Internet of Things equipment position sensing and superposition pilot frequency transmission method

Abstract

The invention discloses an Internet of Things equipment position sensing and superposed pilot frequency transmission method, which relates to the technical field of Internet of Things, and comprises the following steps: firstly, constructing a wireless frame structure comprising two parts of position sensing and superposed pilot frequency transmission; secondly, based on a wireless frame structure, a user carries out position estimation according to a base station pilot signal reflected by a reconfigurable intelligent surface RIS; then, on the basis of a position estimation result, a closed expression of the uplink reachable rate of the user is established; then, according to the closed expression, constructing a weighting and rate maximization problem for optimizing the RIS phase, and solving by utilizing a genetic algorithm to obtain an optimized RIS phase; then, based on the optimized RIS phase, the user sends a pilot frequency and a data superposition signal, and the base station performs channel estimation after receiving the superposition signal; and based on the channel estimation result, the base station performs data detection by adopting a maximum ratio merging detector. According to the method, under the condition that spectrum resources and transmission power are limited, the UE position can be accurately obtained, and the system transmission rate is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a method for position perception and superimposed pilot transmission of Internet of Things devices. Background Art

[0002] Reconfigurable smart surface (RIS) is a promising wireless coverage enhancement technology that has been widely studied. However, many existing works are done under the assumption that instantaneous channel state information (CSI) is available, which needs to be estimated in each channel coherence interval. Estimating CSI is not easy for at least two reasons:

[0003] The first challenge is the communication overhead, especially in scenarios with a large number of UEs. In the conventional pilot (RP) transmission scheme, in order to reduce the interference of pilot and data, the two are transmitted separately, which leads to low resource efficiency. To improve resource efficiency, a new transmission scheme called superimposed pilot (SP) is proposed, which transmits pilot and data simultaneously, so that more samples can be used for channel estimation and data transmission. However, there is little work related to channel estimation and performance analysis in the RIS-assisted SP transmission scenario.

[0004] The second challenge comes from the lack of prior information. In the above works, it is assumed that the UE position is fixed and known a priori. However, this assumption is not always valid because some terminals (such as vehicles and drones) are mobile. Most of the existing RIS-assisted terminal positioning research focuses on analyzing statistical position errors (such as CRB), but does not give clear terminal position estimation results for channel estimation and further data decoding.

[0005] In summary, RIS-assisted SP transfer has not been well investigated and explored. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method for location perception and superimposed pilot transmission of Internet of Things devices. Under the condition of limited spectrum resources and transmission power, this method can accurately obtain the UE position and significantly improve the system transmission rate.

[0007] The present invention adopts the following technical solutions to solve the above technical problems:

[0008] A method for location perception and superimposed pilot transmission of an Internet of Things device proposed in the present invention includes:

[0009] Construct a wireless frame structure including position sensing and superimposed pilot transmission;

[0010] Based on the wireless frame structure, the user estimates the position according to the base station pilot signal reflected by the RIS and obtains the position estimation result;

[0011] Based on the position estimation result, a closed-form expression of the user's uplink achievable rate is established;

[0012] According to the closed-form expression of the user's uplink achievable rate, a weighted sum rate maximization problem for optimizing the RIS phase is constructed, and a genetic algorithm is used to solve the local optimal solution of the problem to obtain the optimized RIS phase.

[0013] Based on the optimized RIS phase, the user sends a pilot and data superposition signal. After receiving the superposition signal, the base station performs channel estimation and obtains the channel estimation result.

[0014] Based on the channel estimation result, the base station performs data detection to obtain a data detection result.

[0015] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, the wireless frame structure includes multiple location coherent intervals, and the frame length of one location coherent interval is τ L =τ P +n C τ C , where τ P is the time domain sequence length occupied by user positioning, τ C is the time domain sequence length of a channel coherence interval, τ L Includes 1 frame length of τ P The location awareness interval, and N C The frame length is τ C The channel coherence interval.

[0016] As a further optimization scheme of the method for location sensing and superimposed pilot transmission of an IoT device described in the present invention, based on the wireless frame structure, in the location sensing interval, the base station BS transmits a pure pilot sequence to the kth user UE. k , where k represents the subscript index value of the user UE, k∈{1,2,…,K}, and K is the number of UEs; then, UE k Use 2D positioning method to estimate UE k The angle information of the user itself is used to obtain the angle vector of the kth user. Estimated value of Angle vector θ k Refers to the distance from RIS to UE in the RIS local coordinate system. k The departure angle AOD includes the UE k Azimuth and the pitch angle of the kth user They are The estimated value of .

[0017] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, a closed-form expression of the uplink achievable rate of a user is established based on the location estimation result, which is as follows:

[0018] R k =log2(1+γ k )

[0019] in,

[0020]

