Transmit beamforming method for millimeter wave communication and sensing integrated system

Abstract

The present invention discloses a transmission beamforming method for a millimeter wave communication and perception integrated system. Under the conditions of meeting the communication user's reception signal-to-interference-noise ratio requirements and total transmission power constraints, the method aims to minimize the gap between the beam pattern of the transmission beam to be designed and the given radar beam pattern, optimizes the design of the transmission beam through an alternating minimization method, and obtains a beamforming matrix without hardware constraints; on this basis, the hardware constraints of the phased array are further considered, and the analog beamforming matrix and the digital beamforming matrix of the millimeter wave communication and perception integrated system are optimized and designed.

Description

Technical Field

[0001] The present invention belongs to the field of millimeter wave wireless communications, and in particular relates to a transmission beamforming method for a millimeter wave communication and sensing integrated system. Background Art

[0002] As a candidate technology for the next generation of wireless communications, Integrated Sensing and Communications (ISAC) is attracting widespread attention and interest from academia, industry, and government. ISAC can fully share space, time, frequency, and power resources, reducing the hardware and software complexity of wireless communications and radar sensing. At present, the research work on ISAC systems is mainly divided into communication-centric design, perception-centric design, and communication-perception joint design. On the other hand, scarce spectrum resources have prompted everyone to develop and utilize higher-frequency millimeter-wave spectrum. Generally speaking, millimeter-wave massive MIMO systems use phased arrays to form highly directional communication beams to reduce the path loss of high-frequency transmission, which has a common feature with radar systems-both use phased arrays.

[0003] One of the challenges facing MIMO ISAC is beamforming design. Reference [1] divides the antenna array into a communication array and a sensing array, which are independent of each other. By optimizing analog beamforming, the mutual interference between the two arrays is minimized as much as possible. Compared with the traditional semi-positive definite programming method, better communication performance and sensing performance can be obtained. [1] F. Dong, W. Wang, Z. Hu and T. Hui, "Low-Complexity Beamformer Design for Joint Radar and Communications Systems," IEEE Commun. Lett., vol. 25, no. 1, pp. 259-263, Jan. 2021.

[0004] However, the above design only considers the design form of separating the sensing array and the communication array, and does not make full use of the antenna resources. At the same time, the above design is not suitable for the millimeter wave frequency band, cannot be used for the communication and sensing integrated system of hybrid beamforming, and has no strict restrictions on the communication performance. Summary of the invention

[0005] The purpose of the present invention is to provide a transmission beamforming method for a millimeter wave communication and perception integrated system, and to obtain an optimized design result of an analog beamforming matrix and a digital beamforming matrix that meet the hardware constraints of a phased array while meeting the requirements of the total transmission power of a base station and the received signal-to-interference-noise ratio of a communication user, so as to solve the technical problems that the existing design only considers the separation of a perception array and a communication array, does not fully utilize antenna resources, and at the same time, the above design is not suitable for the millimeter wave frequency band, cannot be used for a communication and perception integrated system with hybrid beamforming, and has no strict restrictions on communication performance.

[0006] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0007] A transmission beamforming method for a millimeter wave communication and sensing integrated system comprises the following steps:

[0008] Step 1: Design a millimeter wave communication and perception integrated system model;

[0009] Step 2: Design a channel model for millimeter wave communication;

[0010] Step 3: Design a signal transmission model for millimeter wave communication;

[0011] Step 4: Optimization problem modeling is performed based on the system model, channel model and signal transmission model;

[0012] Step 5: Design the transmit beam by solving the modeled optimization problem to obtain a beamforming matrix without hardware constraints;

[0013] Step 6: Design an analog beamforming matrix and a digital beamforming matrix that meet hardware constraints based on the beamforming matrix without hardware constraints.

[0014] Furthermore, the system model in step 1 is as follows:

[0015] Radar communication dual-function base station deployed N BS Antennas serving N c The antennas are deployed as a uniform linear array, the distance between adjacent antennas is half a wavelength, and the number of base station RF links is N. RF , where N c RF links are used to form N c Communication beams, N t RF links are used to form radar sensing beams, that is, N RF =N t +N c .

