Hybrid beamforming method based on one-bit digital-to-analog converter with integrated sensing and communication

Abstract

The present application relates to a method for hybrid beamforming based on a one-bit digital-to-analog converter for integrated communication. The method comprises: designing an analog precoder and a one-bit digital-to-analog converter quantized signal after a digital precoder based on a trade-off optimization problem between radar and communication performance; designing a transmit beam vector according to the designed analog precoder and the one-bit digital-to-analog converter quantized signal, solving the optimization expression of the transmit beam vector to obtain the optimal transmit beam vector; designing an optimization problem of an analog precoder and a one-bit digital-to-analog converter quantized signal according to the optimal transmit beam vector, solving the optimization problem respectively using the Matlab toolbox and the BTSA‑GA algorithm, and generating an integrated hybrid beam according to the optimized analog precoder and the one-bit digital-to-analog converter quantized signal. The adoption of this method can reduce the complexity and power consumption of large-scale MIMO systems.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing. Background Art

[0002] In recent years, commercial demand for intelligent integrated sensing and communication (ISAC) devices has rapidly increased. With the continuous advancement of radar and wireless communication technologies, and the growing expectations for multifunctionality and intelligence in small embedded devices, ISAC systems have entered the public eye and sparked widespread discussion in industry and academia. Both systems utilize the same platform, share the same hardware and spectrum resources, and form a distinct class of RF technologies. Despite their excellent communication and radar detection performance, ISAC systems based on millimeter-wave and massive MIMO technologies require a large number of hardware components, such as RF chains, power amplifiers, and digital-to-analog converters. This significantly increases the total power consumption and hardware cost of base stations. Pure digital beamforming directly designs quantized transmit signal vectors, mapping one RF chain to each antenna. This approach is suitable for systems with a small number of antennas. However, for massive MIMO systems, hybrid beamforming designs are required to reduce system complexity and power consumption. Summary of the Invention

[0003] Based on this, it is necessary to provide a hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing that can reduce the complexity and power consumption of large-scale MIMO systems to address the above technical problems.

[0004] A hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing and communication, the method comprising:

[0005] The signal sent by the base station is acquired and modeled to obtain a signal model; the signal model includes an analog precoder to be solved; the analog precoder and the one-bit digital-to-analog converter quantizing the signal after the digital precoder are designed based on the trade-off optimization problem between radar and communication performance;

[0006] The transmit beam vector is designed based on the designed analog precoder and the one-bit digital-to-analog converter quantized signal, and the optimized expression of the transmit beam vector is established;

[0007] Solve the optimization expression of the transmit beam vector, expand the cost function of the optimization expression, express the expanded expression in Lagrangian form, and obtain the optimality condition of the expanded expression based on the Lagrangian form; solve the optimality condition to obtain the optimal transmit beam vector;

[0008] The optimization problem of designing analog precoder and one-bit DAC quantization signal according to the optimal transmit beam vector is solved by Matlab toolbox and BTSA-GA algorithm respectively, and the optimized analog precoder and one-bit DAC quantization signal are obtained. The sense-communication integrated hybrid beam is generated based on the optimized analog precoder and one-bit DAC quantization signal.