[0021] Among them, for the kth user, R k It is UE k The closed-form solution of the uplink achievable rate, M B is the number of BS antennas; k About UE k of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy represents the mathematical expectation operation of the matrix, and the superscript H represents the conjugate transpose operation of the matrix. The number of rows is M B The unit array, is the linear minimum mean square error estimate LMMSE case g k The estimated value of k From UE k The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS; q k 、p k UE k The pilot normalized transmit power and data normalized transmit power, q i 、p i are the i-th user UE i The normalized pilot transmission power and data transmission power, p j is the jth user UE j The data normalized transmission power of , i,j∈{1,2,…,K}; χ k Yes k The values ​​of the main diagonal elements of the autocorrelation matrix and satisfy Δ k is to calculate g k The intermediate variable of the fourth-order moment satisfies ‖·‖ represents the 2-norm operation of a vector; c k About UE kThe weighting coefficient of the linear estimation that minimizes the linear mean square error between the channel estimation value and the true value; Δ i is to calculate g i The intermediate variable of the fourth-order moment satisfies g i From UE i The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ki Yes k and g i The expected value of the inner product modulus squared and satisfies and Ω ki =Ω ik , |·| represents the modulus operation of complex numbers, Ω ik Yes i and g k The expected value of the square of the inner product modulus, Ξ ij Yes i Vector inner product and g j The expected value of the product of vector inner products and satisfies And ij =Ξ ji ,Ξ ji Yes j Vector inner product and g i The expected value of the vector inner product, g j From UE j The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ij Yes i and g j The expected value of the inner product modulus squared and satisfies and Ω ij =Ω ki ;Ω ki Yes j and g i The expected value of the square of the inner product modulus, χ i Yes i The values ​​of the main diagonal elements of the autocorrelation matrix and satisfy λ i About UE i of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy is the linear minimum mean square error estimate LMMSE case g i Estimated value of ik Yes i Vector inner product and g k The expected value of the product of vector inner products and satisfies And ik=Ξ ki ,Ξ ki Yes k Vector inner product and g i The expected value of the product of vector inner products, γ k is the UE under the maximum ratio combining MRC detector k The closed-form solution of the effective signal-to-interference-and-noise ratio SINR;

[0022] In the uplink channel, for the kth user UE k , g k The expression is as follows:

[0023]

[0024] in, With the RIS phase matrix Λ and About k The LOS channel component of the line-of-sight link is constant in the same position coherence interval, but varies in different position coherence intervals. Indicates size is M B ×1 complex column vector; is the channel matrix of the link from RIS to BS, is the LOS channel component of H0, Indicates size is M R ×M B The complex matrix, M R =M R,x ×M R,z is the number of RIS elements, M R,x and M R,z are the lengths of the rows and columns of Λ respectively;

[0025] and Where m x ∈{1,2,...,M R,x},m z ∈

[0026] {1,2,...,M R,z}, Where e is the base of natural logarithms, J is the imaginary unit, is the phase of the nth reflection element in the RIS phase matrix, m x and m z are the indices of the row dimension and column dimension of Λ respectively, and n satisfies n∈{1,2,...,M R} and n=(m x -1)M R,z +m z, vec(Λ) is the vectorization of Λ, diag(vec(Λ)) is the diagonalization of vector vec(Λ), Indicates size is M R,x ×M R,z The complex matrix of ; From UE k The channel matrix of the link to RIS, Yes k The LOS channel component, Indicates size is M R ×1 complex column vector, ψ k From UE k LOS link amplitude gain to RIS, express The steering vector of

[0027] The expressions for the remaining variables are as follows:

[0028]

[0029] χ k =ρ B,k (ε0ε k |f k | 2 +(ε0+ε k +1)M R )

[0030] χ i =ρ B,k (ε0ε k |f i | 2 +(ε0+ε i +1)M R )

[0031]

[0032] in, express The steering vector, express The steering vector, It is UE i An estimate of the angle vector of ; is the path loss from RIS to BS, From UE k Path loss to RIS, From UE i Path loss to RIS; ρ B,k is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE kThe product of the large-scale fading of the NLOS link to the RIS, ρ B,i is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE i The product of the large-scale fading of the non-line-of-sight link NLOS to RIS; ε0 is the Ricean factor of the link from RIS to BS, ε k From UE k Ricean factor of the link to RIS, ε i From UE i The Rice factor of the link to the RIS; Represents θ R The steering vector, θ R is the angle of arrival AOA from BS to RIS in the RIS local coordinate system; f k is UE at time t k The total phase shift of the received signal, including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector f i is UE at time t i The total phase shift of the received signal, f i Including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector From UE k LOS channel to RIS link; Re{·} represents real part operation.

[0033] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, a weighted sum rate maximization problem of optimizing the RIS phase is constructed according to a closed-form expression of the user's uplink achievable rate. The mathematical model of the weighted sum rate maximization problem of optimizing the RIS phase is:

[0034]

[0035] Among them, κ k is the weight of the kth user, R k It is UE k The closed-form expression for the achievable rate is, Represents a set of UE subscript indices.

[0036] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, a weighted sum rate maximization problem of optimizing the RIS phase is constructed according to a closed-form expression of the user's uplink achievable rate, and a genetic algorithm is used to solve the local optimal solution of the problem to obtain the optimized RIS phase; wherein the steps of the genetic algorithm are as follows:

[0037] Step 61, initialize the population:

[0038] Determine the population size N t , each RIS phase array is regarded as an individual, and each individual in the population contains M R =M R,x ×M R,z chromosomes, the nth chromosome corresponds to the phase of the nth reflection unit in the RIS phase matrix Randomly generated in [0,2π);

[0039] Step 62: Adaptability evaluation and scaling:

[0040] ① Evaluate the fitness of each individual in the current population, fitness evaluation function The objective function of the mathematical model for the weighted sum rate maximization problem;

[0041] ② Measure the original fitness value of the individual according to its ranking in the population, sort the original fitness of the individual, and calculate their scaled fitness as follows:

[0042]

[0043] Among them, rank i′ is the original fitness function of individual i′ after being sorted in descending order, is the proportional fitness of individual i′, N c is a parameter used in the crossover operation, To calculate the individual i' Intermediate variables used when

[0044] Step 63. Select:

[0045] Select N from the current population e indivual Individuals with the largest values ​​serve as elite individuals, and these individuals produce the parents of their offspring;

[0046] Step 64, cross:

[0047] Select 2N based on random sampling c parents, used for crossover operation, to generate N c offspring;

[0048] Step 65, mutation:

[0049] Select the remaining N m =N t -N e -N c individuals are used for mutation operation, and each chromosome of each individual is represented by p m The probability of mutation is a random number between [0,2π), where N m is the number of mutant individuals, p m is the probability of gene mutation;

[0050] Step 66, fitness update:

[0051] Calculate the fitness evaluation function for each offspring

[0052] Step 67, Iteration:

[0053] Repeat steps 62-66 until The absolute value of the difference between the previous and next iterations is less than ε, or the number of iterations of steps 62-66 exceeds the maximum number of iterations N max , the iteration stops, and the optimized RIS phase matrix and weighted sum rate values ​​are output, where is the fitness evaluation function of all individuals is the average value of , and ε is the termination iteration error of the genetic algorithm.

[0054] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, based on the optimized RIS phase, the user sends a pilot and data superimposed signal, and the base station performs channel estimation after receiving the superimposed signal to obtain a channel estimation result; the details are as follows:

[0055]

[0056] in, is the BS received signal within a channel coherence interval, Indicates size is M B ×τ C The complex matrix of y k is the non-Gaussian signal of the kth user after despreading, It is UE k The pilot signal sent, It is UE k′ Pilot signal sent, UE k′ is the k′th user and satisfies and Represents the size as τ C ×1 complex column vector; is an additive white Gaussian noise (AWGN) matrix and the noise value obeys represents a complex Gaussian distribution with a mean of 0 and a variance of 1; v k It is UE k The sending data and v i It is UE i The sending data and Indicates that the mean is 0 and the covariance matrix is The complex Gaussian distribution of The number of rows is τ C The unit array.

[0057] As a further optimization scheme of the method for location perception and superimposed pilot transmission of an IoT device described in the present invention, based on the channel estimation result, the base station uses a maximum ratio combining detector to perform data detection to obtain data detection results of all users. The data detection result of the kth user is The details are as follows:

[0058]

[0059] in, It is UE i The pilot signal sent.

[0060] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0061] The present invention obtains the estimated position of the user through a positioning method, which can transform the CSI from completely unknown to partially known. Under the condition that there is almost no difference between the reachability calculated by the estimated position and the reachability calculated by the real position, the difficulty of channel estimation is greatly reduced and the calculation cost is saved; under the condition that the total energy cost of the system remains unchanged and the number of users is large, the SP scheme has certain advantages over the traditional RP scheme. The use of the SP scheme can improve the weighted sum rate of the user, and the data rate of the SP scheme after power optimization is higher than the data rate of the RP; the use of the genetic algorithm to optimize the RIS phase matrix can improve the channel environment, and the weighted sum rate of the user can be significantly improved through RIS beamforming. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a diagram of the frame structure of the transport protocol under consideration;

[0063] Figure 2 is a schematic diagram of the system model under consideration, UE krepresents the kth user, the dotted line represents the NLOS link between the two devices, the solid line represents the LOS link between the two devices, and “×” indicates that the corresponding link (including LOS and NLOS links) is blocked;

[0064] Figure 3 It is a flow chart of a short packet transmission method in a multi-user uplink mMIMO system based on URLLC transmission in a large-scale wireless network environment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0066] The present invention considers the uplink multiple-input multiple-output (MIMO) system assisted by RIS, and it is necessary to design a location-aware authorized RIS-assisted SP transmission scheme to solve the problems in the background technology. The present invention designs a RIS-assisted transmission protocol frame structure composed of multiple position coherent intervals, each of which has a pure pilot transmission interval and multiple channel coherent intervals. In the pure pilot transmission stage, a two-dimensional inverse Fourier transform 2D-IFFT positioning algorithm is proposed to estimate the UE position. On this basis, a weighted sum rate maximization problem under a given terminal transmission energy and number of terminal users is proposed. The local optimal solution of the problem is found through a genetic algorithm to achieve RIS phase optimization; in the superimposed pilot transmission stage, the channel state is first estimated, and then data detection is performed.

[0067] Symbol Description: In the symbol description, lowercase bold letters are used to represent vectors (such as x), uppercase bold letters are used to represent matrices (such as X), and x is used to represent T ,x H , ⊙,‖x‖, tr{x} and Re{x} represent the transposed matrix (vector), conjugate transposed matrix (vector), Kronecker product, Hadamard product, l2 norm, mathematical expectation, trace and real part of the matrix (vector), respectively. -1 represents the inverse matrix of X, and diag(a) represents the diagonalization of vector a into a matrix with main diagonal elements a. Represents a mean vector as The complex Gaussian vector a with covariance matrix ∑ is expressed as Represents a real Gaussian vector a, using Represents the set of complex numbers, and uses |·| to represent the modulus operation of complex numbers.

[0068] Consider the uplink transmission of a single-cell system, which consists of a base station BS, a RIS, and K single-antenna user UEs, where the BS is equipped with MB Root antenna, RIS equipped with M R Reconfigurable reflector unit. k represents the kth user, k represents the subscript index value of the user UE, and the set of UE subscript indexes is K is the number of UEs. The element spacing of the BS and RIS arrays is d = λ c / 2,λ c is the carrier wavelength. In addition, it is assumed that the BS, RIS and all user UEs are distributed in the far field area, and the transmission beam is a uniform plane wave, that is, the near-field radiation model of the electromagnetic wave is ignored.