[0016] Furthermore, in step 2, the channel model in millimeter wave communication is designed:

[0017] The channel between the nth communication user and the base station can be expressed as:

[0018]

[0019] Among them, L n Represents the number of wireless transmission paths, g l represents the gain of the lth path, represents the starting angle of the lth path, H is the conjugate transpose operation, represents the channel steering vector, which is defined as:

[0020]

[0021] It is N BS and Function, l = 1, 2, ..., L n .

[0022] Furthermore, in step 3, a signal transmission model for millimeter wave communication is designed:

[0023] y=HF RF F BB x+η

[0024] in, represents the simulated beamforming matrix of the base station, represents the digital beamforming matrix of the base station, F BB The modulus and amplitude of each element are variable, and F RF The modulus of each element of is always 1; Represents h according to step 2 n The definition of a new joint channel matrix is ​​defined, Represents the vector of the transmitted signal, Represents the communication signal sent, represents the radar signal sent, and satisfies Indicates that the dimension is N RF The identity matrix of Represents the Gaussian white noise at the receiving end, each element of which is independent and obeys the complex Gaussian distribution

[0025] Furthermore, in step 3, define the matrix F = F RF F BB , and define the i-th column of F as f i , i = 1, 2, ..., N RF , then the base station's transmit beam is

[0026] Furthermore, in step 4, based on the system model, channel model and signal transmission model, the modeling optimization problem is as follows:

[0027]

[0028]

[0029] γ n ≥Γ n , n=1,2,...,N c

[0030] Among them, ‖·‖2 represents the 2-norm of the vector. In the constraints,

[0031]

[0032] represents the actual received signal-to-interference-noise ratio of the nth communication user, Γ n Represents the preset user signal-to-interference-noise ratio threshold, P T represents the total transmission power of the base station; according to the preset M spatial sampling points φ1, φ2…φ M Forming a spatial steering vector matrix Φ = [α(N BS ,φ1),α(N BS ,φ2),...,α(N BS ,φ M )] T , where -1≤φ1<φ2<…<φ M ≤1; represents a given radar beam pattern; D is a preset M-dimensional non-negative diagonal matrix, and the i-th diagonal element of D represents the i The transmit beam to be designed The weight of the closeness of the formed pattern to the given radar beam pattern b is obtained by adjusting Under the constraints of the total transmission power of the base station and the signal-to-interference-to-noise ratio of the communication users, the transmission beam The resulting pattern is close to the given radar beam pattern b.

[0033] Furthermore, the method for solving the optimization problem in step 5 is as follows:

[0034] Step 5.1: Introduce phase vector To improve the degree of freedom in solving the optimization problem, the elements in p satisfy: |p i |=1,i=1,2,...,M;Define an M-dimensional diagonal matrix A=diag{b}, that is, the i-th diagonal element of A is equal to the i-th element of b, i=1,2,...,M;Rewrite the optimization problem as:

[0035]

[0036]

[0037] γ n ≥Γ n , n=1,2,...,N c

[0038] |p i |=1,i=1,2,...,M

[0039] Step 5.2: Convert the optimization problem described in step 5.1 into two sub-problems. In sub-problem 1, fix p to optimize Sub-problem 1 is expressed as follows:

[0040]

[0041]

[0042] γ n ≥Γ n， n=1,2,...,N c

[0043] Sub-issue 2 fixed To optimize p, subproblem 2 is expressed as follows:

[0044]

[0045] st|p i |=1,i=1,2,...,M

[0046] Every element of p satisfies the constraint |p i |=1,i=1,2,...,M,randomly initialize each element of p;

[0047] Step 5.3: For sub-problem 1 described in step 5.2, the solution is as follows:

[0048]

[0049]

[0050] γ n ≥Γ n， n=1,2,...,N c

[0051] |p i |=1,i=1,2,...,M

[0052] Define the auxiliary matrix

[0053]

[0054] Contains N RF Unit Matrix definition

[0055]

[0056] By concatenating multiple beamforming vectors into a long column vector, the base station's transmit beam can be rewritten as:

[0057]

[0058] The total transmission power constraint of the base station is rewritten as

[0059] Define the auxiliary matrix

[0060]

[0061] Contains N RF -1 zero matrix and an identity matrix And in There are i-1 zero matrices before and N zero matrices after RF -i zero matrices; then the actual received signal-to-interference-to-noise ratio constraint of the nth communication user is expressed as:

[0062]

[0063] The above formula can be expressed as:

[0064]

[0065] Taking into account the h n S n f is rotated, that is, the overall phase is adjusted, so that h n S n f becomes a positive real number, so the above formula is expressed in the form of a second-order cone constraint:

[0066]

[0067] The vector t is a computational intermediate quantity introduced to embody the second-order cone constraint form, and t is defined as follows

[0068]

[0069] Therefore, subproblem 1 is expressed as a standard second-order cone programming problem:

[0070]

[0071]

[0072]

[0073] Step 5.4: For sub-problem 2 described in step 5.2, the solution is as follows:

[0074] Ignore constraints first |p i |=1,i=1,2,...,M,then we can get the least squares solution of subproblem 2 under unconstrained conditions:

[0075]

[0076] Where W is a matrix calculated in advance, and W is defined as follows

[0077]

[0078] On this basis, consider the constraint |p i |=1,i=1,2,...,M,then the effective solution of subproblem 2 is The kth element of is:

[0079]

[0080] Step 5.5, alternately execute step 5.3 and step 5.4 until the stopping condition is met; the stopping condition is set as: the number of executions reaches the maximum preset number of times, or the change in the objective function of the optimization problem between two consecutive executions is less than a preset threshold value; the solution to the optimization problem obtained after the stopping condition is met is defined as Define beamforming matrix without hardware constraints

[0081] Furthermore, subproblem 1 in step 5.3 is expressed as a standard second-order cone programming problem and solved using the optimization toolbox.

[0082] Further, in step 6, based on the beamforming matrix without hardware constraints Design analog beamforming matrices and digital beamforming matrices that meet hardware constraints and model the hybrid beamforming design problem as follows:

[0083]

[0084]

[0085] |F RF (m, n)|=1,

[0086] m=1,2,...,N BS , n=1,2,...,N RF

[0087] Step 6.1: For the hybrid beamforming design problem, ignore the constraint of the total transmit power of the base station. At the same time, according to |F RF (m,n)|=1,m=1,2,...,N BS , n=1,2,...,N RF This constraint has an effect on N RF All elements of are randomly initialized;

[0088] Step 6.2: Fix N RF Optimize F BB , through the least squares estimation, we can get F BB Estimates:

[0089]

[0090] Step 6.3, fix F BB Optimize F RF When , the hybrid beamforming design problem is expressed as:

[0091]

[0092] st|F RF (m, n)|=1,

[0093] m=1,2,...,N BS , n=1,2,...,N RF

[0094] The problem is solved by Riemannian manifold optimization;

[0095] Step 6.4, alternately execute step 6.2 and step 6.3 until a termination condition is met, wherein the termination condition is set to: the number of executions reaches a maximum preset number of times, or the change in the objective function of the hybrid beamforming design problem between two consecutive executions is less than a preset threshold value; the solution to the hybrid beamforming design problem obtained after the termination condition is met is and The final designed analog beamforming matrix is Consider the constraints of the total transmission power of the base station The final designed digital beamforming matrix is

[0096] The transmit beamforming method of a millimeter wave communication and perception integrated system designed by the present invention has the following advantages:

[0097] 1. Compared with the existing solutions, the system modeling of the present invention takes into account the hybrid beamforming architecture applied to the millimeter wave frequency band for the radar communication dual-function base station, and further designs the RF link allocation structure for radar beamforming and communication beamforming respectively, which can meet the needs of integrated communication perception in multiple scenarios;