[0009] The above-mentioned hybrid beamforming method based on a one-bit digital-to-analog converter for integrated sensing and communication differs from the design approach of directly designing a quantized transmit signal vector in digital beamforming schemes. This application designs an analog precoder and a one-bit digital-to-analog converter to quantize the signal after the digital precoder. Given an optimal communication beamformer and a desired radar beam pattern, an analog precoder and a one-bit digital-to-analog converter to quantize the signal after the digital precoder are designed. Under the constraints of system transmit power, the constant modulus constraint of the analog precoder, and the discrete non-convex constraint of the quantized signal, the weighted sum of the communication beamforming error and the radar beamforming error is minimized. This allows the ISAC system to achieve a better peak-to-peak ratio of the radar sensing waveform and superior communication bit error rate performance. Then, based on the optimal transmit beam vector, the optimization problem of the analog precoder and the one-bit digital-to-analog converter quantization signal is designed. The Matlab toolbox and the BTSA-GA algorithm are used to alternately solve the analog precoder under the constant envelope constraint of the phase-shifting network and the one-bit digital-to-analog converter quantization signal after digital precoding. By designing a global optimization algorithm, namely the BTSA-GA algorithm, to solve the discrete constrained optimization problem, the performance loss caused by relaxing the non-convex symbol constraints in some classic nonlinear precoding algorithms is avoided. According to the optimized analog precoder and the one-bit digital-to-analog converter quantization signal, a sensing and communication integrated hybrid beam is generated. The radar detection and perception functions are realized while the base station communicates with multiple downlink users, and the scarce spectrum resources are fully utilized to achieve larger channel capacity, higher spectrum efficiency and better target detection performance, which can greatly reduce the hardware cost, system complexity and power consumption of each RF link. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 1 is a flow chart of a hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing and communication in one embodiment;

[0011] Figure 2 A diagram of a massive MIMO ISAC system model based on a one-bit digital-to-analog converter in one embodiment;

[0012] Figure 3 The figure is a flowchart of the overall process of the algorithm of the present application in one embodiment.

[0013] Figure 4 is a flowchart of a BTSA-GA algorithm in one embodiment;

[0014] Figure 5 A diagram comparing the system reachability and rate of the method of the present application in one embodiment;

[0015] Figure 6 A comparison chart of bit error rates of the method of the present application in one embodiment;

[0016] Figure 7 FIG. 4 is a radar beam pattern diagram of the method of the present application in one embodiment. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0018] In one embodiment, Figure 1 As shown, a hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing is provided, which can be applied to Figure 2 The massive MIMO ISAC system based on a single-bit digital-to-analog converter includes the following steps:

[0019] Step 102: Acquire the signal sent by the base station, model the signal sent by the base station, and obtain a signal model; the signal model includes the analog precoder to be solved; design the analog precoder and the one-bit digital-to-analog converter quantized signal after the digital precoder based on the trade-off optimization problem between radar and communication performance.

[0020] Narrowband massive MIMO system, where dual-function base stations are equipped with N T Transmitting antennas and N r receiving antennas, and N RF The base station sends data streams to downlink users. At the same time, it supports the detection and perception of k targets in the same time-frequency resources.

[0021] set up is the downlink channel matrix from the base station to the user equipment, the received signal at the user end can be modeled as

[0022]

[0023] Where, is the power control parameter, To simulate the precoder, is the digital precoder, is a quantized bit, n is zero mean, and the variance is σ 2 The observation noise consists of independent complex Gaussian components. The data flow vector satisfies Among them, P s is the total power of the unprecoded data, N s ×N s The identity matrix of .

[0024] The mmWave channel is modeled as having N c Scattering clusters and each scattering cluster contains N p The propagation environment of the scattering path. In this model, the channel matrix is ​​as follows

[0025]

[0026] Among them, α lr is the small-scale fading coefficient, which satisfies the standard complex Gaussian distribution φ lr and θ lr They are the angle of arrival (AoA) and angle of departure (AoD) respectively. and is the array steering vector that depends only on the transmit and receive antenna array structures.

[0027]

[0028] A single-bit digital-to-analog converter senses the signal The elements in are:

[0029]

[0030] A typical MIMO radar beamformer can be constructed by combining a H (θ)C Rad a(θ) is matched with the desired transmit beam pattern φ(θ), where C Rad is the covariance matrix of the detection signal. Mathematically, solving C Rad The constrained least squares problem is formulated as:

[0031]

[0032] Among them, θ i is a fine grid of points covering the target direction of arrival (DOA), N is the number of points, and δ is the scale parameter. This is a convex optimization problem that can be solved directly using the Matlab toolbox CVX.

[0033] The trade-off optimization problem between radar and communication performance can be described as follows:

[0034]

[0035] Among them, F FDis the optimal unconstrained precoding matrix, which consists of the first n columns of the right singular matrix of H. ρ∈[0,1] is a weighting factor that balances radar performance and communication performance.