[0069] The model geometry parameters are as follows:

[0070] (1) Selection of coordinate system

[0071] A global coordinate system is established based on the entire system, and the origin of the coordinate system is located at the geometric center of the RIS array; a local coordinate system is established based on the BS, and the origin of the coordinate system is located at the geometric center of the BS array; a local coordinate system is established based on the RIS, and the origin of the coordinate system is located at the geometric center of the RIS array. The directional relationship between the global coordinate system and the local coordinate system is determined by the rotation matrix.

[0072] (2) Location coordinates

[0073] The geometric center of BS is located at l B =(5,5,9)m, RIS geometric center is located at l R =(0,0,10)m,UE k Located in k =(l k,x ,l k,y ,0)m.

[0074] (3) Rotation Matrix

[0075] The rotation matrix V of the BS array B =I3, the rotation matrix V of the RIS array R =I3, I3 is the unit matrix with 3 rows.

[0076] (4) Angular coordinates

[0077] Consider a BS antenna array or RIS element array. When the electromagnetic wave leaves or arrives at a uniform planar array (UPA), the angle between the electromagnetic wave ray and the array in the three-dimensional local coordinate system of the array is θ = [θ az ,θ el ](angle of departure AOD or angle of arrival AOA), where θ az and θ elIn the downlink, electromagnetic waves propagate from the BS to the RIS, and then from the RIS to the UEs, so the following definitions apply:

[0078] ① BS-AOD from BS to RIS in BS local coordinate system;

[0079] ② RIS-AOA from BS to RIS in the RIS local coordinate system;

[0080] ③ From RIS to UE in RIS local coordinate system k RIS-AOD.

[0081] Among them, BS-AOD refers to the AOD angle vector of BS, RIS-AOA refers to the AOA angle vector of RIS, and RIS-AOD refers to the AOD angle vector of RIS.

[0082] In the uplink, electromagnetic waves propagate from UEs to RIS, and then from RIS to BS. Using the reciprocity of the channel, the angle vectors of the uplink and downlink are numerically the same and are represented by the same variable symbols. It only requires replacing AOA with AOD, or AOD with AOA. The specific definitions are as follows:

[0083] ① BS-AOA from RIS to BS in BS local coordinate system;

[0084] ② RIS-AOD from RIS to BS in the RIS local coordinate system;

[0085] ③ From UE in RIS local coordinate system k RIS-AOA to RIS.

[0086] Among them, BS-AOA refers to the AOA angle vector of BS, RIS-AOD refers to the AOD angle vector of RIS, and RIS-AOA refers to the AOA angle vector of RIS.

[0087] (5) Steering vector

[0088] Given a uniform planar array (UPA) and its local coordinate system. UPA is a two-dimensional matrix with M = M x ×M z elements, the minimum distance between elements is d, M x and M z are the lengths of the rows and columns of the UPA, respectively. In this coordinate system, the (m x ,m z) elements are located at

[0089]

[0090] Among them, m x and m z are the index of the row dimension and column dimension of UPA, respectively, m x ∈{1,2,...,M x},m z ∈{1,2,...,M z}.

[0091] In this coordinate system, the array steering vector can be calculated Among them, the steering vector ((m x -1)×M z +m z ) elements are

[0092]

[0093] Where ω(θ)=[sin(θ el )cos(θ az ), sin(θ el )sin(θ az ),cos(θ el )] T ω(θ) is the normalized beam vector of θ.

[0094] Before performing the optimization, the specific implementation of the frame structure of the considered transmission protocol is first described. Figure 1 As shown in Figure 1, the wireless frame structure consists of multiple position coherent intervals. The frame length of a position coherent interval is τ L =τ P +N C τ C , where τ P is the time domain sequence length occupied by user positioning, τ C is the time domain sequence length of a channel coherence interval, τ L Includes 1 frame length of τ P The location awareness interval, and N C The frame length is τ C In the position-aware interval, the BS first transmits a pure pilot sequence to the UE, then the UE estimates the position, and finally optimizes the RIS phase based on the estimated position to maximize the weighted sum rate of the UE; in the channel-coherent interval, the UE first transmits a pilot and data superposition sequence to the BS, then performs channel estimation, and finally uses the maximum ratio combining (MRC) detector to detect the UE's data sequence, and uses the Shannon formula to calculate the UE's uplink data achievable rate.

[0095] Assume that the UE position is constant within a position coherence interval and the channel state information CSI is constant within a channel coherence interval. Define the UE movement mode as a random walk model. For the kth user UE k , use n L Indicates the nth L The index of the position coherence interval is l k (0) indicates UE k At the frame header, UE k In the nth L The position of the position coherence interval can be expressed as l k (n L )=l k (n L -1)+Δl k (n L ),in Among them k Indicates that in a length of τ L The covariance matrix of the random walk over the time period .

[0096] Before performing the optimization, the specific implementation of the channel structure under consideration is described below. Figure 2 As shown, assuming that from BS to UE k The line-of-sight LOS link is blocked, but the LOS link from BS to RIS and from RIS to UE k The LOS link is available, that is, the BS and UE k There is a virtual LOS link between them. In addition, according to channel reciprocity, within the channel coherence interval, the uplink and downlink channels have the same CSI, which is mathematically expressed as the transposition relationship of the matrix.