[0098] 2. The problem modeling of the present invention is different from the existing solutions. The present invention uses a communication sensing shared array. Under the premise of ensuring that the communication users meet the requirements of the receiving signal-to-interference-noise ratio, the present invention aims to minimize the gap between the beam pattern of the sensing beam to be designed and the beam pattern of the ideal sensing beam. Under the hardware constraints of the millimeter wave system such as the constant envelope constraint of the phase shifter, the present invention designs a digital beamforming matrix and an analog beamforming matrix respectively. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 A schematic diagram of a millimeter wave communication and sensing integrated system model used in an embodiment of the present invention;

[0100] Figure 2 A flowchart of a scheme designed for an embodiment of the present invention;

[0101] Figure 3 Comparison diagram of two transmit beam patterns with and without hardware constraints and an ideal beam pattern designed for an embodiment of the present invention;

[0102] Figure 4 It is a comparison diagram of the mean square error of the approximation of the two transmission beam patterns with and without hardware constraints designed by the present invention to the ideal radar beam pattern under different user signal-to-interference-noise ratio requirements. DETAILED DESCRIPTION

[0103] In order to better understand the purpose, structure and function of the present invention, the following further describes in detail the transmission beamforming method of the millimeter wave communication and perception integrated system of the present invention in conjunction with the accompanying drawings.

[0104] The present invention comprises the following steps:

[0105] Step 1: Design a millimeter wave communication and perception integrated system model;

[0106] The millimeter wave communication perception integrated system model designed by the present invention is as follows Figure 2 As shown. The radar communication dual-function base station deploys N BS Antennas serving N c The antennas are deployed as a uniform linear array, the distance between adjacent antennas is half a wavelength, and the number of base station RF links is N. RF , where N c RF links are used to form N c Communication beams, N tRF links are used to form radar sensing beams, that is, N RF =N t +N c .

[0107] Step 2: Design a channel model for millimeter wave communication;

[0108] The channel between the nth communication user and the base station is expressed as:

[0109]

[0110] Among them, L n Represents the number of wireless transmission paths, g l represents the gain of the lth path, represents the starting angle of the lth path, H is the mathematical conjugate transpose operation, represents the channel steering vector, which is defined as:

[0111]

[0112] It is N BS and Function, l = 1, 2, ..., L n .

[0113] Step 3: Design a signal transmission model for millimeter wave communication;

[0114] y=HF RF F BB x+η

[0115] in, represents the simulated beamforming matrix of the base station, represents the digital beamforming matrix of the base station, F BB The modulus and amplitude of each element are variable, and F RF The modulus of each element of is always 1; Represents h according to step 2 n The definition of a new joint channel matrix is ​​defined, Represents the vector of the transmitted signal, Represents the communication signal sent, represents the radar signal sent, and satisfies Indicates that the dimension is N RF The identity matrix of Represents the Gaussian white noise at the receiving end, each element of which is independent and obeys the complex Gaussian distribution To simplify the representation, define the matrix F = F RF F BB , and further define the i-th column of F as f i, i = 1, 2, ..., N RF , then the base station's transmit beam is

[0116] Step 4: Optimization problem modeling is performed based on the system model, channel model and signal transmission model; the modeling optimization problem is:

[0117]

[0118]

[0119] γ n ≥Γ n， n=1,2,...,N c

[0120] in,

[0121]

[0122] represents the actual received signal-to-interference-noise ratio of the nth communication user, Γ n Represents the preset user signal-to-interference-noise ratio threshold, P T Represents the total transmission power of the base station. According to the preset M spatial sampling points φ1, φ2…φ M Forming a spatial steering vector matrix Φ = [α(N BS ,φ1),α(N BS ,φ2),...,α(N BS ,φ M )] T , where -1≤φ1<φ2<…<φ M ≤1. represents a given radar beam pattern. D is a preset M-dimensional non-negative diagonal matrix, and the i-th diagonal element of D represents the i The transmit beam to be designed The weight of the closeness of the formed pattern to the given radar beam pattern b, the larger the weight, the higher the closeness requirement, and the smaller the weight, the lower the closeness requirement. Under the constraints of the total transmission power of the base station and the signal-to-interference-to-noise ratio of the communication users, the transmission beam The resulting pattern is as close as possible to the given radar beam pattern b.