[0036] Step 104 : Design a transmit beam vector based on the designed analog precoder and the one-bit digital-to-analog converter quantized signal, and establish an optimized expression for the transmit beam vector.

[0037] First, design the transmit beam vector The optimization problem of y can be equivalent to:

[0038]

[0039] in,

[0040] The transmit beam vector is designed based on the designed analog precoder and the one-bit digital-to-analog converter quantized signal, and an optimized expression of the transmit beam vector is established to ensure that it can meet the dual requirements of radar and communication.

[0041] Step 106: Solve the optimization expression for the transmit beam vector, expand the cost function of the optimization expression, express the expanded expression in Lagrangian form, and obtain the optimality condition of the expanded expression based on the Lagrangian form; solve the optimality condition to obtain the optimal transmit beam vector.

[0042] Define Q = D H D, G = D H B, further expand the cost function of (8) into:

[0043]

[0044] The optimization problem (9) is a matrix version of the trust region subproblem (TRS) that satisfies strong duality, with a zero duality gap. Its Lagrangian form is expressed as:

[0045]

[0046] where λ is the dual variable associated with the equality constraint. The optimality condition for (9) is

[0047]

[0048]

[0049] in, λ opt is the optimal dual variable, y opt It is the optimal solution to the dual problem and also the optimal solution to the original optimization problem. (12) and (13) guarantee the feasibility of the original problem and the feasibility of the dual problem respectively.

[0050] based on Obviously, It is the matrix M H +M's eigenvalues, which satisfy the corresponding eigenvectors of (12) and (13), are exactly the optimal solution y opt .

[0051] The cost function is expanded into a Lagrangian form to obtain the optimality condition and then solve the optimal transmit beam vector.

[0052] An analog precoder is designed, and a single-bit digital-to-analog converter (DAC) is used after the digital precoder to sense the quantized signal. Given an optimal communication beamformer and a desired radar beam pattern, the analog precoder and the DAC after the digital precoder are designed to sense the quantized signal. Under the constraints of system transmit power, the constant modulus constraint of the analog precoder, and the discrete non-convexity of the quantized signal, the ISAC system minimizes the weighted sum of the communication beamforming error and the radar beamforming error. This allows for a better peak-to-side-error ratio (PSLR) of the radar sensing waveform and superior communication bit error rate performance.

[0053] Step 108, designing the optimization problem of the analog precoder and the one-bit digital-to-analog converter quantization signal based on the optimal transmit beam vector, using the Matlab toolbox and the BTSA-GA algorithm to solve the optimization problem respectively, to obtain the optimized analog precoder and the one-bit digital-to-analog converter quantization signal; generating an integrated sensing and communication hybrid beam based on the optimized analog precoder and the one-bit digital-to-analog converter quantization signal.

[0054] fixed F A The optimization problem can be written as:

[0055]

[0056] The optimization problem (14) is a typical Riemannian manifold optimization problem and can be solved using the Matlab toolbox Manopt. The direction in which the cost function decreases the most at x in the Riemannian manifold is its Riemannian gradient:

[0057]

[0058] The cost function is its Euclidean gradient, described in detail as follows:

[0059]

[0060] Fixed F A , The optimization problem can be written as:

[0061]

[0062] By quantizing the real and imaginary parts by one bit respectively, the above optimization problem can be further rewritten as:

[0063]

[0064] in, Represents the set {1,2,...,2N RF},

[0065]

[0066] The optimization problem (18) is solved by the binary tree species-genetic optimization algorithm (BTSA-GA). The specific process of the algorithm is as follows.

[0067] In the initialization phase of the BTSA-GA algorithm, a batch of trees are generated:

[0068] T i,j =l j +r i,j (h j -l j ), (twenty two)

[0069] Among them, l j is the lower bound of the search space -1, h j is the upper bound of the search space 1, r i,j It is a random number generated for each dimension at each position in the range [0,1].

