel fading of the RIS-to-Rx path.

[0108] Second step: design of AP probing signal and RIS dynamic phase encoding

[0109] Assume that every L (L > K) time slots constitute a time block, and the total number of time blocks required to estimate the target (r k , θ k ) is Q. In the same time block q, q = 1, …, Q, the AP repeatedly transmits the same sensing signal in each time slot l, l = 1, …, L, denoted as which satisfies n = 1, …, 2N + 1, CN(0, 1) denotes a complex Gaussian distribution with mean 0 and variance 1, Let be the transmit beamforming matrix of the sensing signal, in order to achieve omnidirectional sensing, set W as a unitary matrix, then the signal transmitted by the AP in the lth time slot of the qth time block can be expressed as:

[0110]

[0111] The signal received by the RIS in the lth time slot of the qth time block is:

[0112]

[0113] Design the RIS to use the same reflection pattern on different time blocks, and the reflection pattern is different between different time slots in the same time block, that is, the reflection phase shift matrix satisfies the following two conditions:

[0114] Φ ((q-1)L+1 )≠Φ ((q-1)L+2) ≠…≠Φ (qL)

[0115]

[0116] where l = 1, …, L, q = 1, …, Q, is the constant RIS reflection pattern used in the lth time slot of all time blocks. Therefore, the signal received by Rx in the lth time slot of the qth time block is expressed as:

[0117]

[0118] where n (( q-1)L+l ) ~ CN(0,σ 2 ) represents an additive white Gaussian noise, σ 2 represents the noise power.

[0119] The above equation can be further written as:

[0120]

[0121] Third step: Establishing a virtual multi-dimensional received signal model

[0122] Recombine the signal terms in the above equation, where the equivalent channel of K targets to Rx is written as

[0123] (r, θ) = [(r1, θ1), (r2, θ2), …, (r K , θ K )] T The signals reflected at the K targets are written as Therefore, the above equation can be further equivalent to:

[0124]

[0125] Cumulate the received signals in each time block consisting of L > K consecutive time slots to form a multi-dimensional signal. Define as the signal vector received by Rx in the qth time block, which can be expressed as:

[0126]

[0127] where satisfies I L represents an L x L dimensional identity matrix, and has:

[0128]

[0129] where

[0130] Considering the RIS reflection pattern Φ (1) ≠ Φ (2) ≠ … ≠ Φ (L) , we have where rank(·) represents the rank operation of a matrix. Similarly, from we have which does not change with time q.

[0131] The system provided in this embodiment can be equivalent to a virtual multi-antenna receiving system, where the probe signals transmitted by the AP pass through K targets and are received by a virtual Rx using L receiving antennas, and the virtual Rx has a virtual steering vector in the direction θ k where the phase difference generated by the L virtual receiving antennas is simulated by designing different RIS reflection matrices on L time slots to perform positioning.

[0132] Step 4: Multi-target positioning

[0133] Using the above signal model, the signal received by the Rx satisfies the condition of using the Multiple Signal Classification (MUSIC) algorithm, that is:

[0134] ① L > K

[0135] ②

[0136] ③ does not change over time

[0137] where is the covariance matrix of .

[0138] Next, the 2D-MUSIC algorithm is used to realize joint estimation of angle and distance parameters through the orthogonality of the signal subspace and the noise subspace. First, the covariance matrix of the received signal on Q time blocks is obtained as:

[0139]

[0140] Then, R is subjected to Eigenvalue Decomposition (EVD) to obtain:

[0141]

[0142] where Λ = diag(λ1, λ2, …, λ L ), the diagonal elements of the matrix Λ are the eigenvalues of , U = [u1, …, u L ] is composed of the corresponding eigenvectors, and without loss of generality, it is assumed that λ1≥ λ2≥ … ≥ λ L , and is defined as the noise subspace, therefore, the spatial spectrum used in the MUSIC algorithm can be defined as:

[0143]

[0144] ​So the estimated value K peaks of the above spatial spectrum are obtained through two-dimensional search, so that K estimated distances And angles

[0145] To sum up, the embodiment proposes a multi-target positioning system based on RIS, which deploys RIS in the near field of multiple targets, uses a uniform spherical wave model to describe the near field channel between RIS and targets, and establishes a signal propagation model. Compared with the traditional plane wave-based model, the model can not only accurately estimate the distance-angle information of different targets when multiple targets are at different positions, but also estimate the position information of different targets when multiple targets are at the same angle.