[0097] Taking the uplink as an example, for the kth user UE k , from UE k The uplink concatenated channel vector consisting of the channel to RIS and the channel from RIS to BS is g k , the expression is as follows:

[0098]

[0099] in, is the channel matrix of the link from RIS to BS, is the LOS channel component of H0, Indicates size is M R ×M B The complex matrix of ; represents the RIS phase matrix, and Where m R,x∈{1,2,...,M R,x},m R,z ∈{1,2,...,M R,z}. Where e is the base of natural logarithms, J is the imaginary unit, is the phase of the nth reflection element in the RIS phase matrix, n satisfies n∈{1,2,...,M R} and n=(m R,x -1)M R,z +m R,z , vec(Λ) is the vectorization of Λ, diag(vec(Λ)) is the diagonalization of vector vec(Λ), Indicates size is M R,x ×M R,z The complex matrix, M R,x and M R,z are the lengths of the rows and columns of Λ respectively; From UE k The channel matrix of the link to RIS, Yes k The LOS channel component, Indicates size is M R ×1 complex column vector; Yes k The LOS channel component, is constant in the same position coherence interval, but varies in different position coherence intervals. Indicates size is M B ×1 complex column vector; R B,k Yes k The covariance matrix of . P In a time slot, the RIS phase is random and known at any time, and no channel feedback is required in the position estimation process; C In a time slot, the RIS phase is the phase optimized by the genetic algorithm, and the RIS phase is fixed at any time.

[0100] For the kth user, the model structure of each link is as follows:

[0101] (1) RIS-BS link

[0102] Channel fading follows the Rice distribution, and the channel matrix in and They represent the LOS channel component and the non-line-of-sight NLOS channel component of channel H0 respectively. RIS-BS refers to the distance from RIS to the base station.

[0103] ①LOS channel component

[0104]

[0105] in, represents the LOS link amplitude gain of RIS-BS, represents the path loss of RIS-BS, ε0 represents the Rice factor, Represents θ R The steering vector, Represents θ B The steering vector.

[0106] ②NLOS channel component

[0107] The elements in the matrix are subject to A random variable.

[0108] (2)RIS-UE k link

[0109] The channel fading follows the Rice distribution, and the channel vector in and They represent the channel h k The LOS channel component and NLOS channel component of RIS-UE k It refers to the distance from RIS to the kth user.

[0110] ①LOS channel component

[0111]

[0112] in, Indicates RIS-UE k The LOS link amplitude gain, Represents RIS-UE k The path loss, ε k represents the Rice factor, express The steering vector.

[0113] ②NLOS channel component

[0114] The elements in a vector are subject to A random variable.

[0115] like Figure 3 As shown, a flowchart of a method for RIS-assisted position sensing and superimposed pilot transmission in an uplink mMIMO system in a large-scale wireless network environment provided by an embodiment of the present invention includes the following steps:

[0116] Step 1: Position estimation. Based on the frame structure of the transmission protocol under consideration, position estimation is performed in the position sensing interval. Considering the MIMO downlink, for the kth user UE k ,UE k is the receiving end, and UE k The height of UE k To estimate the 2D position, a two-stage method is used: first, the 2D-IFFT algorithm is used to obtain θ k Estimated value of Then according to the geometric relationship, we can get l k Estimated value of That is, according to the azimuth, elevation and UE k The height of the UE is determined k Position. The formula is as follows:

[0117]

[0118] Step 2: Problem formulation: First, a closed-form solution to the achievable uplink rate of the user is given, and on this basis, a weighted sum rate maximization problem of optimizing the RIS phase is constructed, and finally a genetic algorithm is used to solve the local optimal solution of the problem.

[0119] Step a: Calculate the traversal reachability rate. Based on the location estimation result, establish a closed-form expression for the user uplink reachability rate. The calculation result of the closed-form expression for the user uplink reachability rate is as follows:

[0120] R k =log2(1+γ k )

[0121] in,

[0122]

[0123] Among them, for the kth user, R k It is UE k The closed-form solution of the uplink achievable rate λ k About UE k of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy The number of rows is M B The unit array, is the linear minimum mean square error estimate LMMSE case g k The estimated value of q k 、p k UE k The pilot normalized transmit power and data normalized transmit power, q i 、p i are the i-th user UEi The normalized pilot transmission power and data transmission power, p j is the jth user UE j The data normalized transmission power of , i,j∈{1,2,…,K}; χ k Yes k The values ​​of the main diagonal elements of the autocorrelation matrix and satisfy Δ k is to calculate g k The intermediate variable of the fourth-order moment satisfies c k About UE k The weighting coefficient of the linear estimation that minimizes the linear mean square error between the channel estimation value and the true value; Δ i is to calculate g i The intermediate variable of the fourth-order moment satisfies g i From UE i The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ki Yes k and g i The expected value of the inner product modulus squared and satisfies and Ω ki =Ω ik ,Ω ik Yes i and g k The expected value of the square of the inner product modulus, Ξ ij Yes i Vector inner product and g j The expected value of the product of vector inner products and satisfies And ij =Ξ ji ,Ξ ji Yes j Vector inner product and g i Expected value of vector inner product, g j From UE j The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ij Yes i and g j The expected value of the inner product modulus squared and satisfies and Ω ij =Ω ji ;Ω ji Yes j and g i The expected value of the square of the inner product modulus, χ i Yes i The values ​​of the main diagonal elements of the autocorrelation matrix and satisfy λ i About UE i of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy is the linear minimum mean square error estimate LMMSE case g i Estimated value of ik Yes i Vector inner product and g k The expected value of the product of vector inner products and satisfies And ik =Ξ ki ,Ξ ki Yes k Vector inner product and g i The expected value of the product of vector inner products, γ k is the UE under the maximum ratio combining MRC detector k The closed-form solution of the effective signal-to-interference-and-noise ratio SINR is