[0070] T i,j Binary discretization:

[0071]

[0072] Find the current best solution using the following formula:

[0073] S=minf(T i ),i∈1,2,…,2N RF , (twenty four)

[0074]

[0075] Trees generate seeds, and the positions of trees and seeds in the n-dimensional search space can be viewed as possible solutions to the optimization problem. The positions of seeds are obtained through two search equations:

[0076]

[0077] Among them, S i,j is the jth dimension of the i-th seed generated by the i-th tree, Ti,j is the jth dimension of the i-th tree, B j is the jth dimension of the best tree position obtained so far, T r,j is the jth dimension of the rth tree randomly selected from the population, β i,j is a randomly generated scaling factor in the range [-1, 1]. The first equation in (22) focuses on global search to prevent the algorithm from entering a local optimum, while the second equation focuses on local search to facilitate rapid convergence. γ is a random number generated in the range (0, 1) for each position and dimension, and ST is the selected control parameter.

[0078] In order to avoid the tendency of iteration moving to the local optimum, a threshold H is set. i The number of generations of generating seeds exceeds H and there is no better seed to update T i When T is updated through crossover and mutation operations in the genetic algorithm i , expand the search space, and thus reach the global optimal solution.

[0079] According to the optimal transmit beam vector, an optimization problem of designing an analog precoder and a one-bit digital-to-analog converter quantization signal is solved respectively using the Matlab toolbox and the BTSA-GA algorithm to obtain the optimized analog precoder and the one-bit digital-to-analog converter quantization signal. The optimized analog precoder and the one-bit digital-to-analog converter quantization signal can generate a more accurate integrated sensing and communication hybrid beam, thereby providing a larger channel capacity, higher spectrum efficiency and better target detection performance. At the same time, the one-bit digital-to-analog converter sensing is deployed in a large-scale MIMO ISAC system, which greatly reduces the hardware cost, system complexity and power consumption of each RF link. The non-convex subproblem of the one-bit digital-to-analog converter quantization signal is solved. The BTSA-GA algorithm proposed in this application has lower performance loss than the linear quantization precoder.

[0080] In the above-mentioned hybrid beamforming method based on integrated sensing and communication using a single-digital-analog converter, unlike the design concept of directly designing a quantized transmit signal vector in the digital beamforming scheme, this application designs an analog precoder and a single-digital-analog converter to quantize the signal after the digital precoder. Given the optimal communication beamformer and the desired radar beam pattern, the analog precoder and the single-digital-analog converter quantized signal after the digital precoder are designed. Under the constraints of the system transmit power, the constant modulus constraint of the analog precoder, and the discrete non-convex constraint of the quantized signal, the weighted sum of the communication beamforming error and the radar beamforming error is minimized. The ISAC system can achieve a better peak-to-peak ratio of the radar sensing waveform and superior communication bit error rate performance. Then, based on the optimal transmit beam vector, the optimization problem of the analog precoder and the one-bit digital-to-analog converter quantization signal is designed. The Matlab toolbox and the BTSA-GA algorithm are used to alternately solve the analog precoder under the constant envelope constraint of the phase-shifting network and the one-bit digital-to-analog converter quantization signal after digital precoding. By designing a global optimization algorithm, namely the BTSA-GA algorithm, to solve the discrete constrained optimization problem, the performance loss caused by relaxing the non-convex symbol constraints in some classic nonlinear precoding algorithms is avoided. According to the optimized analog precoder and the one-bit digital-to-analog converter quantization signal, a sensing and communication integrated hybrid beam is generated. The radar detection and perception functions are realized while the base station communicates with multiple downlink users, and the scarce spectrum resources are fully utilized to achieve larger channel capacity, higher spectrum efficiency and better target detection performance, which can greatly reduce the hardware cost, system complexity and power consumption of each RF link.

[0081] In one embodiment, an analog precoder and a one-bit digital-to-analog converter quantized signal after the digital precoder are designed based on a trade-off optimization problem between radar and communication performance, including:

[0082] The analog precoder and the one-bit DAC quantized signal after the digital precoder are modeled as the objective function of the trade-off optimization problem between radar and communication performance. The trade-off optimization problem model between radar and communication performance is established under the constraints of system transmit power constraint, constant modulus constraint of analog precoder and discrete non-convex constraint of quantized signal.

