A Phase-Shift Control Method for Intelligent Reflective Surface-Assisted Vehicle-to-Everything (V2X) Sensor Integration

Abstract

This invention provides a phase-shift control method for intelligent reflector-assisted vehicle-to-everything (V2X) communication and sensing integration, comprising: Step 1, the base station performs channel estimation based on information fed back from the vehicle, and obtains the channel matrix and vector information; Step 2, a communication model is established, and the beamforming matrix of the base station and the phase-shift matrix of the intelligent reflector are initialized; Step 3, a sensing model is established; Step 4, the sensing performance matrix is ​​derived; Step 5, the joint beam optimization objective and various constraints are established. This invention uses an onboard RIS (Radio Resonance System) to handle user-end beamforming, while the base station only needs to transmit a wide-beam signal, thus significantly reducing the burden on the base station. By adjusting the sensing accuracy constraints, the performance of communication and sensing services can be flexibly balanced under different accuracy scenarios.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and in particular relates to a phase shift control method for intelligent reflective surface-assisted vehicle networking with integrated sensing. Background Technology

[0002] Vehicle-to-everything (V2X) technology refers to an intelligent system that enables interconnection between vehicles and everything else through advanced wireless communication, sensing, and networking technologies. In existing V2X systems, communication and sensing functions are typically separate, leading to low spectrum efficiency, high hardware costs, and difficulty in achieving wide-area collaboration. Especially in complex urban areas and high-speed moving scenarios, vehicles face severe non-line-of-sight congestion and channel fading. Traditional roadside unit deployment solutions are costly, while communication signal-based assisted sensing technologies lack sufficient accuracy.

[0003] Because reconfigurable intelligent surfaces (RIS) lack signal processing capabilities, obtaining channel state information related to RIS is difficult and costly, making beam tracking and beam alignment challenging. Utilizing target location sensed by integrated sensing and communications ISAC technology enables fast and efficient beam scanning, pairing, and alignment processes, particularly suitable for highly mobile scenarios. Furthermore, angle and timing information can be further used for communication channel estimation. Summary of the Invention

[0004] Purpose of the Invention: The technical problem this invention aims to solve is to address the shortcomings of existing technologies by providing a phase-shift control method for integrated communication and sensing in vehicle-to-everything (V2X) networks, specifically addressing the issues of low spectral efficiency and high hardware costs caused by the separation of communication and sensing functions in V2X. Innovations include: achieving spectrum and hardware sharing between communication and sensing through an intelligent reflector; establishing an integrated communication and sensing model; utilizing sensing information (such as target location) to assist in rapid beam alignment; designing a joint optimization framework to maximize user and communication rates while satisfying sensing performance constraints (measured by the Cramer-Rao lower bound), transmit power limits, and RIS phase mode constraints; and employing an alternating optimization algorithm to iteratively update the precoding matrix and phase-shift parameters, ultimately achieving communication-sensing synergistic gain.

[0005] The method includes the following steps:

[0006] Step 1: The base station performs channel estimation based on the information fed back by the vehicle, and obtains the channel matrix and vector information;

[0007] Step 2: Establish the communication model and initialize the beamforming matrix and intelligent reflector phase shift matrix of the base station;

[0008] Step 3: Establish a perception model;

[0009] Step 4: Derive the perception performance matrix;

[0010] Step 5: Establish the joint beam optimization objective and various constraints.

[0011] Step 1 includes: the channel matrix includes: direct channels between the integrated inductive base station and the roof reflector. ,in This indicates that the matrix elements are complex numbers; This represents the total number of RIS units in the communication system. M represents the total number of RIS units in the radar system; M represents the number of antennas at the base station.

[0012] Channel between the roof reflector and the k-th in-vehicle user , ; base station and the Non-line-of-sight channel between users inside the vehicle K represents the total number of users inside the vehicle.

[0013] Step 2 includes: using and These refer to the beamforming matrix for communication and the beamforming matrix for radar systems, respectively.

[0014] Set the transmission phase shift matrix of the smart reflector as follows: ;in Represents a diagonal matrix function; A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal One element;

[0015] Set the reflection phase shift matrix of the smart reflector as follows: , A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal One element;

[0016] Used to represent the global phase shift matrix of RIS This is the global phase shift vector of RIS;

[0017] In the Signals transmitted within each time slot Set to:

[0018] ,

[0019] in, For joint precoding matrix; transmitted signal , , This represents the total number of time slots.

[0020] For communication signals of K users, and covariance matrix identity matrix : ;

[0021] A dedicated sensing signal for achieving full degrees of freedom in target sensing across M channels, and The covariance matrix is ​​the identity matrix. ,Right now ;

[0022] Furthermore, the communication signals and the sensing signals are independent of each other. .

[0023] Step 3 includes: transmitting the sensing signal from the base station back to the base station receiving antenna via a reflecting element. The model is as follows:

[0024] ,

[0025] in The radar cross section is T, which represents the transpose. Let be the target response matrices of the base station and the target, and let the noise vector follow the order of the base station and the target response matrices. The mean is zero vector and the covariance matrix is Complex Gaussian distribution, Let Variance be the variance of the noise vector. It is an M-dimensional identity matrix;

[0026] This is the channel between the reflecting unit and the base station. For the transmitting end steering vector, The starting angle for the signal to travel from the base station to the target. For the receiver steering vector, These represent the azimuth and elevation angles at which the signal arrives at the reflecting unit, respectively.