[0124] The expressions for the remaining variables are as follows:

[0125]

[0126] χ k =ρ B,k (ε0ε k |f k | 2 +(ε0+ε k +1)M R )

[0127] χ i =ρ B,k (ε0ε k |f i | 2 +(ε0+ε i +1)M R )

[0128]

[0129] in, express The steering vector, express The steering vector, It is UE i An estimate of the angle vector of ; is the path loss from RIS to BS, From UE k Path loss to RIS, From UE i Path loss to RIS; ρB,k is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE k The product of the large-scale fading of the NLOS link to the RIS, ρ B,i is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE i The product of the large-scale fading of the non-line-of-sight link NLOS to RIS; ε0 is the Ricean factor of the link from RIS to BS, ε k From UE k Ricean factor of the link to RIS, ε i From UE i The Rice factor of the link to the RIS; Represents θ R The steering vector, θ R is the angle of arrival AOA from BS to RIS in the RIS local coordinate system; f k is UE at time t k The total phase shift of the received signal, including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector f i is UE at time t i The total phase shift of the received signal, f i Including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector From UE k LOS channel to RIS link; Re{·} represents real part operation.

[0130] Step b: Establish an optimization problem. Based on the closed-form expression of the user's uplink achievable rate, construct a weighted sum rate maximization problem for optimizing the RIS phase. The mathematical model for maximizing the weighted sum rate for optimizing the RIS phase is:

[0131]

[0132] Among them, κ k is the weight of the kth user, R k It is UE k Closed-form expression of achievable rate. The rate of each user can be expressed as a function of the RIS phase. This is a non-convex optimization problem that requires finding the local optimal solution of the RIS phase, with the goal of maximizing the weighted sum of the rates of all users.

[0133] Step c: Solve the optimization problem. Use genetic algorithm to search for the local optimal solution of the above problem. Steps:

[0134] Step c1: Initialize the population

[0135] Determine the population size N t , each RIS phase array is regarded as an individual, and each individual in the population contains M R =M R,x ×M R,z chromosomes, the nth chromosome corresponds to the phase of the nth reflection unit in the RIS phase array Randomly generated in [0,2π);

[0136] Step c2: Adaptability evaluation and scaling

[0137] ① Evaluate the fitness of each individual in the current population, fitness evaluation function The objective function of the mathematical model for the weighted sum rate maximization problem;

[0138] ② Measure the original fitness value of the individual according to its ranking in the population, sort the original fitness of the individual, and calculate their scaled fitness as follows:

[0139]

[0140] Among them, rank i′ is the original fitness function of individual i′ after being sorted in descending order, is the proportional fitness of individual i′, N c is a parameter used in the crossover operation, To calculate the individual i' Intermediate variables used when

[0141] Step c3: Select

[0142] Select N from the current population e f i scaled The larger individuals serve as elite individuals, and these individuals produce the parents of the offspring;

[0143] Step C4: Cross

[0144] Select 2N based on random sampling c parents, used for crossover operation, to generate N c offspring;

[0145] Step c5: Mutation

[0146] Select the remaining N m =Nt -N e -N c individuals are used for mutation operation, and each chromosome of each individual is represented by p m The probability of mutation is a random number between [0,2π), where N m is the number of mutant individuals, p m is the probability of gene mutation;

[0147] Step c6: Fitness update

[0148] Calculate the fitness evaluation function for each offspring

[0149] Step c7: Iteration

[0150] Repeat steps c2-c6 until The absolute value of the difference between the previous and next iterations is less than ε, or the number of iterations of steps c2-c6 exceeds the maximum number of iterations N max , the iteration stops, and the optimized RIS phase matrix and weighted sum rate values ​​are output, where is the fitness evaluation function of all individuals is the average value of , and ε is the termination iteration error of the genetic algorithm.

[0151] Step 3: Channel estimation. Based on the optimized RIS phase, the user sends a pilot and data superposition signal. After receiving the superposition signal, the base station despreads the data received at the BS end, performs channel estimation based on the despread received signal, and obtains the channel estimation result. The steps are as follows:

[0152] Step a: Multiply the received signal Y at the BS by produce:

[0153]

[0154] in, y k is the non-Gaussian signal after despreading of the kth user, y k is a non-Gaussian signal; is the BS received signal within a channel coherence interval, Indicates size is M B ×τ C The complex matrix of ; It is UE k The pilot signal sent, It is UE k′ Pilot signal sent, UE k′ is the k′th user and satisfies and Represents the size as τC ×1 complex column vector; is an additive white Gaussian noise (AWGN) matrix and the noise value obeys represents a complex Gaussian distribution with a mean of 0 and a variance of 1; v k It is UE k The sending data and v i It is UE i The sending data and Indicates that the mean is 0 and the covariance matrix is The complex Gaussian distribution of The number of rows is τ C The unit array.