[0123] Step 5: Design the transmit beam by solving the modeled optimization problem to obtain a beamforming matrix without hardware constraints;

[0124] Step 5.1: Introduce phase vector To improve the degree of freedom in solving the optimization problem, the elements in p satisfy: |p i|=1,i=1,2,...,M. Define an M-dimensional diagonal matrix A=diag{b}, that is, the i-th diagonal element of A is equal to the i-th element of b, i=1,2,...,M. Rewrite the optimization problem as:

[0125]

[0126]

[0127] γ n ≥Γ n， n=1,2,...,N c

[0128] |p i |=1,i=1,2,...,M

[0129] Step 5.2: Convert the optimization problem described in step 5.1 into two sub-problems. In sub-problem 1, fix p to optimize Sub-problem 1 is expressed as follows:

[0130]

[0131]

[0132] γ n ≥Γ n , n=1,2,...,N c

[0133] Sub-issue 2 fixed To optimize p, subproblem 2 is expressed as follows:

[0134]

[0135] st|p i |=1,i=1,2,...,M

[0136] Every element of p satisfies the constraint |p i |=1,i=1,2,…,M, each element of p is randomly initialized.

[0137] Step 5.3: For sub-problem 1 described in step 5.2, the solution is as follows:

[0138]

[0139]

[0140] γ n ≥Γ n , n=1,2,...,N c

[0141] |p i |=1,i=1,2,...,M

[0142] Define the auxiliary matrix

[0143]

[0144] Contains N RF Unit Matrix Furthermore, define

[0145]

[0146] Multiple beamforming vectors are concatenated into a long column vector. The base station's transmit beam can be rewritten as:

[0147]

[0148] The total transmission power constraint of the base station can be rewritten as

[0149] Define the auxiliary matrix

[0150]

[0151] Contains N RF -1 zero matrix and an identity matrix And in There are i-1 zero matrices before and N zero matrices after RF -i zero matrices. Then the actual received signal-to-interference-to-noise ratio constraint of the nth communication user can be expressed as:

[0152]

[0153] The above formula can be further simplified as follows:

[0154]

[0155] Taking into account the h n S n f is rotated, that is, the overall phase is adjusted, so that h n S n f becomes a positive real number, so the above formula can be further expressed in the form of a second-order cone constraint:

[0156]

[0157] in

[0158]

[0159] Therefore, subproblem 1 can be expressed as a standard second-order cone programming problem:

[0160]

[0161]

[0162]

[0163] This problem can be solved using existing optimization toolboxes.

[0164] Step 5.4: For sub-problem 2 described in step 5.2, the solution is as follows:

[0165] Ignore constraints first |p i |=1,i=i,2,...,M,then we get the least squares solution of subproblem 2 under unconstrained conditions:

[0166]

[0167] Where W is a matrix that can be calculated in advance, and W is defined as follows

[0168]

[0169] On this basis, consider the constraint |p i |=1,i=1,2,...,M,then the effective solution of subproblem 2 is The kth element of is:

[0170]

[0171] Step 5.5: Alternately execute step 5.3 and step 5.4 until the stopping condition is met. The stopping condition can be set as: the number of executions reaches the maximum preset number of times, or the change in the objective function of the optimization problem between two consecutive executions is less than a preset threshold value. The solution to the optimization problem obtained after the stopping condition is met is defined as Define beamforming matrix without hardware constraints

[0172] Step 6: Design an analog beamforming matrix and a digital beamforming matrix that meet hardware constraints based on the beamforming matrix without hardware constraints. This problem can be expressed as:

[0173]

[0174]

[0175] |F RF (m,n)|=1,

[0176] m=1,2,...,NBS ,n=1,2,...,N RF

[0177] Step 6.1: For the hybrid beamforming design problem, ignore the constraint of the total transmit power of the base station. At the same time, according to |F RF (m,n)|=1,m=1,2,...,N BS , n=1,2,...,N RF This constraint has an impact on F RF Each element of is initialized randomly.