[0027] Combined echo signals Each sample, the base station will receive the signal Described as:

[0028] ,

[0029] in For the equivalent echo channel matrix, the signal matrix ,in For the signal in time slot L; noise matrix ,in The noise in time slot L.

[0030] Step 4 includes:

[0031] Step 4-1, vectorize the signal received by the base station:

[0032] ,

[0033] in For the signal section, For the noise part, vec indicates vectorization. Represent the vectorized signal; set the unknown parameters to be estimated. ; This refers to the two-dimensional angle information of the vehicle relative to the base station, where Let be a two-dimensional real vector composed of the real and imaginary parts of the complex gain of the target reflection. and These correspond to the real and imaginary parts of the complex gain of reflection, respectively;

[0034] Considering noise compliance If the complex Gaussian distribution is followed, then ;estimate The (i,j)th element of the Fisher Information Matrix (FIM) The expression is as follows: , Corresponding to The i-th component of the vector, Corresponding to The j-th component of the vector; Re represents a partial differential; Re{⋅} represents taking the real part of a complex number;

[0035] Step 4-2, extract angle information from the echo signal. Calculate partial derivatives to establish the Fisher Information Matrix (FIM); about The partial derivatives are shown below:

[0036] ,

[0037] ,

[0038] ,

[0039] in and These represent the echo channels. right Find the partial derivative, with respect to Find the partial derivative;

[0040] Obtain the required elements of the Fisher Information Matrix (FIM) ;

[0041] The elements of the matrix are represented as follows:

[0042] Fisher information matrix of angle parameters ,

[0043] The mutual information vector between the angle parameter and the complex gain ,

[0044] transpose ,

[0045] Among them, the mutual information between azimuth angle and complex gain ,

[0046] Mutual information between pitch angle and complex gain ,

[0047] Fisher information for the complex gain itself;

[0048] For submatrix The expression is as follows:

[0049] Self-information of azimuth ,

[0050] Mutual information between azimuth and elevation angles ,

[0051] Mutual information between pitch and azimuth ,

[0052] Self-information of pitch angle ;

[0053] Where Tr represents the trace of the matrix; the superscript H represents the conjugate transpose;

[0054] Step 4-3, Separation Angle Cramer-Rao Lower Bound (CRB): Invert the Fisher Information Matrix (FIM) to obtain the Cramer-Rao Lower Bound (CRB) matrix. By separating the two angles of Cramer-Lower Boundary (CRB), we obtain... ,in and These correspond to the lower bounds of Cramer-Rao for azimuth and elevation, respectively.

[0055] Step 5 includes the following steps:

[0056] Step 5-1: The base station initializes the algorithm and sets the initial number of iterations. With maximum iteration Set precision ;

[0057] Step 5-2, Initialize the base station precoding matrix The phase shift matrix of the intelligent reflective surface And towards quantification: The precoded vector is obtained by expanding it by rows and columns. , Then, extracting the elements of the main pair of lines yields the phase shift vector. ; Keep parameters updated during iteration: Let , Save the current parameters as the state of the previous round;

[0058] Step 5-3: Based on the known channel state information, establish and solve the optimization problem (Ⅰ), and extract and transform the coupling parameters. For non-convex problems and non-convex constraints, transform them to calculate the user's sum and communication rate. Consider the problem of maximizing the sum and communication rate, while satisfying: the sensing system should meet the minimum sensing performance requirements, the base station transmitter power should not exceed the set maximum power constraint, and the phase mode constraint of the intelligent reflector.

[0059] The optimization problem (I) is expressed as:

[0060] ,

[0061] in, Indicates the base station and the first Non-line-of-sight channel between users inside the vehicle transpose; This represents the channel between the roof reflector and the k-th user inside the vehicle. transpose; This is the direct channel between the base station and the roof reflector; This is the pre-encoding vector corresponding to user k; For other users' precoded vectors; For noise variance;

[0062] Constraint 1:

[0063] ,

[0064] Constraint 2:

[0065] ,

[0066] Constraint 3:

[0067] ,

[0068] in The CRB threshold; Represents the maximum transmission power; Reflection phase shift matrix The nth diagonal element;

[0069] Step 5-4, Solve the optimization problem (III) and update the auxiliary variables. and Solve optimization problem (V) to update the sending precode w, and solve optimization problem (VI) to obtain the transmission phase shift matrix of the smart reflector. The phase shift matrix of the intelligent reflector is obtained by aligning the phase of the wave with the incoming wave from the base station. Calculate user and communication rates At the same time, the following conditions must be met: the sensing system should meet the minimum sensing performance requirements, the power of the base station transmitter should not exceed the set maximum power constraint, and the phase mode constraint of the intelligent reflector should be met.

[0070] Step 5-5: Adjust the penalty parameters until all variables converge; if convergence fails, utilize the obtained intermediate variables, i.e., the precoding vector. and phase shift vector Repeat steps 5-3 to 5-4 to update and obtain new convergence variables, including the converged precoding vector. and Then let t = t + 1, and return to step 5-2;

[0071] Steps 5-6: Recombine the precoding vector and the smart reflector phase shift vector into a matrix, and output the optimal transmission precoding matrix obtained through iteration. Optimal phase shift control matrix of the intelligent reflector .

