Beam forming method of 5G multiple-input multiple-output technology based on geometric prior

Abstract

The invention provides a beam forming method of a 5G multiple-input multiple-output technology based on geometric priori. The beam forming method comprises the following steps: acquiring an angle domain channel matrix of a wireless channel between two multiple-input multiple-output arrays; according to the relative orientation included angle of the array, determining an area where non-zero elements may appear in the angle domain channel matrix by using geometric prior information; an orthogonal matching pursuit algorithm is adopted to search an analog beam forming matrix of the transmitting end and an analog beam forming matrix of the receiving end in an area where non-zero elements may appear in the angle domain channel matrix; and according to the analog beamforming matrix of the transmitting end, the analog beamforming matrix of the receiving end and a known channel matrix, solving a digital beamforming matrix of the transmitting end and a digital beamforming matrix of the receiving end through a mathematical optimization method to obtain a shaped target beam. The calculation complexity of beam forming is reduced.

Description

Technical Field

[0001] The present invention relates to the field of 5G communication and beamforming technology, and particularly to a beamforming method for 5G multiple-input multiple-output technology based on geometric prior knowledge. Background Art

[0002] Multiple-input multiple-output (MIMO) technology is a key technology for realizing high-capacity, high-rate, and low-power wireless transmission in 5G, and has been incorporated into international 5G technical standards such as 3GPP TS 38.211, 3GPP TS 38.213, 3GPP TS 38.300, and 3GPP TS 38.321. In order to fully utilize the array gain of MIMO in 5G scenarios, a low-complexity robust beamforming algorithm is crucial. Although traditional 5G MIMO beamforming algorithms have been quite mature, these algorithms still cannot cope with the geometrically growing scale of two-way MIMO channels. Specifically, traditional MIMO beamforming is usually implemented digitally at the baseband, and high array gain is achieved by flexibly adjusting the phase and amplitude of the electromagnetic waves radiated by each antenna. However, such a digital baseband processing scheme requires a dedicated variable-gain radio frequency amplifier for each antenna. For small-scale MIMO (the number of antennas for small-scale MIMO is specified as 2 to 8 according to the 3GPP TS 38.101 standard), this hardware cost is relatively acceptable; but for a slightly larger MIMO array (especially a two-dimensional uniform planar array), according to the 3GPP TS 38.300 standard, the number of its antennas may be as high as 64, 128, or even 256, which results in too high a hardware cost for equipping each antenna with a variable-gain radio frequency amplifier, thus severely disabling traditional beamforming algorithms. In addition, a large number of variable-gain radio frequency amplifiers will cause a sharp increase in system power consumption, unable to meet the low-power requirements of 5G communication systems. Therefore, wireless communication systems equipped with MIMO arrays usually adopt an analog-digital hybrid method for beamforming. Specifically, in the hardware deployment of an actual MIMO array, each antenna element is connected to each radio frequency link channel through a phase shifter. Without directly adjusting the amplitude of the electromagnetic waves radiated by each antenna element, the radiation pattern of the array is changed by adjusting the phase, thereby significantly saving hardware costs and reducing system power consumption. Mathematically speaking, the goal of the MIMO beamforming algorithm is to derive the optimal phase shifts of all phase shifters at the transceiver ends and the baseband precoding matrix of the radio frequency links based on the known channel matrix to maximize the channel capacity. Given that hybrid beamforming MIMO has entered multiple 3GPP standards and has been widely deployed in 5G communication systems, designing a beamforming algorithm with lower computational complexity has become the key to further optimizing 5G communication systems.

[0003] At present, a large amount of work has been carried out on the mixed analog-digital beamforming algorithms for MIMO. These studies generally use the Angular Domain (AD) codebook for beamforming, that is, the discrete Fourier transform matrix is used as the dictionary matrix. Specifically, according to the 3D-SV channel model of 5G MIMO specified in 3GPP TS 38.815, the channel between the MIMO transceiver arrays can be decomposed into the superposition of several plane wave propagation components in the angular domain (i.e., performing two-dimensional discrete Fourier transform on the channel matrix). Therefore, using the discrete Fourier transform matrix as the dictionary matrix for beamforming naturally fits the wireless propagation environment described by the 3D-SV channel model. However, the existing methods for beamforming using the angular domain codebook need to search the entire angular domain codebook and compare it with the channel, resulting in a high computational complexity. Summary of the Invention

[0004] The present invention provides a beamforming method for 5G multiple-input multiple-output technology based on geometric prior, which is used to solve the defect of high computational complexity of beamforming using the angular domain codebook in the prior art, and to reduce the computational complexity of beamforming.