[0083]

[0084] Among them, F A represents the analog precoder, represents a one-bit DAC quantized signal, ρ∈[0,1] is a weighting factor that balances radar performance and communication performance, the superscript H represents conjugate transpose, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, represents an element in a one-bit DAC quantized signal, represents the quantized value set of a one-bit digital-to-analog converter, Represents the set {1,2,...,N RF}, N T Represents the number of transmit antennas of the dual-function base station, n represents the column subscript index of the element in the m-th row and n-th column of the analog precoder, and m represents the row subscript index of the element in the m-th row and n-th column of the quasi-precoder.

[0085] In one embodiment, a transmit beam vector is designed based on the designed analog precoder and a one-bit digital-to-analog converter quantized signal, and an optimized expression for the transmit beam vector is established, including:

[0086] According to the designed analog precoder and one-bit DAC quantization signal, the transmit beam vector is designed as follows: The optimal expression for establishing the transmit beam vector is:

[0087]

[0088] in, F A represents the analog precoder, represents a one-bit DAC quantized signal, ρ∈[0,1] is a weighting factor that balances radar performance and communication performance, the superscript H represents the transpose operation, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t It represents the total transmit power of the dual-function base station, and the superscript T represents transposition.

[0089] In one embodiment, the cost function of the optimization expression is expanded, including:

[0090] Define Q = D H D, G = D H B. Expand the cost function of the optimization expression into

[0091]

[0092] Where y represents the transmit beam vector, ρ∈[0,1] is the weighting factor that balances radar performance and communication performance. The superscript H represents the transpose operation. FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t It represents the total transmit power of the dual-function base station, and the superscript T represents transposition.

[0093] In one embodiment, the expanded expression is expressed in Lagrangian form, including:

[0094] The expanded expression is expressed in Lagrangian form as

[0095]

[0096] where λ is the dual variable associated with the equality constraint.

[0097] In one embodiment, obtaining the optimality condition of the expanded expression based on the Lagrangian form includes:

[0098] The optimality condition for the expanded expression based on the Lagrangian form is:

[0099]

[0100] Where Q = D H D, G = D H B, λ opt is the optimal dual variable, y opt is the candidate solution for the optimal transmit beam vector, ρ∈[0,1] is the weighting factor that weighs radar performance and communication performance, the superscript H represents the transpose operation, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t Indicates the total transmit power of the dual-function base station, Indicates dimension is 2N T ×2N T The superscript T indicates the transpose.

[0101] In one embodiment, the optimization problem of designing an analog precoder and a one-bit digital-to-analog converter to quantize a signal according to an optimal transmit beam vector includes:

[0102] The optimization problem of designing the analog precoder based on the optimal transmit beam vector by fixing the one-bit DAC quantization signal is:

[0103]

[0104] Among them, F A represents the analog precoder, represents a one-bit DAC quantized signal, y opt is the candidate solution for the optimal transmit beam vector, N T Indicates the number of transmitting antennas of the dual-function base station, and n represents the analog precoder F A The element in row m and column n [F A ]m,n The column subscript index, m represents the analog precoder F A The element in row m and column n [F A ] m,n The row subscript index of .

[0105] In one embodiment, the analog precoder is fixed, and the optimization problem for designing a one-bit digital-to-analog converter quantized signal based on the optimal transmit beam vector is:

[0106]

[0107] in, represents an element in a one-bit DAC quantized signal, represents the quantized value set of a one-bit digital-to-analog converter, Represents the set {1,2,...,N RF}.

[0108] The optimization problem of the above one-bit DAC quantized signal is further rewritten as follows:

[0109]

[0110] in, Represents the set {1,2,...,2N RF},

[0111]

[0112] It represents the operator for taking the real part of a complex number. Represents the operator for taking the imaginary part of a complex number.