[0072] Step 5-3 includes the following steps:

[0073] Step 5-3-1: Perform a fractional transformation on the optimization problem. After performing a Lagrange dual transformation on the objective function to remove the outer logarithmic function, the objective function is transformed using fractional programming as follows to obtain a new objective function:

[0074] ,

[0075] in This is an auxiliary variable introduced for the ratio term; it is used to expand the third term of the objective function and introduce the auxiliary variable. Then, the objective function is transformed into:

[0076] ,

[0077] in This is a simplified form of the new objective function;

[0078] Transform into about and The quadratic polynomial:

[0079] ,

[0080] in , Used to refer to the objective function that does not contain and Other items, while defining , The remaining items are described below:

[0081] ,

[0082] for The coefficient vector of the first-order terms, and These are two auxiliary variables corresponding to user k;

[0083] ,

[0084] for The quadratic coefficient matrix;

[0085] ,

[0086] This is the weighted equivalent channel vector for the k-th user and the i-th beam direction, corresponding to the noise interference term; The selection matrix represents the extraction of the i-th beam from the total precoding vector;

[0087] ,

[0088] for The vector of coefficients of the first-order terms;

[0089] ,

[0090] for The quadratic coefficient matrix;

[0091] Step 5-3-2: Process constraint 1 by introducing auxiliary variables. Constraint 1 is transformed into the following two constraints:

[0092] Constraint 1-1:

[0093] ,

[0094] in For matrix The reversed trace;

[0095] Constraint 1-2:

[0096] ,

[0097] Using Schur complement conditions, for Constraint 1-2 is equivalent to:

[0098] ;

[0099] Step 5-3-3: Introduce auxiliary variables Will and Extracting the semidefinite constraints and using the augmented Lagrangian method, the optimization problem (I) is transformed into the optimization problem (II):

[0100] ,

[0101] Constraint 1:

[0102] ,

[0103] Constraint 2:

[0104] ,

[0105] Constraint 3:

[0106] ,

[0107] Constraint 4:

[0108] ,

[0109] Constraint 5:

[0110] ,

[0111] in,

[0112] ,

[0113] ,

[0114] ,

[0115] ,

[0116] ,

[0117] ,

[0118] in As dual variables, It is a second-order identity matrix. As a penalty factor, Let i represent the i-th auxiliary variable. Let i represent the i-th dual variable. Used in referential expressions where it is not explicitly included For the other terms, i takes values ​​from 1 to 6.

[0119] Step 5-4 includes the following steps:

[0120] Step 5-4-1, Update the auxiliary variable penalty term With Lagrange multipliers: auxiliary variables for user k and The update is achieved through calculation and It is concluded that; at this time , ;

[0121] Solve The optimization problem for f is as follows:

[0122] Optimization Problem (III):

[0123] ,

[0124] Constraint 1:

[0125] ,

[0126] Constraint 2:

[0127] ;

[0128] Step 5-4-2: Fix the phase shift matrix of the intelligent reflector to solve the transmission precoding; at this time, solve... The optimization problem is as follows:

[0129] ,

[0130] Optimization Problem (Ⅳ):

[0131] ,

[0132] Constraint 1:

[0133] ,

[0134] First, define to refer to and The irrelevant terms. Next, the penalty function is expanded to obtain formula (1):

[0135] (1),

[0136] The second term of formula (1) is processed to define the Hermitian matrix.

[0137] Construct a negative semidefinite matrix ,in for The largest eigenvalue. For Perform a Taylor expansion at the t-th iteration: .

[0138] Organized ,in , ;

[0139] Process the third term of formula (1) and define... Construct a negative semidefinite matrix ,in for The largest eigenvalue;

[0140] Considering power constraints and define irrelevant items for:

[0141] ,

[0142] in, ;

[0143] at this time Transformed into:

[0144] ,

[0145] Applying a first-order Taylor expansion again to remove the fourth-order terms in the formula, we get:

[0146] ,

[0147] The final solution The problem becomes a convex problem as follows:

[0148] Optimization problem (V):

[0149] ,

[0150] Constraint 1:

[0151] ;

[0152] Step 5-4-3, for The optimization problem is as follows:

[0153] Optimization problem (VI):

[0154] ,

[0155] Constraint 1:

[0156] .

[0157] The present invention also provides an electronic device, including a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method.

[0158] The present invention also provides a storage medium storing a computer program or instructions that, when the computer program or instructions are run on a computer, execute the steps of the method described.

[0159] The present invention has the following beneficial effects:

[0160] 1) In highly mobile vehicle-to-everything (V2X) scenarios dominated by line-of-sight, using vehicle status information obtained from a sensor-integrated system to assist beam alignment will significantly reduce pilot overhead and processing latency.

[0161] 2) This invention uses an on-board RIS to handle beamforming at the user end, while the base station only needs to transmit a wide beam signal, thereby significantly reducing the burden on the base station.

[0162] 3) By adjusting the perception accuracy constraints, the performance of communication and perception services can be flexibly balanced in different accuracy scenarios.

[0163] 4) It fully considers the power constraints under the existing network architecture and has good compatibility with actual deployment. Attached Figure Description

[0164] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0165] Figure 1 This is a schematic diagram of the model scene.

[0166] Figure 2 This is a flowchart of the method of the present invention.

[0167] Figure 3 This is a graph comparing algorithm performance.

[0168] Figure 4 This is a graph showing the relationship between communication rate and transmit power under different CRB values. Detailed Implementation

[0169] like Figure 1 As shown, this embodiment of the invention provides a phase-shift control method for intelligent reflective surface-assisted vehicle-to-everything (V2X) sensing integration, comprising the following steps:

[0170] Step 1: The base station performs channel estimation based on the information fed back by the vehicle, and obtains the channel matrix and vector information.