[0146] The embodiment provides a time-domain multi-dimensional parameter joint estimation method based on time-varying RIS, which continuously transmits the same probe signal in multiple consecutive time slots, while the RIS dynamically adjusts its reflection phase shift matrix in each time slot. The receiver Rx synthesizes the received signals of different time slots to construct a virtual multi-dimensional received signal model, wherein the created virtual steering vector contains the distance-angle information of different targets, so that the Rx can realize the angle-distance joint positioning of multiple targets with only one receiving antenna.

[0147] In some specific embodiments of the application, in order to further prove the effect of the application, simulation experiments are carried out on the Matlab platform to evaluate the positioning performance of the method proposed in the above embodiment.

[0148] In Figures 2 to 4 , the distance and angle estimation spectrum of the method proposed in the above embodiment is shown, wherein the wavelength is set to 0.03m, the number of AP transmitting antennas is set to 2N+1=65, the number of RIS reflection units is set to 2M+1=201, the number of estimated time blocks is set to 12, the number of time slots of each time block is set to 6, the distance between the AP and the target is set to 100m, 110m, the distance between the target and the center of the RIS array is set to 5m, 6m, the angle of the target to the center of the RIS array is set to 30°, 30°, the distance between the center of the RIS array and the Rx is set to 75m, the angle of arrival of the RIS to the Rx is set to 60°, and it is assumed that the AP and the RIS are on the same horizontal line. The numerical results show that the estimated distance and angle of the paths of the two users to the RIS are (5.01m, 30°) and (5.99m, 30°) respectively, which are very close to the true values, and the positions of different targets can be estimated from the same angle.

[0149] In the description of the specification, the description of the terms "one embodiment", "an example", "a specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Also, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0150] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above embodiments, and the above embodiments and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application.

Claims

1. An intelligent metasurface-assisted multi-target positioning system, characterized in that, The method comprises the following steps: an intelligent metasurface is arranged in the near-field region of the target; a receiver is arranged in the far-field region of the intelligent metasurface; a plurality of wireless access points are arranged to periodically emit probe signals, and the probe signals can reach the intelligent metasurface through a plurality of targets and reach the receiver after being reflected by the intelligent metasurface; a received signal model construction module is configured to divide a probe signal period into a plurality of time blocks, each time block comprises a plurality of time slots, the reflection mode of the intelligent metasurface is the same in each time block, the reflection mode is different between different time slots in the same time block, and expressions of signals received by the intelligent metasurface and the receiver in different time block time slots are obtained; a virtual multi-dimensional signal construction module is configured to accumulate the signals received by the receiver in each time block composed of a plurality of continuous time slots to form a multi-dimensional signal, and calculate the multi-dimensional signal according to the signals received by the receiver to obtain a virtual steering vector received by the receiver; and a spatial information positioning module is configured to realize joint estimation of angle and distance parameters by orthogonality of a signal subspace and a noise subspace of the signals received by the receiver through a multiple signal classification algorithm, obtain a spatial spectrum of the multiple signal classification algorithm, and search the spatial spectrum to obtain steering vectors corresponding to a plurality of peaks in the spatial spectrum to obtain angles and distances of a plurality of targets. The signals received by the receiver satisfy the condition of using the multiple signal classification algorithm, that is: ①L>K In the multiple signal classification algorithm, joint estimation of angle and distance parameters is realized by orthogonality of a signal subspace and a noise subspace, and first, a covariance matrix of signals received in Q time blocks is obtained as follows: ② ③ does not change over time wherein is the covariance matrix; K is the number of targets to be detected; L is the number of slots contained in each time block; The wireless access points and the intelligent metasurface are both configured with uniform linear arrays, and the reflection units of the intelligent metasurface can control the phase of incident signals. represents the signal vector received by the receiver in the qth time block; Then the Eigenvalue decomposition is performed to obtain: where Λ=diag(λ1,λ2,…,λ L The diagonal elements of matrix Λ are... The eigenvalues, U = [u1, ..., u] L The vectors are composed of the corresponding eigenvectors. Without loss of generality, we assume that λ1≥λ2≥…≥λ L ,definition For the noise subspace, the spatial spectrum used in multi-signal classification algorithms is defined as: estimated values The K peaks of the spatial spectrum are found by a two-dimensional search, thus obtaining K estimated distances r k and angles θ k .