[0178] Step 6.2, fix F RF Optimize F BB , through the least squares estimation, we can get F BB Estimates:

[0179]

[0180] Step 6.3, fix F BB Optimize F RF When , the hybrid beamforming design problem is expressed as:

[0181]

[0182] st|F RF (m, n)|=1,

[0183] m=1,2,...,N BS ,n=1,2,...,N RF

[0184] This problem can be solved by Riemannian manifold optimization.

[0185] Step 6.4: Alternately execute step 6.2 and step 6.3 until a termination condition is met. The termination condition can be set as: the number of executions reaches a maximum preset number of times, or the change in the objective function of the hybrid beamforming design problem between two consecutive executions is less than a preset threshold value. The solution to the hybrid beamforming design problem obtained after the termination condition is met is and The final designed analog beamforming matrix is Consider the constraints of the total transmission power of the base station The final designed digital beamforming matrix is

[0186] The overall algorithm flow chart of the above steps is as follows Figure 2 shown.

[0187] The present invention is further described below in conjunction with simulation conditions and results:

[0188] Figure 3 A comparison diagram of two transmit beam patterns with and without hardware constraints designed by the present invention and an ideal beam pattern is shown. Figure 3 In the example, the number of base station antennas is N BS =128, number of radio frequency links N RF =3, number of users N c =3, and the channel between the base station and each user has a line-of-sight path and two non-line-of-sight paths, and the gain of the line-of-sight path satisfies The gains of the two non-line-of-sight paths satisfy The number of spatial sampling points is M = 400, and the sampling range is [-1, 1], that is, -1≤φ1<φ2<…<φ M ≤1, where i = 1, 2, ..., M. The user's channel line-of-sight directions are -70°, -40°, and -10°, respectively. The minimum requirement for the user's received signal-to-interference-to-noise ratio is Γ1=Γ2=Γ3=30dB. At the same time, assuming that the ideal radar target beam pattern b covers a range of [10°, 30°]∪[40°, 60°]. Accordingly, in order to ensure that the calculated beam pattern gain is a relative value compared with the omnidirectional antenna gain, the amplitude of the beam pattern b is set to

[0189]

[0190] After completing the above specific parameter settings, execute Figure 2 The algorithm flow chart described above can obtain the beamforming matrix without hardware constraints. and the hybrid beamforming matrix under hardware constraints The beam patterns generated by the two beamforming matrices are further obtained as follows: Figure 3 As shown in the solid and dotted lines, Figure 3 The middle dotted line represents the ideal radar beam pattern. It can be seen from the simulation results that the three peaks on the left are all accurately pointed to the line-of-sight directions of the three user channels. The peak value is determined by the user's signal-to-noise ratio requirements. The greater the signal-to-noise ratio required by the user, the higher the peak value. At the same time, it can be seen that the two beam patterns designed in this paper accurately cover the coverage range required by the ideal radar beam. Since the communication beams formed are very narrow, they occupy a small part of the energy. Therefore, the amplitude of the beam patterns generated by the two beamforming matrices within this coverage range is slightly lower than the amplitude of the radar beam pattern. And due to the constant modulus constraints of the elements in the analog beamforming matrix, there are some energy leakages in the transmit beam generated by the hybrid beamforming matrix, which also causes its performance to be slightly worse than that without hardware constraints.

[0191] Figure 4The figure shows the mean square error of the two transmit beam patterns designed by the present invention with and without hardware constraints to the ideal radar beam pattern under different user signal-to-interference-to-noise ratio requirements. BS =32, 64, 128) were used as controls. The three groups of experiments were performed Figure 2 The algorithm flow chart described above can obtain the beamforming matrix without hardware constraints. and the hybrid beamforming matrix under hardware constraints Furthermore, the approximation error is calculated

[0192]