[0113] In one embodiment, the optimization problem is solved using the Matlab toolbox and the BTSA-GA algorithm to obtain an optimized analog precoder and a one-bit digital-to-analog converter quantized signal, including:

[0114] The Matlab toolbox Manopt is used to solve the optimization problem of the analog precoder and obtain the optimized analog precoder.

[0115] In one embodiment, the BTSA-GA algorithm is used to solve the optimization problem of a 1-bit digital-to-analog converter quantized signal. In the initialization phase of the BTSA-GA algorithm, a batch of trees are generated as follows:

[0116] T i,j =l j +r i,j (h j -l j ),

[0117] Among them, l j is the lower bound of the search space -1, h j is the upper bound of the search space 1, r i,j It is a random number generated for each position and each dimension in the range of [0,1];

[0118] T i,j Binary discretization:

[0119]

[0120] Find the current best solution using the following formula:

[0121] S=min f(T i ),i∈1,2,…,2N RF ,

[0122]

[0123] Trees generate seeds, and the positions of trees and seeds in the n-dimensional search space are regarded as possible solutions to the optimization problem. The positions of seeds are obtained through two search equations:

[0124]

[0125] Among them, S i,j is the jth dimension of the i-th seed generated by the i-th tree, T i,j is the jth dimension of the i-th tree, B j is the jth dimension of the best tree position obtained so far, T r,j is the jth dimension of the rth tree randomly selected from the population, β i,j is a randomly generated scaling factor in the range [-1, 1], γ is a random number generated for each position and dimension in the range (0, 1), and ST is a selection control parameter;

[0126] Set the threshold H, when T i The number of generations of generating seeds exceeds H and there is no better seed to update T i When T is updated through crossover and mutation operations in the genetic algorithm i , and obtain the global optimal solution, which is the optimized one-bit DAC quantized signal.

[0127] In a specific embodiment, combining Figure 3 As shown, based on the embodiment of the present application, a digital-to-analog converter sensing-communication integrated hybrid beamforming method is proposed in this application. A two-variable alternating minimization algorithm is first designed to transmit the beam y of the MIMO ISAC receiver. opt Then, according to the designed transmit beam, alternately solve the analog precoder F under the constant envelope constraint of the phase shift networkA and 1-bit quantized signal after digital precoding because The solution to the quantized signal after the digital precoder is obtained by exhaustive search Second, when there are many RF chains, the solution process is too complicated. To solve this problem, this application designs a global optimization algorithm, namely the improved binary tree genetic optimization algorithm (BTSA-GA) algorithm to solve the discrete constraint optimization problem. The algorithm flow chart is as follows: Figure 4 shown.

[0128] like Figure 5 As shown in Figure 2, this application compares the achievable communication and rate of the proposed one-bit hybrid beamforming design method with other linear one-bit digital beamforming design methods based on Monte Carlo experiments. Consider a medium-sized 32×4 MIMO system. Figure 5 It can be clearly seen that as the signal-to-noise ratio increases, the achievable communication sum rate also increases accordingly. Among the precoding methods, the zero forcing algorithm (ZF) is a linear precoding method with lower complexity. Obviously, the achievable communication sum rate of the nonlinear precoder with a one-bit digital-to-analog converter designed in this application is better than the linear precoding method, and as the signal-to-noise ratio increases, the gap in achievable sum rate also increases accordingly. When the signal-to-noise ratio is 10dB, the achievable communication sum rate of the nonlinear precoder is almost twice that of the linear precoding method. The algorithm proposed in this application has a achievable communication sum rate performance close to that of the nonlinear digital beamforming design method, and has significant advantages in achievable sum rate performance compared to the linear digital beamforming design method.

[0129] Figure 6 This is a comparison chart of the bit error rate performance of the one-bit hybrid beamforming design method proposed in this application and the bit error rate of other precoding schemes. Compared with the one-bit linear digital beamforming design scheme, such as ZF (zero forcing)-1bit, the one-bit hybrid beamforming design method proposed in this application has significant advantages in bit error rate performance.