2. The intelligent metasurface-assisted multi-target positioning system of claim 1, wherein, The near-field channel between the intelligent metasurface and the target can be modeled by using a uniform spherical wave model, and the modeling method comprises the following steps:

3. The intelligent metasurface-assisted multi-target positioning system of claim 1, wherein, Based on Fresnel approximation, high-order terms higher than the second power are ignored, and therefore the phase difference can be approximated as: Let r k ,θ k denote the distance and the angle of arrival of the kth target to the center of the reflective unit of the smart surface, respectively, and d R denote the distance between two adjacent reflective units of the smart surface; then the steering vector of the kth target to the smart surface is: wherein, denotes the complex field, a m (r k ,θ k ), m = -M,..., M denotes the m-th element of the steering vector; let r k,m denotes the distance of the k-th target to the m-th reflecting element of the intelligent meta-surface, then the phase difference of the k-th target to the m-th reflecting element of the intelligent meta-surface and the center reflecting element of the intelligent meta-surface is calculated as: the intelligent meta-surface has 2M+1 reflecting elements; the wireless access point is equipped with 2N+1 antennas; For r k,m Using a second order Taylor expansion gives: Where d0 is a reference distance; the channel modeling from the wireless access point to the kth target to the intelligent metasurface is as follows: Consider all targets are in the far field region of the wireless access point, assume d A denotes the inter-element spacing of the wireless access point antenna array, denotes the angle of departure of the wireless access point antenna array to the kth target, then the steering vector of the wireless access point to the kth target is: Let the distance from the wireless access point to the kth target be d AT,k , the distance from the intelligent meta-surface to the receiver be d RR , the path loss constant from the wireless access point to the kth target to the intelligent meta-surface be γ ATR,k , the path loss constant from the intelligent meta-surface to the receiver be γ RR ; and the channel fading from the wireless access point to the kth target to the intelligent meta-surface be: The steering vector from the intelligent metasurface to the receiver is as follows: where ε k represents the radar cross section of the kth target.

4. The intelligent metasurface-assisted multi-target positioning system of claim 3, wherein, γ represents the angle of departure from the intelligent metasurface to the receiver; Therefore, the channel modeling from the intelligent metasurface to the receiver is as follows:

5. An intelligent metasurface-assisted multi-target positioning method, comprising the following steps: h RR = δb(γ) wherein denotes the channel fading of the smart metasurface to receiver path. controlling the wireless access points to periodically emit probe signals, and the probe signals can reach the intelligent metasurface through a plurality of targets and reach the receiver after being reflected by the intelligent metasurface; the period comprises a plurality of time blocks, each time block comprises a plurality of time slots, the reflection mode of the intelligent metasurface is the same in each time block, the reflection mode is different between different time slots in the same time block, and expressions of signals received by the intelligent metasurface and the receiver in different time block time slots are obtained; accumulating the signals received by the receiver in each time block composed of a plurality of continuous time slots to form a multi-dimensional signal, and calculating the multi-dimensional signal according to the expressions of the signals received by the intelligent metasurface to obtain a virtual steering vector. ​ The orthogonality of the signal subspace and the noise subspace of the signal received by the receiver is used to realize the joint estimation of the angle and distance parameters by the multiple signal classification algorithm, and the spatial spectrum of the multiple signal classification algorithm is obtained; The steering vectors corresponding to the peaks in the spatial spectrum are obtained by searching the spatial spectrum, and the angles and distances of the targets are obtained; The signal received by the receiver satisfies the condition of using the multiple signal classification algorithm, that is: ① L > K ② ③ does not change over time wherein is the covariance matrix; K is the number of targets to be detected; L is the number of slots contained in each time block; In the multiple signal classification algorithm, the orthogonality of the signal subspace and the noise subspace is used to realize the joint estimation of the angle and distance parameters, and the covariance matrix of the signal received in Q time blocks is obtained as follows: represents the signal vector received by the receiver in the qth time block; Then the Eigenvalue decomposition is performed to obtain: where Λ=diag(λ1,λ2,…,λ L The diagonal elements of matrix Λ are... The eigenvalues, U = [u1, ..., u] L The vectors are composed of the corresponding eigenvectors. Without loss of generality, we assume that λ1≥λ2≥…≥λ L ,definition For the noise subspace, the spatial spectrum used in multi-signal classification algorithms is defined as: estimated values The K peaks of the spatial spectrum are found by a two-dimensional search, thereby obtaining K estimated distances r k and angles θ k .

6. The intelligent metasurface-assisted multi-target positioning method of claim 5, wherein, The calculation method of the expression of the signal received by the intelligent metasurface includes the following steps: Each time block includes L time slots, the estimation target (r k , θ k ) requires a total of Q time blocks; within the same time block q, q = 1,..., Q, the wireless access point repeatedly transmits the same sensing signal in each time slot l, l = 1,..., L, which is set as which satisfies n = 1,..., 2N + 1, CN(0, 1) represents a complex Gaussian distribution with a mean of 0 and a variance of 1, Let To perceive the signal of the transmit beamforming matrix, set W as the unitary matrix, then the signal transmitted by the wireless access point in the lth time slot of the qth time block is represented as: The signal received by the intelligent metasurface in the lth time slot of the qth time block is: where H k is the channel model from the wireless access point to the kth target to the smart hypersurface; Satisfy the following two conditions: Φ ((q-1)L+1) ≠Φ ((q-1)L+2 )≠…≠Φ (qL) where l = 1,..., L, q = 1,..., Q, is the constant intelligent surface reflection pattern used in the l-th time slot of all time blocks; The calculation method of the expression of the signal received by the receiver includes the following steps: the signal received by the receiver in the lth time slot of the qth time block is expressed as: wherein denotes additive white Gaussian noise, σ 2 denotes the noise power; The above formula is further written as: b(y) is the steering vector from the intelligent metasurface to the receiver; a(r k , θ k ) is the steering vector from the position of the target k to the intelligent metasurface; ε k represents the radar cross section of the kth target; ∧ k is the channel fading from the wireless access point to the kth target to the intelligent metasurface; K is the total number of targets to be detected.

7. The intelligent metasurface-assisted multi-target positioning method of claim 6, wherein, The signal received by the receiver in the lth time slot of the qth time block is equivalent to: accumulating the signals it receives in each time block consisting of L > K consecutive time slots to form a multi-dimensional signal; for K targets to the equivalent channel of the receiver; for K targets to the signal reflected at the K targets; Definitions denotes the signal vector received by Rx at the qth time block, is denoted as: wherein satisfies I L denotes an L x L dimensional identity matrix and has: wherein By the reflection phase shift matrix of the smart metasurface: Φ (1) ≠ Φ (2) ≠... ≠ Φ (L) , we have where rank(·) denotes the rank operation of a matrix. get does not change over time q.

8. A computer-readable storage medium storing instructions, the instructions comprising: When the instructions are executed, the intelligent metasurface assisted multi-target positioning method of any one of claims 5-7 can be realized.

9. A spatial positioning device, characterized by The computer readable storage medium of claim 8 or the intelligent metasurface assisted multi-target positioning system of any one of claims 1-4.

Images