[0171] The channel matrix includes: a direct channel between the integrated inductive base station and the roof reflector. ,in This indicates that the matrix elements are complex numbers; N is the total number of RIS (Reconfigurable Intelligent Surface Units). This represents the total number of RIS units in the communication system. M represents the total number of RIS units in the radar system; M represents the number of antennas at the base station.

[0172] Channel between the roof reflector and the k-th in-vehicle user , ; base station and the Non-line-of-sight channel between users inside the vehicle K represents the total number of users inside the vehicle.

[0173] Step 2: Establish the communication model and initialize the beamforming matrix and intelligent reflector phase shift matrix of the base station.

[0174] use and These refer to the beamforming matrix of communication systems and the beamforming matrix of radar systems, respectively.

[0175] Set the transmission phase shift matrix of the smart reflector as follows: ;in Represents a diagonal matrix function; A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal Each element.

[0176] Set the reflection phase shift matrix of the smart reflector as follows: , A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal Each element.

[0177] In the Signals transmitted within each time slot Set to:

[0178] ,

[0179] in, For joint precoding matrix; transmitted signal , , This represents the total number of time slots.

[0180] For communication signals of K users, and The covariance matrix is ​​the identity matrix. , ;

[0181] A dedicated sensing signal for achieving full degrees of freedom in target sensing across M channels, and The covariance matrix is ​​the identity matrix. , .

[0182] Furthermore, the communication signals and the sensing signals are independent of each other. .

[0183] Step 3: Establish a sensing model. The sensing signal transmitted from the base station is sent back to the base station receiving antenna via a reflective element. The model is as follows:

[0184] ,

[0185] in The radar cross section represents the complex gain of the reflection, and T denotes the transpose. This is the target response matrix between the base station and the target (the reflective element part of the reflective surface). It is a complex Gaussian noise vector. Let Variance be the variance of the noise vector. It is an M-dimensional identity matrix;

[0186] This is the channel between the reflecting unit and the base station. For the transmitting end steering vector, , For the receiver steering vector, These represent the azimuth and elevation angles of the target relative to the base station, respectively. The combined echo signal... Each sample, the base station will receive the signal Described as:

[0187] ,

[0188] in For simplified formula, the signal matrix is: ,in For the signal in time slot L; noise matrix ,in The noise in time slot L.

[0189] Step 4, Derive the perception performance matrix. Consider... The target parameters for sensing are used, and the Cramer-Rao lower bound is adopted as the sensing performance matrix. The signal received by the base station... Vectorization: ,in For the signal section, For the noise component, the Fisher Information Matrix (FIM) is obtained by taking the partial derivative of the two-dimensional angle information. After inverting the FIM, its main diagonal elements are extracted as an evaluation of the perception performance.

[0190] Step 5: Establish the joint beamforming objective and constraints, and perform joint optimization. Utilize the obtained communication beamforming matrix... and radar beamforming matrix Intelligent reflective surface transmission phase shift Reflection phase shift Given the known channels, the user's sum and communication rate can be calculated. Considering the problem of maximizing the sum and communication rate, while simultaneously satisfying: 1) the sensing system should meet the minimum sensing performance requirements; 2) the base station transmitter power should not exceed the set maximum power constraint; and 3) the phase mode constraint of the intelligent reflector.

[0191] Step 6: The base station sends the optimized phase shift control matrix information of the intelligent reflector to the intelligent reflector for phase shift control; simultaneously, the base station sends the optimal precoding matrix information to the transmitter for precoding matrix control. This establishes a communication link between the base station, the RIS transmission unit, and the user, as well as a sensing link between the base station, the RIS reflection unit, and the base station.

[0192] Step 4 includes the following steps:

[0193] Step 4-1: Derive the Cramér-Rao Bound (CRB) for two-dimensional angle information from the received echo signal. This requires vectorizing the signal received by the base station. ,in , Set the unknown parameters to be estimated. , This refers to the two-dimensional angle information of the vehicle relative to the base station. , and These correspond to the real and imaginary parts of the complex reflection gain vector, respectively;

[0194] Considering noise compliance If the complex Gaussian distribution is followed, then .estimate The (i,j)th element of the Fisher Information Matrix (FIM) is expressed as: , Corresponding to The i-th component of the vector, Corresponding to The j-th component of the vector.

[0195] Step 4-2, extract angle information from the echo signal. Calculate partial derivatives to establish the Fisher Information Matrix (FIM); about The partial derivatives are shown below:

[0196] ,

[0197] ,

[0198] ,

[0199] in and These represent the echo channels. right Find the partial derivative, with respect to Find the partial derivative;

[0200] Substituting the above three equations into the Fisher Information Matrix (FIM), we obtain the required elements of the Fisher Information Matrix (FIM). .

[0201] The elements of the matrix are represented as follows:

[0202] ,

[0203] ,

[0204] ,

[0205] ,

[0206]

[0207] Specifically, for submatrices The expression is as follows:

[0208] ,

[0209] ,

[0210] ,

[0211] ;

[0212] Step 4-3, Separation Angle Cramer-Rao Lower Bound (CRB): Invert the Fisher Information Matrix (FIM) to obtain the Cramer-Rao Lower Bound (CRB) matrix. By separating the two angles of Cramer-Lower Boundary (CRB), we obtain... .in and These correspond to the lower bounds of Cramer-Rao for azimuth and elevation, respectively.

[0213] Step 5 includes the following steps:

[0214] Step 5-1: The base station initializes the algorithm and sets the initial number of iterations. With maximum iteration Set precision .