[0155] The expression of Y is:

[0156]

[0157] Step b: Use the LMMSE estimator to obtain the corresponding channel estimate:

[0158]

[0159] in,

[0160]

[0161] Step 4: Data detection. Based on the channel estimation results, the base station uses an MRC detector to perform data detection and obtains data detection results for all users. k , and the data detection results are It includes the following sub-steps:

[0162] Step a: Channel estimation matrix The receiving matrix of the MRC detector

[0163] Step b: Use the receiving matrix to process the received signal and obtain in represents the estimated value of all user datasets;

[0164] Step c: Restore the data of the kth user to

[0165]

[0166] in represents the data detection result of the kth user, It is UE i a pilot signal sent;

[0167] Step d: Using the use-and-forget (UatF) technique, the recovered data is rewritten as

[0168]

[0169] in,

[0170]

[0171] in, is used to represent the k-th user’s z k The intermediate variable represents the expected signal of the kth user, represents the uncertainty of the beamforming gain of the kth user, z k represents the interference and noise of the kth user;

[0172] Step e: Recover the data of the kth user after rewriting to obtain the effective uplink spectrum efficiency of the kth user in the superimposed pilot mode

[0173]

[0174] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for location perception and superimposed pilot transmission of an Internet of Things device, characterized in that: include: Construct a wireless frame structure including position sensing and superimposed pilot transmission; Based on the wireless frame structure, the user estimates the position according to the base station pilot signal reflected by the RIS and obtains the position estimation result; Based on the position estimation result, a closed-form expression of the user's uplink achievable rate is established; According to the closed-form expression of the user's uplink achievable rate, a weighted sum rate maximization problem for optimizing the RIS phase is constructed, and a genetic algorithm is used to solve the local optimal solution of the problem to obtain the optimized RIS phase. Based on the optimized RIS phase, the user sends a pilot and data superposition signal. After receiving the superposition signal, the base station performs channel estimation and obtains the channel estimation result. Based on the channel estimation result, the base station performs data detection to obtain a data detection result.

2. A method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 1, characterized in that: The wireless frame structure includes multiple position coherent intervals, and the frame length of one position coherent interval is τ L =τ P +N C τ C , where τ P is the time domain sequence length occupied by user positioning, τ C is the time domain sequence length of a channel coherence interval, τ L Includes 1 frame length of τ P The location awareness interval, and N C The frame length is τ C The channel coherence interval.

3. A method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 2, characterized in that: Based on the radio frame structure, in the location sensing interval, the base station BS transmits a pure pilot sequence to the kth user UE. k , where k represents the subscript index value of the user UE, k∈{1,2,…,K}, and K is the number of UEs; then, UE k Use 2D positioning method to estimate UE k The angle information of the user itself is used to obtain the angle vector of the kth user. Estimated value of Angle vector θ k Refers to the distance from RIS to UE in the RIS local coordinate system. k The departure angle AOD includes the UE k Azimuth and the pitch angle of the kth user They are The estimated value of .

4. A method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 3, characterized in that: Based on the location estimation result, a closed-form expression for the user uplink achievable rate is established as follows: R k =log2(1+γ k ) in, Among them, for the kth user, R k It is UE k The closed-form solution of the uplink achievable rate, M B is the number of BS antennas; k About UE k of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy represents the mathematical expectation operation of the matrix, and the superscript H represents the conjugate transpose operation of the matrix. The number of rows is M B The unit array, is the linear minimum mean square error estimate LMMSE case g k The estimated value of k From UE k The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS; q k 、p k UE k The normalized pilot transmission power and data transmission power, q i 、p i are the i-th user UE i The normalized pilot transmission power and data transmission power, p j is the jth user UE j The data normalized transmission power of , i,j∈{1,2,…,K}; χ k Yes k The values ​​of the main diagonal elements of the autocorrelation matrix satisfy Δ k is to calculate g k The intermediate variable of the fourth-order moment satisfies ‖·‖ represents the 2-norm operation of a vector; c k About UE k The weighting coefficient of the linear estimation that minimizes the linear mean square error between the channel estimation value and the true value; Δ i is to calculate g i The intermediate variable of the fourth-order moment satisfies g i From UE i The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ki Yes k and g i The expected value of the inner product modulus squared and satisfies and Ω ki =Ω ik , |·| represents the modulus operation of complex numbers, Ω ik Yes i and g k The expected value of the square of the inner product modulus, Ξ ij Yes i Vector inner product and g j The expected value of the product of vector inner products and satisfies And ij =Ξ ji ,Ξ ji Yes j Vector inner product and g i Expected value of vector inner product, g j From UE j The uplink concatenated channel vector consists of the channel to RIS and the channel from RIS to BS, Ω ij Yes i and g j The expected value of the inner product modulus squared and satisfies and Ω ij =Ω ji ;Ω ji Yes j and g i The expected value of the square of the inner product modulus, χ i Yes i The values ​​of the main diagonal elements of the autocorrelation matrix satisfy λ i About UE i of The values ​​of the main diagonal elements of the autocorrelation matrix of the vector satisfy is the linear minimum mean square error estimate LMMSE case g i Estimated value of ik Yes i Vector inner product and g k The expected value of the product of vector inner products and satisfies And ik =Ξ ki ,Ξ ki Yes k Vector inner product and g i The expected value of the product of vector inner products, γ k is the UE under the maximum ratio combining MRC detector k The closed-form solution of the effective signal-to-interference-and-noise ratio SINR; In the uplink channel, for the kth user UE k , g k The expression is as follows: in, With the RIS phase matrix Λ and About k The LOS channel component of the line-of-sight link is constant in the same position coherence interval, but varies in different position coherence intervals. Indicates size is M B ×1 complex column vector; is the channel matrix of the link from RIS to BS, is the LOS channel component of H0, Indicates size is M R ×M B The complex matrix, M R =M R,x ×M R,z is the number of RIS elements, M R,x and M R,z are the lengths of the rows and columns of Λ respectively; and Where m x ∈{1,2,...,M R,x },m z ∈ {1,2,...,M R,z }, Where e is the base of natural logarithms, J is the imaginary unit, is the phase of the nth reflection element in the RIS phase matrix, m x and m z are the indices of the row dimension and column dimension of Λ respectively, and n satisfies n∈{1,2,...,M R } and n=(m x -1)M R,z +m z , vec(Λ) is the vectorization of Λ, diag(vec(Λ)) is the diagonalization of vector vec(Λ), Indicates size is M R,x ×M R,z The complex matrix of ; From UE k The channel matrix of the link to RIS, Yes k The LOS channel component, Indicates size is M R ×1 complex column vector, ψ k From UE k LOS link amplitude gain to RIS, express The steering vector of The expressions for the remaining variables are as follows: x k =ρ B,k (e0e k |f k | 2 +(ε0+ε k +1)M R ) x i =ρ B,k (e0e k |f i | 2 +(ε0+ε i +1)M R ) in, express The steering vector, express The steering vector, It is UE i An estimate of the angle vector of ; is the path loss from RIS to BS, From UE k Path loss to RIS, From UE i Path loss to RIS; ρ B,k is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE k The product of the large-scale fading of the NLOS link to the RIS, ρ B,i is the large-scale fading of the non-line-of-sight link NLOS from RIS to BS and from UE i The product of the large-scale fading of the non-line-of-sight link NLOS to RIS; ε0 is the Ricean factor of the link from RIS to BS, ε k From UE k Ricean factor of the link to RIS, ε i From UE i The Rice factor of the link to the RIS; Represents θ R The steering vector, θ R is the angle of arrival AOA from BS to RIS in the RIS local coordinate system; f k is UE at time t k The total phase shift of the received signal, including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector f i is UE at time t i The total phase shift of the received signal, f i Including the first part a(θ R ) produced by the phase shift, the second part of the phase modulation of the RIS phase matrix, and the third part The steering vector From UE k LOS channel to RIS link; Re{·} represents real part operation.