[0005] The present invention provides a beamforming method for 5G multiple-input multiple-output technology based on geometric prior, including: Obtain the angular domain channel matrix of the wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing two-dimensional discrete Fourier transform on the initial channel matrix; According to the relative orientation angle between the two multiple-input multiple-output arrays, use the geometric prior information to determine the region where non-zero elements in the angular domain channel matrix may appear, and the geometric prior information is the mapping relationship between the wave departure angle at the transmitter and the wave arrival angle at the receiver combined with the lobe width of the array beamforming; Use the orthogonal matching pursuit algorithm to search for the analog beamforming matrix at the transmitter and the analog beamforming matrix at the receiver in the region where non-zero elements in the angular domain channel matrix may appear; According to the analog beamforming matrix at the transmitter, the analog beamforming matrix at the receiver, and the known channel matrix, solve the digital beamforming matrix at the transmitter and the digital beamforming matrix at the receiver through a mathematical optimization method to obtain the target beam after beamforming.

[0006] In a possible implementation manner, the method further includes: Determine the closed-form expression of the geometric prior information according to the relative orientation angle between the two multiple-input multiple-output arrays; Determine the region where non-zero elements in the angular domain channel matrix may appear according to the closed-form expression.

[0007] In a possible implementation, the method further includes: Initialize the residual matrix; In each round of loop, calculate the squared modulus of the product of the residual matrix and each column of the dictionary matrix at the transmitter, and select the column with the largest squared modulus of the product to be added to the set of used codewords; Update the residual matrix and the analog beamforming matrix at the transmitter based on the existing set of codewords until the number of used codewords reaches the number of radio frequency links at the transmitter.

[0008] In a possible implementation, the method further includes: According to the analog beamforming matrix at the transmitter, the analog beamforming matrix at the receiver, and the known channel matrix, use a convex optimization algorithm to estimate the angle between the normals of the two multiple-input multiple-output arrays; Based on the angle between the normals of the two multiple-input multiple-output arrays, optimize and solve the digital beamforming matrix at the transmitter and the digital beamforming matrix at the receiver to obtain the shaped target beam.

[0009] In a possible implementation, the method further includes: Define a residual function according to the known pilot signal, the beamforming matrix, the noise, and the transmit power; Use a convex optimization method to solve for the angle value that minimizes the residual function, and use the angle value that minimizes the residual function as the angle between the normals of the two multiple-input multiple-output arrays.

[0010] In a possible implementation, the method further includes: When searching for the angle of arrival of the wave at the receiver, use the geometric prior information to narrow the search range and calculate the squared modulus of the product of the rows that meet specific conditions and the residual matrix.

[0011] The present invention also provides a beamforming device for 5G multiple-input multiple-output technology based on geometric prior, including the following modules: An acquisition module, configured to acquire an angular domain channel matrix of the wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on the initial channel matrix; A determination module, configured to determine, according to the relative orientation angle between the two multiple-input multiple-output arrays, a region where non-zero elements in the angular domain channel matrix may appear by using geometric prior information, where the geometric prior information is the mapping relationship between the wave departure angle at the transmitter and the wave arrival angle at the receiver combined with the lobe width of the array beamforming; A search module, configured to search for the analog beamforming matrix at the transmitter and the analog beamforming matrix at the receiver in the region where non-zero elements may appear in the angular domain channel matrix by using the orthogonal matching pursuit algorithm; A beamforming module, configured to solve the digital beamforming matrix at the transmitter and the digital beamforming matrix at the receiver by a mathematical optimization method according to the analog beamforming matrix at the transmitter, the analog beamforming matrix at the receiver, and the known channel matrix, so as to obtain a shaped target beam.