[0193] Among them, F is and They are used to calculate the mean square error of the approximation of the ideal radar beam pattern by beamforming without hardware constraints and hybrid beamforming with hardware constraints. Figure 4 In the figure, the solid lines marked with stars and circles represent the mean square error of the approximation of the ideal radar beam pattern by the transmit beam pattern with and without hardware constraints, respectively. The results show that the greater the user's signal-to-interference-noise ratio requirement, the greater the mean square error of the approximation of the two beam patterns to the ideal radar beam pattern. This indicates that the improvement in communication quality leads to a decrease in the approximation performance of the radar beam pattern. At the same time, by comparing the experimental results with different numbers of antennas, it can be seen that the increase in the number of antennas increases the degree of freedom in designing the beam and the better the approximation effect. In addition, the increase in the number of antennas makes the simulated beamforming matrix F RF The more constant modulus constraints there are on the elements in and The error between them increases, which leads to a greater difference in the approximation performance of the transmit beam pattern without hardware constraints and the transmit beam pattern satisfying hardware constraints to the ideal radar beam pattern.

[0194] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A transmit beamforming method for a millimeter wave communication and sensing integrated system, characterized in that: The following steps are involved: Step 1: Design a millimeter wave communication and perception integrated system model; Step 2: Design a channel model for millimeter wave communication; Step 3: Design a signal transmission model for millimeter wave communication; Step 4: Optimization problem modeling is performed based on the system model, channel model and signal transmission model; Step 5: Design the transmit beam by solving the modeled optimization problem to obtain a beamforming matrix without hardware constraints; Step 6: designing an analog beamforming matrix and a digital beamforming matrix that meet hardware constraints based on the beamforming matrix without hardware constraints; The system model in step 1 is as follows: Radar communication dual-function base station deployed N BS Antennas serving N c The antennas are deployed as a uniform linear array, the distance between adjacent antennas is half a wavelength, and the number of base station RF links is N. RF , where N c RF links are used to form N c Communication beams, N t RF links are used to form radar sensing beams, that is, N RF =N t +N c ; In step 2, the channel model in millimeter wave communication is designed: The channel between the nth communication user and the base station is expressed as: Among them, L n Represents the number of wireless transmission paths, g l represents the gain of the lth path, represents the starting angle of the lth path, H is the conjugate transpose operation, represents the channel steering vector, and defines for: It is N BS and Function, l=1,2,…,L n ; In step 3, design the signal transmission model for millimeter wave communication: and=HF RF F BB x+η in, represents the simulated beamforming matrix of the base station, represents the digital beamforming matrix of the base station, F BB The modulus and amplitude of each element are variable, and F RF The modulus of each element of is always 1; Represents h according to step 2 n The definition of a new joint channel matrix is ​​defined, Represents the vector of the transmitted signal, Represents the communication signal sent, represents the radar signal sent, and satisfies Indicates that the dimension is N RF The identity matrix of Represents the Gaussian white noise at the receiving end, each element of which is independent and obeys the complex Gaussian distribution In step 4, based on the system model, channel model and signal transmission model, the modeling optimization problem is as follows: c n ≥C n ,n=1,2,…,N c Among them, ‖·‖2 represents the 2-norm of the vector. In the constraints, represents the actual received signal-to-interference-noise ratio of the nth communication user, Γ n Represents the preset user signal-to-interference-noise ratio threshold, P T represents the total transmission power of the base station; according to the preset M spatial sampling points φ1, φ2…φ M Forming a spatial steering vector matrix Φ = [α(N BS ,φ1),α(N BS ,φ2),…,α(N BS ,φ M )] T , where -1≤φ1<φ2<…<φ M ≤1; represents a given radar beam pattern; D is a preset M-dimensional non-negative diagonal matrix, and the i-th diagonal element of D represents the i The transmit beam to be designed The weight of the closeness of the formed pattern to the given radar beam pattern b. This optimization problem is solved by adjusting Under the constraints of the total transmission power of the base station and the