5. A method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 4, characterized in that: According to the closed-form expression of the user's uplink achievable rate, the weighted sum rate maximization problem of optimizing the RIS phase is constructed. The mathematical model of the weighted sum rate maximization problem of optimizing the RIS phase is: Among them, κ k is the weight of the kth user, R k It is UE k The closed-form expression for the achievable rate is, Represents a set of UE subscript indices.

6. A method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 4, characterized in that: According to the closed-form expression of the user's uplink achievable rate, a weighted sum rate maximization problem for optimizing the RIS phase is constructed, and a genetic algorithm is used to solve the local optimal solution of the problem to obtain the optimized RIS phase. The steps of the genetic algorithm are as follows: Step 61, initialize the population: Determine the population size N t , each RIS phase array is regarded as an individual, and each individual in the population contains M R =M R,x ×M R,z chromosomes, the nth chromosome corresponds to the phase of the nth reflection unit in the RIS phase matrix Randomly generated in [0,2π); Step 62: Adaptability evaluation and scaling: ① Evaluate the fitness of each individual in the current population, fitness evaluation function The objective function of the mathematical model for the weighted sum rate maximization problem; ② Measure the original fitness value of the individual according to its ranking in the population, sort the original fitness of the individual, and calculate their scaled fitness as follows: Among them, rank i′ is the original fitness function of individual i′ after being sorted in descending order, is the proportional fitness of individual i′, N c is a parameter used in the crossover operation, To calculate individual i ′ of Intermediate variables used when Step 63. Select: Select N from the current population e indivual Individuals with the largest values ​​serve as elite individuals, and these individuals produce the parents of their offspring; Step 64, cross: Select 2N based on random sampling c parents, used for crossover operation, to generate N c offspring; Step 65, mutation: Select the remaining N m =N t -N e -N c individuals are used for mutation operation, and each chromosome of each individual is represented by p m The probability of mutation is a random number between [0,2π), where N m is the number of mutant individuals, p m is the probability of gene mutation; Step 66, fitness update: Calculate the fitness evaluation function for each offspring Step 67, Iteration: Repeat steps 62-66 until The absolute value of the difference between the previous and next iterations is less than ε, or the number of iterations of steps 62-66 exceeds the maximum number of iterations N max , the iteration stops, and the optimized RIS phase matrix and weighted sum rate values ​​are output, where is the fitness evaluation function of all individuals is the average value of , and ε is the termination iteration error of the genetic algorithm.

7. The method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 1, characterized in that: Based on the optimized RIS phase, the user sends a pilot and data superposition signal. After receiving the superposition signal, the base station performs channel estimation and obtains the channel estimation result; the details are as follows: in, is the BS received signal within a channel coherence interval, Indicates size is M B ×τ C The complex matrix of y k is the despread non-Gaussian signal of the jth user, It is UE k The pilot signal sent, It is UE k′ Pilot signal sent, UE k′ For the kth ′ Users who meet τ C ≥K and k≠k ′ , Represents the size as τ C ×1 complex column vector; is an additive white Gaussian noise (AWGN) matrix and the noise value obeys represents a complex Gaussian distribution with a mean of 0 and a variance of 1; v k It is UE k The sending data and v i It is UE i The sending data and Indicates that the mean is 0 and the covariance matrix is The complex Gaussian distribution of The number of rows is τ C The unit array.

8. The method for location perception and superimposed pilot transmission of an Internet of Things device according to claim 4, characterized in that: Based on the channel estimation results, the base station uses the maximum ratio combining detector to perform data detection and obtains the data detection results of all users. The data detection result of the kth user is The details are as follows: in, It is UE i The pilot signal sent.

Images