[0215] Step 5-2, Initialize the base station precoding matrix The phase shift matrix of the intelligent reflective surface And vectorize it ( Expand by rows and columns, Then extract its main opposite element). During the iteration process, keep the parameters updated, that is, let... , Save the current parameters as the state of the previous round.

[0216] Step 5-3: Based on the known channel state information, establish and solve optimization problem (Ⅰ), and extract and transform the coupling parameters, transforming non-convex problems and non-convex constraints. Calculate the user's sum and communication rate. Consider the problem of maximizing the sum and communication rate, while satisfying: 1) the sensing system should meet the minimum sensing performance requirements; 2) the base station transmitter power should not exceed the set maximum power constraint; 3) the phase mode constraint of the intelligent reflector.

[0217] Optimization Problem (I):

[0218] ,

[0219] Constraint 1:

[0220] ,

[0221] Constraint 2:

[0222] ,

[0223] Constraint 3:

[0224] ,

[0225] Constraint 1 is a CRB constraint. The first constraint is the CRB threshold; constraint 2 is a power constraint. Represents the maximum transmit power; constraint 3 is a RIS modulo-1 constraint. Reflection phase shift matrix The nth diagonal element; this constraint relates to the physical properties of RIS.

[0226] Step 5-4, Solve the optimization problem (III) and update the auxiliary variables. and Solve optimization problem (V) to update the sending precode w, and solve optimization problem (VI) to obtain the transmission phase shift matrix of the smart reflector. After obtaining the above parameters, calculate the user's and communication rates. At the same time, it should meet the following requirements: 1) the sensing system should meet the minimum sensing performance requirements; 2) the power of the base station transmitter should not exceed the set maximum power constraint; and 3) the phase mode constraint of the intelligent reflector.

[0227] Step 5-5: Adjust the penalty parameters until all variables converge; if convergence fails, utilize the obtained intermediate variables. and Repeat steps 5-3 to 5-4 to update and obtain new variables. and Then let t = t + 1, and return to step 5-2;

[0228] Steps 5-6: Recombine the precoding vector and the smart reflector phase shift vector into a matrix, and output the optimal transmission precoding matrix obtained through iteration. Optimal phase shift control matrix of the intelligent reflector .

[0229] Step 5-3 includes the following steps:

[0230] Step 5-3-1: Perform a fractional transformation on the optimization problem. After removing the outer logarithmic function through a Lagrange dual transformation of the objective function, the objective function can be transformed using fractional programming to obtain a new, more manageable objective function:

[0231] ,

[0232] in This is an auxiliary variable introduced for the ratio term. Expand the third term of the objective function and introduce the auxiliary variable. Then, the objective function is transformed into:

[0233] ,

[0234] The above formula can be transformed into the following expressions: and The quadratic polynomial:

[0235] ,

[0236] in , Used to refer to the objective function that does not contain and Other items, while defining , The remaining items are described below:

[0237] ,

[0238] ,

[0239] ,

[0240] ,

[0241] ,

[0242] By processing the objective function, it is made into a convex function with respect to the variables, which facilitates the subsequent iterative solution.

[0243] Step 5-3-2 addresses constraint 1 by optimizing the complex CRB matrix. Considering the positive semi-definiteness of the CRB matrix, the trace of its inverse is monotonically decreasing on a positive definite cone. An auxiliary variable is then introduced. This ensures the positive semidefiniteness of the CRB matrix. Thus, constraint 1 is transformed into the following two constraints:

[0244] Constraint 1-1:

[0245] ,

[0246] in For matrix The traces of reversal.

[0247] Constraint 1-2:

[0248] ,

[0249] Using Schur complement conditions, for The above constraints 1-2 are equivalent to:

[0250] ;

[0251] Step 5-3-3 introduces an auxiliary variable f to handle coupling variables. Solving the optimization problem at this point remains quite complex because the precoding matrix needs to be optimized. With the transmission phase shift matrix They are coupled, and the perceived performance constraints are also related to... and reflection phase shift matrix This is relevant. At this point, an auxiliary variable is introduced. Will and Extracting the semidefinite constraints and using the augmented Lagrangian method, the optimization problem (I) is transformed into:

[0252] Optimization Problem (II):

[0253] ,

[0254] Constraint 1:

[0255] ,

[0256] Constraint 2:

[0257] ,

[0258] Constraint 3:

[0259] ,

[0260] Constraint 4:

[0261] ,

[0262] Constraint 5:

[0263] ,

[0264] in,

[0265] ,

[0266] ,

[0267] ,

[0268] ,

[0269] ,

[0270] ,

[0271] in As dual variables, This serves as the penalty factor. At this point, the variables in the problem and constraints are separated, allowing for the next step of alternating optimization.

[0272] Step 5-4 includes the following steps:

[0273] Step 5-4-1: Update the auxiliary variable penalty term and Lagrange multipliers. While fixing other variables, [the following steps are taken]. and The update can be calculated and The conclusion is reached. At this point... , Similarly, when other variables are fixed, it is used to solve... and The optimization problem is as follows:

[0274] Optimization Problem (III):

[0275] ,

[0276] Constraint 1:

[0277] ,

[0278] Constraint 2:

[0279] ,

[0280] This problem is an SDP problem, which can be solved using existing mature algorithms.

[0281] Step 5-4-2: Fix the phase shift matrix of the smart reflector to solve for the transmission precoding. At this point, the solution... The optimization problem is as follows:

[0282] ,

[0283] Optimization Problem (Ⅳ):

[0284] ,

[0285] Constraint 1:

[0286] ,

[0287] Clearly, the objective function is affected by the penalty function of the constraint terms. This results in the entire item being non-convex, which can be simplified by performing some processing on this item.