signal-to-interference-noise ratio of the communication users, the transmission beam The pattern formed is close to the given radar beam pattern b; The method for solving the optimization problem in step 5 is as follows: Step 5.1: Introduce phase vector To improve the degree of freedom in solving the optimization problem, the elements in p satisfy: |p i |=1,i=1,2,…,M; define an M-dimensional diagonal matrix A=diag{b}, that is, the i-th diagonal element of A is equal to the i-th element of b, i=1,2,…,M; rewrite the optimization problem as: c n ≥C n ,n=1,2,…,N c |p i |=1,i=1,2,…,M Step 5.2: Convert the optimization problem described in step 5.1 into two sub-problems. In sub-problem 1, fix p to optimize Sub-problem 1 is expressed as follows: c n ≥C n ,n=1,2,…,N c Sub-issue 2 fixed To optimize p, subproblem 2 is expressed as follows: s.t.|p i |=1,i=1,2,…,M Every element of p satisfies the constraint |p i |=1,i=1,2,…,M, each element of p is randomly initialized; Step 5.3: For sub-problem 1 described in step 5.2, the solution is as follows: c n ≥C n ,n=1,2,…,N c |p i |=1,i=1,2,…,M Define the auxiliary matrix Contains N RF Unit Matrix definition By concatenating multiple beamforming vectors into a long column vector, the base station's transmit beam can be rewritten as: The total transmission power constraint of the base station is rewritten as Define the auxiliary matrix Contains N RF -1 zero matrix and an identity matrix And in There are i-1 zero matrices before and N zero matrices after RF -i zero matrices; then the actual received signal-to-interference-to-noise ratio constraint of the nth communication user is expressed as: The above formula can be expressed as: Taking into account the h n S n f is rotated, that is, the overall phase is adjusted, so that h n S n f becomes a positive real number, so the above formula is expressed in the form of a second-order cone constraint: The vector t is an intermediate quantity introduced to embody the second-order cone constraint form, and t is defined as follows: Therefore, subproblem 1 is expressed as a standard second-order cone programming problem: Step 5.4: For sub-problem 2 described in step 5.2, the solution is as follows: Ignore constraints first |p i |=1,i=1,2,…,M, we get the least squares solution of subproblem 2 under unconstrained conditions: Where W is a matrix calculated in advance, and W is defined as follows On this basis, consider the constraint |p i |=1,i=1,2,…,M, the effective solution to subproblem 2 is The kth element of is: Step 5.5, alternately execute step 5.3 and step 5.4 until the stopping condition is met; the stopping condition is set as: the number of executions reaches the maximum preset number of times, or the change in the objective function of the optimization problem between two consecutive executions is less than a preset threshold value; the solution to the optimization problem obtained after the stopping condition is met is defined as Define beamforming matrix without hardware constraints In step 6, the beamforming matrix without hardware constraints is Design analog beamforming matrices and digital beamforming matrices that meet hardware constraints and model the hybrid beamforming design problem as follows: |F RF (m,n)|=1, m=1,2,…,N BS ,n=1,2,…,N RF Step 6.1: For the hybrid beamforming design problem, ignore the constraint of the total transmit power of the base station. At the same time, according to |F RF (m,n)|=1,m=1,2,…,N BS ,n=1,2,…,N RF This constraint has an impact on F RF All elements of are randomly initialized; Step 6.2, fix F RF Optimize F BB , and obtain F by least squares estimation BB Estimates: Step 6.3, fix F BB Optimize F RF When , the hybrid beamforming design problem is expressed as: s.t.|F RF (m,n)|=1, m=1,2,…,N BS ,n=1,2,…,N RF The problem is solved by Riemannian manifold optimization; Step 6.4, alternately execute step 6.2 and step 6.3 until a termination condition is met, wherein the termination condition is set to: the number of executions reaches a maximum preset number of times, or the change in the objective function of the hybrid beamforming design problem between two consecutive executions is less than a preset threshold value; the solution to the hybrid beamforming design problem obtained after the termination condition is met is and The final designed analog beamforming matrix is Consider the constraints of the total transmission power of the base station The final designed digital beamforming matrix is 2. The transmit beamforming method of the millimeter wave communication and perception integrated system according to claim 1, characterized in that: In step 3, define the matrix F = F RF F BB , and define the i-th column of F as f i ,i=1,2,…,N RF , then the base station's transmit beam is 3. The transmit beamforming method of the millimeter wave communication and perception integrated system according to claim 1, characterized in that: The subproblem 1 in step 5.3 is expressed as a standard second-order cone programming problem and solved using the Optimization Toolbox.

Images