[0012] The present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of the above is implemented.

[0013] The present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of the above is implemented.

[0014] The present invention further provides a computer program product, including a computer program. When the computer program is executed by a processor, the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of the above is implemented.

[0015] The beamforming method of the 5G multiple-input multiple-output technology based on geometric prior provided by the present invention includes: obtaining an angular domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; determining, according to a relative orientation angle between the two multiple-input multiple-output arrays, a region where non-zero elements may appear in the angular domain channel matrix by using geometric prior information, where the geometric prior information is a mapping relationship between a wave departure angle at a transmitter and a wave arrival angle at a receiver combined with a lobe width of array beamforming; searching for an analog beamforming matrix at the transmitter and an analog beamforming matrix at the receiver in the region where non-zero elements may appear in the angular domain channel matrix by using the orthogonal matching pursuit algorithm; and solving a digital beamforming matrix at the transmitter and a digital beamforming matrix at the receiver by a mathematical optimization method according to the analog beamforming matrix at the transmitter, the analog beamforming matrix at the receiver, and the known channel matrix, so as to obtain a shaped target beam. Compared with the defect that the beamforming calculation complexity is high by using an angular domain codebook in the prior art, in this method, beamforming is performed by using geometric prior in a wireless communication propagation path, so as to reduce the beamforming calculation complexity. Description of the Drawings

[0016] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0017] Figure 1 It is one of the schematic flowcharts of the beamforming method for 5G multiple-input multiple-output technology based on geometric prior provided by the present invention.

[0018] Figure 2 It is the second schematic flowchart of the beamforming method for 5G multiple-input multiple-output technology based on geometric prior provided by the present invention.

[0019] Figure 3 It is the schematic diagram of wireless transmission between 5G multiple-input multiple-output arrays provided by the present invention.

[0020] Figure 4 It is the normal angle Schematic diagram of the possible area where non-zero elements of the angle domain determined by geometric prior at degrees may appear.

[0021] Figure 5 It is the normal angle Schematic diagram of the possible area where non-zero elements of the angle domain determined by geometric prior at degrees may appear.

[0022] Figure 6 It is the normal angle Schematic diagram of the possible area where non-zero elements of the angle domain determined by geometric prior at degrees may appear.

[0023] Figure 7 It is the schematic flowchart of estimating the normal angle of the array provided by the present invention.

[0024] Figure 8 It is the schematic structural diagram of the beamforming device for 5G multiple-input multiple-output technology based on geometric prior provided by the present invention.

[0025] Figure 9 It is the schematic structural diagram of the electronic device provided by the present invention. Detailed implementation manners

[0026] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0027] For ease of understanding of the embodiments of the present invention, the following will further explain and illustrate with specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation to the embodiments of the present invention.

[0028] Figure 1 is one of the schematic flowcharts of the beamforming method for 5G multiple-input multiple-output technology based on geometric prior provided by the present invention. As Figure 1 shown, the method includes the following: S11. Obtain the angular domain channel matrix of the wireless channel between two multiple-input multiple-output arrays.

[0029] The embodiments of the present invention will be described in detail in combination with Figure 2 wherein the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on the initial channel matrix.

[0030] Based on the angular domain codebook that conforms to the 3GPP TS 38.815 standard, the embodiments of the present invention propose an algorithm for beamforming using the geometric prior in the wireless communication propagation path. Specifically, in the actual wireless transceiver scenario, the propagation direction of each plane wave component remains almost unchanged during the wireless propagation process. This means that when constructing a wireless communication link with two multiple-input multiple-output (MIMO) arrays, there is a specific mapping relationship between the angle of departure (AoD) of the wave from the transmitting MIMO array and the angle of arrival (AoA) of the wave at the receiving MIMO array. The AoA-AoD mapping relationship provided by the included angle between the relative orientations of the two arrays is called geometric prior.