[0288] First, define Using the above formula to refer to and The irrelevant terms. Next, the penalty function is expanded to obtain formula (1):

[0289] (1),

[0290] The second term of formula (1) is processed to define the Hermitian matrix. ;

[0291] Construct a negative semidefinite matrix ,in for The largest eigenvalue. For Perform a Taylor expansion at the t-th iteration: .

[0292] Organized ,in , .

[0293] Process the third term of formula (1) and define... Construct a negative semidefinite matrix ,in for The largest eigenvalue.

[0294] Considering power constraints and define Irrelevant terms are .in, .

[0295] at this time Transformed into:

[0296] ,

[0297] We can use a first-order Taylor expansion again to remove the fourth-order terms in the above equation:

[0298] ,

[0299] The final solution The problem becomes a convex problem as follows:

[0300] Optimization problem (V)

[0301] ,

[0302] Constraint 1:

[0303] ,

[0304] This problem can be solved directly using standard convex optimization algorithms.

[0305] Step 5-4-3, since the RIS phase shift is divided into and ,in Its function is to enhance the echo signal, and its optimization only requires phase alignment with the incoming signal from the base station. After fixing other variables, for The optimization problem is as follows:

[0306] Optimization problem (VI)

[0307] ,

[0308] Constraint 1:

[0309] ,

[0310] Due to the presence of the modulo 1 constraint, this problem cannot be solved directly using existing convex optimization algorithms. We can consider using SDR to handle the constraint, or we can directly use a manifold optimization solver to solve it.

[0311] In this embodiment, the specific architecture of the method is as follows: Figure 1 As shown (orange units correspond to the RIS reflection unit of the auxiliary communication system, and blue and white units correspond to the RIS transmission unit of the auxiliary sensing system), the transmitter is equipped with There are 1 receiving and transmitting antennas, arranged at a spacing of 1 / 2. A surface-type antenna array. The receiver is... Each vehicle has a single-antenna user. Meanwhile, the RIS (Radio System) mounted on the roof is divided into two parts, one with reflection capabilities and the other with transmission capabilities, to perform communication and sensing functions respectively. For ease of distinction, they are named RISc (communication system) and RISr (radar system) according to their corresponding systems. Communication services will be assisted by the RISc section on the roof. The number of RISc and RISr units are Nc and Nr, respectively.

[0312] Meanwhile, the sensing service treats the RISr on the vehicle roof as the sensing target. In the RIS-assisted sensing scenario considered in this method, the spatial distribution of the road and base station antennas is known, and under normal circumstances, the vehicle travels within the space constrained by the road plane. Under the above assumptions, if the azimuth and elevation angles of the vehicle relative to the base station are obtained, the vehicle's coordinates on the road plane will be easily determined. At this point, the system focuses on the reflected echo from the BS-RISr-BS link. The carried two-dimensional angle information can be processed.

[0313] The following simulation settings are recommended: See simulation results below. Figure 3 and Figure 4 .

[0314] Number of base station transmit and receive antennas The number of users is 8. The sample size is 4. =1024. The total number of RIS cells is 65, the Nc to Nr ratio is 7:2, and the noise power is... Meanwhile, the BS-RISc distance is set to 90 meters, and the BS-UE distance is set to 91 meters. The distances between the users inside the vehicle and the center of the RISc are 0.86, 0.86, 0.94, and 0.75 meters.

[0315] Overall control process as follows Figure 2 As shown, after the vehicle feeds back uplink information, the information dimensions are separated; based on this, an optimization problem is established and joint optimization is performed (see step 5 for the specific process); finally, the base station outputs downlink signals to the vehicle-mounted reflector controller; through the reflector controller, the communication link between the base station, the reflector transmission unit, and the user, and the sensing link between the base station, the reflector reflection unit, and the base station are finally established.

[0316] Figure 3 Simulation results show that ( Figure 3 In this paper, "Communication rate" refers to the communication speed; "Cramér-Rao Bound" (CRB) is the Cramer-Rao Bound. Under the same sensing accuracy constraint, the communication rate achieved by the joint optimization scheme proposed in this invention is superior to that of the traditional separate optimization scheme. As the sensing accuracy requirement decreases, the communication rate of this scheme rapidly increases to the performance boundary, while the rate of the separate scheme remains at a low level. This fully demonstrates that this scheme has significant performance advantages over the traditional separate optimization, providing an efficient and feasible design approach for the engineering implementation of integrated sensing systems.

[0317] according to Figure 4 The simulation results shown ( Figure 4 In this paper, "Communication rate" (bps / Hz represents bits per hertz) and "transmit power" (dBm represents decibels per milliwatt) refer to the communication rate. As the transmit power budget increases, the power available for the system to allocate to the communication signal increases accordingly, thus the user communication rate exhibits a monotonically increasing trend. Simultaneously, under the same transmit power, stricter sensing accuracy requirements (i.e., smaller CRB values) lead to a decrease in the communication rate. This is because, within a limited power budget, meeting high-precision sensing requires allocating more power for sensing waveforms, thereby crowding out the power resources of the communication signal, creating a trade-off between communication and sensing services. This result demonstrates that the joint optimization scheme proposed in this invention can effectively characterize the power allocation pattern under different sensing accuracy constraints, providing a basis for performance evaluation of integrated sensing systems in power-constrained scenarios.