[0130] In order to evaluate the radar detection and sensing performance, this application sets three target directions of interest, and the beam tracking angles are [θ1 = -30°, θ2 = 0°, θ1 = 30°]. The radar beam pattern is normalized and measured in dB. Figure 7 As shown, the three peaks correspond to beams pointing in the directions of three targets of interest, indicating that the sensory-communication integrated hybrid beamforming method proposed in this application meets the detection and perception functions of the radar. Figure 7 It indicates that the quantization error makes the beam pattern quality slightly worse based on a one-bit DAC, and the performance loss of less than 1 dB is tolerable.

[0131] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0132] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A hybrid beamforming method based on a one-bit digital-to-analog converter and integrated sensing, the method being applied to a narrowband massive MIMO system including a dual-function base station, characterized in that: The method comprises: Acquire a signal sent by a base station, model the signal sent by the base station, and obtain a signal model; the signal model includes an analog precoder to be solved; design the analog precoder and a one-bit digital-to-analog converter quantized signal after the digital precoder based on a trade-off optimization problem between radar and communication performance; designing a transmit beam vector based on the designed analog precoder and a one-bit digital-to-analog converter quantized signal, and establishing an optimized expression for the transmit beam vector; Solving an optimization expression for the transmit beam vector, expanding a cost function of the optimization expression, expressing the expanded expression in Lagrangian form, and obtaining an optimality condition for the expanded expression based on the Lagrangian form; solving the optimality condition to obtain an optimal transmit beam vector; Designing an optimization problem for the analog precoder and the one-bit digital-to-analog converter quantized signal based on the optimal transmit beam vector, alternately solving the optimization problem using a Matlab toolbox and a BTSA-GA algorithm, respectively, to obtain optimized analog precoders and one-bit digital-to-analog converter quantized signals, namely, analog precoders under constant envelope constraints of a phase shift network and one-bit digital-to-analog converter quantized signals after digital precoding; Generate a sensory-communication integrated hybrid beam based on the optimized analog precoder and a one-bit digital-to-analog converter quantized signal; The method further comprises: The BTSA-GA algorithm is used to solve the optimization problem of the quantized signal of the one-bit digital-to-analog converter. In the initialization stage of the BTSA-GA algorithm, a batch of trees are generated as follows: T i,j =l j +r i,j (h j -l j ), Among them, l j is the lower bound of the search space -1, h j is the upper bound of the search space 1, r i,j It is a random number generated for each position and each dimension in the range of [0,1]; T i,j Binary discretization: Find the current best solution using the following formula: S=min f(T i ),i∈1,2,…,2N RF , Trees generate seeds, and the positions of trees and seeds in the n-dimensional search space are regarded as possible solutions to the optimization problem. The positions of seeds are obtained through two search equations: Among them, S i,j is the jth dimension of the i-th seed generated by the i-th tree, T i,j is the jth dimension of the i-th tree, B j is the jth dimension of the best tree position obtained so far, T r,j is the jth dimension of the rth tree randomly selected from the population, β i,j is a randomly generated scaling factor in the range [-1, 1], γ is a random number generated for each position and dimension in the range (0, 1), and ST is a selection control parameter; Set the threshold H, when T i The number of generations of generating seeds exceeds H and there is no better seed to update T i When T is updated through crossover and mutation operations in the genetic algorithm i , and obtain the global optimal solution, which is the optimized one-bit DAC quantized signal.