[0031] Given the included angle between the relative orientations of two MIMO arrays, by utilizing the mapping relationship between AoD and AoA and combining with the beamwidth (BW) of array beamforming, the AoA-AoD combinations for wireless transmission can be effectively restricted within a smaller range. By using geometric prior, during the beamforming process, there is no need to search the entire angular domain codebook, and only a small number of AoA-AoD combinations need to be searched to complete fast beamforming, thereby significantly reducing the computational complexity. At the same time, considering that the included angle between the relative orientations of two MIMO arrays is usually unknown, the embodiment of the present invention proposes a pilot-based method for estimating the relative orientation of the arrays to ensure the smooth implementation of the proposed beamforming algorithm.

[0032] First, as Figure 3 shown in the schematic diagram of wireless transmission between 5G MIMO arrays, consider the wireless communication between two MIMO arrays in a 5G system. This communication scenario is defined in 5G international standards such as 3GPP TS 38.211, 3GPP TS 38.300, and 3GPP TS 38.901. The communication between two MIMO arrays uses time division duplex mode (3GPP TS 38.211 standard), so the two-way channel has symmetry. Without loss of generality, consider the case where the upper MIMO array is the transmitter and the lower MIMO array is the receiver. The MIMO array at the transmitter is a uniform linear array (ULA) with a size of . The MIMO array at the receiver is a uniform linear array with a size of . Let the carrier frequency be , the carrier wavelength be , and the propagation constant be . The antenna spacing of the MIMO array is half wavelength (3GPP TS 38.211 standard). The included angle between the normals of the two ULAs is . The MIMO array at the transmitter is equipped with radio frequency links. The MIMO array at the receiver is equipped with radio frequency links.

[0033] Let be the channel matrix, then the baseband signal model is: where is the received baseband signal. is the transmit power. is the receive digital beamforming matrix, which is implemented by baseband signal processing. is the receive analog beamforming matrix, which is implemented by phase shifters. For transmitting the analog beamforming matrix, it is implemented by phase shifters. For transmitting the digital beamforming matrix, it is implemented by baseband signal processing. For the transmitted baseband signal, it satisfies . Is the additive white Gaussian noise, , where is the noise power.

[0034] Channel model: According to the 3GPP TR 38.901 report, in the frequency bands of Sub-6 GHz and millimeter wave, the path gain of line-of-sight propagation is 15 - 30 dB higher than that of non-line-of-sight propagation. In this case, non-line-of-sight propagation can be naturally ignored. Assume that the coordinates of the th antenna element at the transmitter are , and the coordinates of the th antenna element at the receiver are . In the case of line-of-sight communication between two MIMO arrays, the th element of the channel matrix can be given by the scalar Green's function, that is where, is the path gain, and its typical values in different propagation environments are elaborated in detail in the 3GPP TR 38.901 report.

[0035] Channel capacity: When the baseband signals at the transmitter are independent of each other and both follow Gaussian distribution, the spectral efficiency (SE) between the baseband signals at the transmitter and receiver is In the 5G system, the channel bandwidth is usually preset. For example, the 5G channel bandwidth specified in the 3GPP TS 38.104 standard has three configurations: 100 MHz, 200 MHz, and 400 MHz. Which channel bandwidth to adopt specifically depends on the requirements of the operator. Since the channel capacity is equal to the spectral efficiency multiplied by the channel bandwidth, it can be considered that the channel capacity is proportional to the spectral efficiency. Therefore, maximizing the channel capacity of MIMO is equivalent to maximizing the spectral efficiency of MIMO.

[0036] Furthermore, the closed-form expression of the geometric prior adopts the angle-domain channel representation method that conforms to the 3GPP TS 38.815 standard, that is, performing a two-dimensional discrete Fourier transform on the initial channel matrix to obtain the angle-domain channel matrix. Specifically, the angle-domain channel matrix has the following relationship with the initial channel matrix Among them, and are the Fourier transform matrices of the receiving end and the sending end respectively, that is Specifically, The th column represents the plane wave component with the propagation direction of .