[0318] The symbols used in this embodiment are explained in Table 1.

[0319] Table 1

[0320]

[0321] This invention provides a phase-shift control method for intelligent reflective surface-assisted vehicle-to-everything (V2X) sensing integration. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A phase-shift control method integrating intelligent reflective surface-assisted vehicle-to-everything (V2X) sensing, characterized in that, Includes the following steps: Step 1: The base station performs channel estimation based on the information fed back by the vehicle, and obtains the channel matrix and vector information; Step 2: Establish the communication model and initialize the beamforming matrix and intelligent reflector phase shift matrix of the base station; Step 3: Establish a perception model; Step 4: Derive the perception performance matrix; Step 5: Establish the joint beam optimization objective and various constraints.

2. The method according to claim 1, characterized in that, Step 1 includes: the channel matrix includes: direct channels between the integrated inductive base station and the roof reflector. ,in This indicates that the matrix elements are complex numbers; This represents the total number of RIS units in the communication system. M represents the total number of RIS units in the radar system; M represents the number of antennas at the base station. Channel between the roof reflector and the k-th in-vehicle user , ; base station and the Non-line-of-sight channel between users inside the vehicle K represents the total number of users inside the vehicle.

3. The method according to claim 2, characterized in that, Step 2 includes: using and These refer to the beamforming matrix for communication and the beamforming matrix for radar systems, respectively. Set the transmission phase shift matrix of the smart reflector as follows: ;in Represents a diagonal matrix function; A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal One element; Set the reflection phase shift matrix of the smart reflector as follows: , A vector that refers to the elements on the main diagonal of a matrix; For matrix The first on the main diagonal One element; Used to represent the global phase shift matrix of RIS This is the global phase shift vector of RIS; In the Signals transmitted within each time slot Set to: ， in, For joint precoding matrix; transmitted signal , , This represents the total number of time slots. For communication signals of K users, and covariance matrix identity matrix : ; A dedicated sensing signal for achieving full degrees of freedom in target sensing across M channels, and The covariance matrix is ​​the identity matrix. ,Right now ; Furthermore, the communication signals and the sensing signals are independent of each other. .

4. The method according to claim 3, characterized in that, Step 3 includes: transmitting the sensing signal from the base station back to the base station receiving antenna via a reflecting element. The model is as follows: ， in The radar cross section is T, which represents the transpose. Let be the target response matrices of the base station and the target, and let the noise vector follow the order of the base station and the target response matrices. The mean is zero vector and the covariance matrix is Complex Gaussian distribution, Let Variance be the variance of the noise vector. It is an M-dimensional identity matrix; This is the channel between the reflecting unit and the base station. For the sending end steering vector, The starting angle for the signal to travel from the base station to the target. For the receiver steering vector, These represent the azimuth and elevation angles at which the signal arrives at the reflecting unit, respectively. Combined echo signals Each sample, the base station will receive the signal Described as: ， in For the equivalent echo channel matrix, the signal matrix ,in For the signal in time slot L; noise matrix ,in The noise in time slot L.

5. The method according to claim 4, characterized in that, Step 4 includes: Step 4-1, vectorize the signal received by the base station: ， in For the signal section, For the noise part, vec indicates vectorization. Represent the vectorized signal; set the unknown parameters to be estimated. ; This refers to the two-dimensional angle information of the vehicle relative to the base station, where Let be a two-dimensional real vector composed of the real and imaginary parts of the complex gain of the target reflection. and These correspond to the real and imaginary parts of the complex gain of reflection, respectively; Considering noise compliance If the complex Gaussian distribution is followed, then ;estimate The (i,j)th element of the Fisher Information Matrix (FIM) The expression is as follows: , Corresponding to The i-th component of the vector, Corresponding to The j-th component of the vector; Re represents a partial differential; Re{⋅} represents taking the real part of a complex number; Step 4-2, extract angle information from the echo signal. Calculate partial derivatives to establish the Fisher Information Matrix (FIM); about The partial derivatives are shown below: ， ， ， in and These represent the echo channels. right Find the partial derivative, with respect to Find the partial derivative; Obtain the required elements of the Fisher Information Matrix (FIM) ; The elements of the matrix are represented as follows: Fisher information matrix of angle parameters , The mutual information vector between the angle parameter and the complex gain , transpose , Among them, the mutual information between azimuth angle and complex gain , Mutual information between pitch angle and complex gain , Fisher information for the complex gain itself; For submatrix The expression is as follows: Self-information of azimuth , Mutual information between azimuth and elevation angles , Mutual information between pitch and azimuth , Self-information of pitch angle ; Where Tr represents the trace of the matrix; the superscript H represents the conjugate transpose; Step 4-3, Separation Angle Cramer-Rao Lower Bound (CRB): Invert the Fisher Information Matrix (FIM) to obtain the Cramer-Rao Lower Bound (CRB) matrix. By separating the two angles of Cramer-Lower Boundary (CRB), we obtain... ,in and These correspond to the lower bounds of Cramer-Rao for azimuth and elevation, respectively.