2. The method according to claim 1, characterized in that The analog precoder and the one-bit digital-to-analog converter quantized signal after the digital precoder are designed based on the trade-off optimization problem between radar and communication performance, including: The analog precoder and the one-bit DAC quantized signal after the digital precoder are modeled as the objective function of the trade-off optimization problem between radar and communication performance. The trade-off optimization problem model between radar and communication performance is established with the system transmit power constraint, the constant modulus constraint of the analog precoder, and the discrete non-convex constraint of the quantized signal as the constraint conditions. Among them, F A represents the analog precoder, represents a one-bit DAC quantized signal, ρ∈[0,1] is a weighting factor that balances radar performance and communication performance, the superscript H represents conjugate transpose, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t Indicates the total transmit power of the dual-function base station, represents an element in a one-bit DAC quantized signal, represents a one-bit DAC quantization set, Represents the set {1,2,...,N RF }, N RF Indicates the number of RF chains of the dual-function base station, N T Indicates the number of transmitting antennas of the dual-function base station, and n represents the analog precoder F A The element in row m and column n [F A ] m,n The column subscript index, m represents the analog precoder F A The element in row m and column n [F A ] m,n The row subscript index of .

3. The method according to claim 1, characterized in that A transmit beam vector is designed based on the designed analog precoder and a one-bit digital-to-analog converter quantized signal, and an optimized expression for the transmit beam vector is established, including: According to the designed analog precoder and one-bit DAC quantization signal, the transmit beam vector is designed as follows: The optimized expression for the transmit beam vector is established as in, F A represents the analog precoder, represents a one-bit DAC quantized signal, ρ∈[0,1] is a weighting factor that balances radar performance and communication performance, the superscript H represents the transpose operation, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t It represents the total transmit power of the dual-function base station, and the superscript T represents transposition.

4. The method according to claim 1, wherein Expand the cost function of the optimization expression, including: Define Q = D H D, G = D H B. Expand the cost function of the optimization expression into Where y represents the transmit beam vector, is a weighting factor that balances radar performance and communication performance. The superscript H represents the transpose operation. FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t It represents the total transmit power of the dual-function base station, and the superscript T represents transposition.

5. The method according to claim 4, characterized in that The expanded expression is expressed in Lagrangian form, including: The expanded expression is expressed in Lagrangian form as where λ is the dual variable associated with the equality constraint.

6. The method according to claim 1, characterized in that The optimality conditions of the expanded expression are obtained based on the Lagrangian form, including: The optimality condition for the expanded expression based on the Lagrangian form is: Where Q = D H D, G = D H B, λ opt is the optimal dual variable, y opt is the candidate solution for the optimal transmit beam vector, ρ∈[0,1] is the weighting factor that weighs radar performance and communication performance, the superscript H represents the transpose operation, and F FD is the optimal unconstrained precoding matrix, C Rad represents the radar beam covariance matrix, P t Indicates the total transmit power of the dual-function base station, Indicates dimension is 2N T ×2N T The superscript T indicates the transpose.

7. The method according to claim 1, characterized in that The optimization problem of designing the analog precoder and the one-bit digital-to-analog converter quantized signal according to the optimal transmit beam vector includes: The optimization problem of designing the analog precoder based on the optimal transmit beam vector by fixing a one-bit digital-to-analog converter quantized signal is: Among them, F A represents the analog precoder, represents a one-bit DAC quantized signal, y opt is the candidate solution for the optimal transmit beam vector, N T represents the number of transmitting antennas of the dual-function base station, n represents the column subscript index of the element in the m-th row and n-th column of the analog precoder, and m represents the row subscript index of the element in the m-th row and n-th column of the analog precoder.

8. The method according to claim 7, characterized in that The method further comprises: The optimization problem of designing a one-bit digital-to-analog converter to quantize the signal based on the optimal transmit beam vector with a fixed analog precoder is: in, represents an element in a one-bit DAC quantized signal, represents a set of quantized values ​​of a one-bit digital-to-analog converter, Represents the set {1,2,...,N RF }; The optimization problem of the above one-bit DAC quantized signal is further rewritten as follows: in, Represents the set {1,2,...,2N RF }, It represents the operator for taking the real part of a complex number. Represents the operator for taking the imaginary part of a complex number.

9. The method according to claim 1, characterized in that The optimization problem is solved using the Matlab toolbox and the BTSA-GA algorithm to obtain an optimized analog precoder and a one-bit digital-to-analog converter quantized signal, including: The optimization problem of the analog precoder is solved using the Matlab toolbox Manopt to obtain an optimized analog precoder.

Images