[0037] Among them, is the length of the receiving - end MIMO array. The AoA corresponding to this propagation direction is Similarly, The th column represents the plane wave component with the propagation direction of . The AoD corresponding to this propagation direction is Among them, is the length of the sending - end MIMO array. According to the relationship between the AoA, AoD and the normal angle in the propagation path, the mapping relationship between AoA and AoD can be obtained under the given normal angle Then substituting the receiving propagation direction and the transmitting propagation direction can further obtain Let and , then there is a mapping relationship Substituting the propagation direction represented by the th column of and the propagation direction represented by the th column of into the above formula, we get This formula gives the mapping relationship between . According to the physical meaning of the angle - domain channel, when the beam width of the array is infinitely small, the non - zero elements of the angle - domain channel matrix in the th column must appear in the th row or the However, the directivity of the actual array is limited, meaning a beamwidth greater than 0. The beamwidth can be defined by the 3 dB beamwidth, half-power beamwidth, or first null beamwidth, mainly depending on the array characteristics (such as the number of antennas, antenna spacing, and specific structure of the antenna elements), which is a pre-known constant. Let the beamwidth of the array be (unit: rad / m). In this way, if the th element in the angle-domain channel matrix is non-zero, then needs to satisfy Given the angle between the normals of two arrays, the geometric figure of the combinations that satisfy the above equation is approximately an elliptical ring, as shown by the yellow area in Figure 2 . Understanding from the physical meaning of the angle-domain channel matrix, during the wireless propagation between MIMOs, only the transmit-receive angle-domain codeword combinations that satisfy this equation can effectively transmit information. This characteristic is called the geometric prior in MIMO communication. As shown in Figure 4 , 5 , and 6, for degrees, degrees, and degrees, are the regions where non-zero elements determined by the geometric prior may appear in the angle domain.

[0038] S12. Determine the region where non-zero elements in the angle-domain channel matrix may appear according to the relative orientation angle between the two multiple-input multiple-output arrays by using the geometric prior information.

[0039] The geometric prior information is the mapping relationship between the wave departure angle at the transmitter and the wave arrival angle at the receiver combined with the lobe width of the array beamforming.

[0040] In order to utilize the geometric prior in the 5G MIMO wireless propagation process, the angle between the normals of two MIMO arrays needs to be known in advance. As shown in the process flow of the pilot-based array normal angle estimation method in Figure 7 , the pilot signal used in this method can be shared with the pilot signal for channel estimation, so there will be no additional pilot overhead. Specifically, assume that the th baseband pilot signal transmitted by the transmitter MIMO is , the th baseband pilot signal received by the receiver MIMO is , and the corresponding noise is . Then the relationship between and is Since the estimation of the angle between the array normal and the channel estimation share the pilot sequence, the channel matrix is unknown. Assume that the length of the pilot signal is . Combine all pilot signals into and . It is necessary to estimate the angle of the array normal according to the known pilot signals and , the beamforming matrix , , and , as well as the transmit power . For simplicity, denote and . At the same time, define the subscript set as Then, define the residual function as where is the Frobenius norm. represents 's th column. represents 's rd row. Since is a convex function, its minimum value can be solved by traditional convex optimization methods, and the specific process will not be elaborated here. After that, set the frequency resolution , and search in the range of 0 to 180 degrees to make the smallest .

[0041] S13. Use the orthogonal matching pursuit algorithm to search for the analog beamforming matrix at the transmitter and the analog beamforming matrix at the receiver in the region where non-zero elements of the channel matrix in the angle domain may appear.

[0042] S14. According to the analog beamforming matrix at the transmitter and the analog beamforming matrix at the receiver and the known channel matrix, solve the digital beamforming matrix at the transmitter and the digital beamforming matrix at the receiver through mathematical optimization methods to obtain the shaped target beam.

[0043] Optimal beamforming of the MIMO system without hardware limitations: According to the 3GPP TS 36.211 standard, the optimal beamforming of the MIMO system without hardware limitations is given by the SVD decomposition of the channel matrix. Specifically, for the channel matrix The result of performing SVD decomposition is where and are both unitary matrices. is a diagonal matrix, and its diagonal elements are arranged in descending order from the upper left to the lower right. Then, the optimal beamforming matrix at the transmitter is is the result after energy normalization of the matrix composed of the first rows of where energy normalization means multiplying by a positive real number such that Similarly, the optimal beamforming matrix at the receiver is the result of normalizing the matrix composed of the first columns of That is, multiplying and by a positive real number such that