6. The method according to claim 5, characterized in that, Step 5 includes the following steps: Step 5-1: The base station initializes the algorithm and sets the initial number of iterations. With maximum iteration Set precision ; Step 5-2, Initialize the base station precoding matrix The phase shift matrix of the intelligent reflective surface And towards quantification: The precoded vector is obtained by expanding it by rows and columns. , Then, extracting the elements of the main pair of lines yields the phase shift vector. ; Keep parameters updated during iteration: Let , Save the current parameters as the state of the previous round; Step 5-3: Based on the known channel state information, establish and solve the optimization problem (Ⅰ), and extract and transform the coupling parameters. For non-convex problems and non-convex constraints, transform them to calculate the user's sum and communication rate. Consider the problem of maximizing the sum and communication rate, while satisfying: the sensing system should meet the minimum sensing performance requirements, the base station transmitter power should not exceed the set maximum power constraint, and the phase mode constraint of the intelligent reflector. The optimization problem (I) is expressed as: ， in, Indicates the base station and the first Non-line-of-sight channel between users inside the vehicle Transpose of; This represents the channel between the roof reflector and the k-th user inside the vehicle. Transpose of; This is the direct channel between the base station and the roof reflector; This is the pre-encoding vector corresponding to user k; For other users' precoded vectors; For noise variance; Constraint 1: ， Constraint 2: ， Constraint 3: ， in The CRB threshold; Represents the maximum transmission power; Reflection phase shift matrix The nth diagonal element; Step 5-4, Solve the optimization problem (III) and update the auxiliary variables. and Solve optimization problem (V) to update the transmission precode w, and solve optimization problem (VI) to obtain the transmission phase shift matrix of the smart reflector. The phase shift matrix of the intelligent reflector is obtained by aligning the phase of the wave received from the base station with the wave. Calculate user and communication rates At the same time, the following conditions must be met: the sensing system should meet the minimum sensing performance requirements, the power of the base station transmitter should not exceed the set maximum power constraint, and the phase mode constraint of the intelligent reflector should be met. Step 5-5: Adjust the penalty parameters until all variables converge; if convergence fails, utilize the obtained intermediate variables, i.e., the precoding vector. and phase shift vector Repeat steps 5-3 to 5-4 to update and obtain new convergence variables, including the converged precoding vector. and Then let t = t + 1, and return to step 5-2; Steps 5-6: Recombine the precoding vector and the smart reflector phase shift vector into a matrix, and output the optimal transmission precoding matrix obtained through iteration. Optimal phase shift control matrix of the intelligent reflector .

7. The method according to claim 6, characterized in that, Step 5-3 includes the following steps: Step 5-3-1: Perform a fractional transformation on the optimization problem. After performing a Lagrange dual transformation on the objective function to remove the outer logarithmic function, the objective function is transformed using fractional programming as follows to obtain a new objective function: ， in This is an auxiliary variable introduced for the ratio term; it is used to expand the third term of the objective function and introduce the auxiliary variable. Then, the objective function is transformed into: ， in This is a simplified form of the new objective function; Transform into about and The quadratic polynomial: ， in , Used to refer to the objective function that does not contain and Other items, while defining , The remaining items are described below: ， for The coefficient vector of the first-order terms, and These are two auxiliary variables corresponding to user k; ， for The quadratic coefficient matrix; ， This is the weighted equivalent channel vector for the k-th user and the i-th beam direction, corresponding to the noise interference term; The selection matrix represents the extraction of the i-th beam from the total precoding vector; ， for The vector of coefficients of the first-order terms; ， for The quadratic coefficient matrix; Step 5-3-2: Process constraint 1 by introducing auxiliary variables. Constraint 1 is transformed into the following two constraints: Constraint 1-1: ， in For matrix The reversed trace; Constraint 1-2: ， Using Schur complement conditions, for Constraint 1-2 is equivalent to: ； Step 5-3-3: Introduce auxiliary variables Will and Extracting the semidefinite constraints and using the augmented Lagrangian method, the optimization problem (I) is transformed into the optimization problem (II): ， Constraint 1: ， Constraint 2: ， Constraint 3: ， Constraint 4: ， Constraint 5: ， in, ， ， ， ， ， ， in As dual variables, It is a second-order identity matrix. As a penalty factor, Let i represent the i-th auxiliary variable. Let i represent the i-th dual variable. Used in referential expressions where it is not explicitly included For the other terms, i takes values ​​from 1 to 6.

8. The method according to claim 7, characterized in that, Step 5-4 includes the following steps: Step 5-4-1, Update the auxiliary variable penalty term With Lagrange multipliers: auxiliary variables for user k and The update is achieved through calculation and It is concluded that; at this time , ; Solve The optimization problem for f is as follows: Optimization Problem (III): ， Constraint 1: ， Constraint 2: ； Step 5-4-2: Fix the phase shift matrix of the intelligent reflector to solve the transmission precoding; at this time, solve... The optimization problem is as follows: ， Optimization Problem (Ⅳ): ， Constraint 1: ， First, define to refer to and Irrelevant terms; next, expanding the penalty function yields formula (1): （1）， The second term of formula (1) is processed to define the Hermitian matrix. Construct a negative semidefinite matrix ,in for The largest eigenvalue; for Perform a Taylor expansion at the t-th iteration: ; Organized ,in , ; Process the third term of formula (1) and define... Construct a negative semidefinite matrix ,in for The largest eigenvalue; Considering power constraints and define irrelevant items for: ， in, ; at this time Transformed into: ， Applying a first-order Taylor expansion again to remove the fourth-order terms in the formula, we get: ， The final solution The problem becomes a convex problem as follows: Optimization problem (V): ， Constraint 1: ； Step 5-4-3, for The optimization problem is as follows: Optimization problem (VI): ， Constraint 1: 。 9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, It stores a computer program or instructions that, when run on a computer, perform the steps of the method as described in any one of claims 1 to 8.

Images