[0044] However, According to the 3GPP TS 38.214 standard, in an actual MIMO system, in the of the overall beamforming matrix at the transmitter and in the of the overall beamforming matrix at the receiver are usually implemented by phase shifters to save hardware costs and reduce system power consumption. That is, This constraint results in and not being able to take arbitrary matrices. To obtain a beamforming matrix close to the optimal one, it is necessary to approximate the optimal beamforming matrix with the actual hardware structure, that is, to solve the following two mathematical optimization problems The solutions to the above two optimization problems are the MIMO beamforming methods. Next, an efficient algorithm for jointly solving these two optimization problems is proposed using geometric priors.

[0045] 5G MIMO Fast Beamforming Algorithm Based on Geometric Priors In existing work, the way to solve the two optimization problems P1 and P2 is to use the orthogonal matching pursuit algorithm to search in the entire angular domain codebook. Different from the existing work, in the embodiments of the present invention, by using geometric priors, after solving one of P1 and P2, the codebook search range for solving the other problem can be significantly reduced, thereby significantly reducing the computational complexity of solving the other problem. We first solve P1 as in the existing work. Specifically, let be the residual matrix, initialized as . In each round of loop, calculate multiplied by for the modulus square of each column, and capture the column with the largest modulus square and add it to the set of used codewords. Then, update and based on the existing set of codewords. If the number of used codewords has reached the number of transmitter RF links at this time, stop the loop, otherwise enter the next round of loop. After stopping the loop, the set of used codewords is , because each column of the discrete Fourier transform matrix satisfies the constant modulus constraint.

[0046] In the above process of solving P1, the main running time is consumed in the process of "calculating the modulus square of each column of multiplied by and capturing the column with the largest modulus square". If the search range can be reduced during the process of solving P2, that is, only the modulus square of the product of the residual and a small number of codewords in the codebook needs to be calculated, the computational complexity can be significantly reduced. Specifically, assume that the column numbers of the transmitter dictionary matrix captured after solving P1 are , according to the geometric prior, the transmission directions corresponding to these numbers can only be received in the receiving directions that satisfy the following formula numbers (where ) Therefore, in the process of solving P2, there is no need to calculate the modulus square of each row of multiplied by , but only need to calculate the modulus square of the rows of satisfying the above formula label multiplied by . In this way, the computational complexity of solving P2 can be significantly reduced. Considering P1 and P2 comprehensively, the overall computational complexity of solving , and can be reduced by about 40% compared with the existing scheme.

[0047] The beamforming method based on geometric prior for 5G multiple-input multiple-output (MIMO) technology provided by the present invention obtains the angular domain channel matrix of the wireless channel between two MIMO arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on the initial channel matrix; according to the relative orientation angle between the two MIMO arrays, uses geometric prior information to determine the region where non-zero elements in the angular domain channel matrix may appear, and the geometric prior information is the mapping relationship between the wave departure angle at the transmitter and the wave arrival angle at the receiver combined with the lobe width of array beamforming; uses the orthogonal matching pursuit algorithm to search for the analog beamforming matrix at the transmitter and the analog beamforming matrix at the receiver in the region where non-zero elements in the angular domain channel matrix may appear; according to the analog beamforming matrix at the transmitter, the analog beamforming matrix at the receiver, and the known channel matrix, solves the digital beamforming matrix at the transmitter and the digital beamforming matrix at the receiver through a mathematical optimization method to obtain the target beam after beamforming. Compared with the defect of high computational complexity of beamforming using an angular domain codebook in the prior art, by this method, beamforming is performed using geometric prior in the wireless communication propagation path, achieving a reduction in the computational complexity of beamforming.

[0048] The beamforming device based on geometric prior for 5G MIMO technology provided by the present invention will be described below. The beamforming device based on geometric prior for 5G MIMO technology described below can be correspondingly referred to the beamforming method based on geometric prior for 5G MIMO technology described above.

[0049] Figure 8 FIG. is a schematic structural diagram of the beamforming device based on geometric prior for 5G MIMO technology provided by the present invention, specifically including: An acquisition module 801, configured to acquire the angular domain channel matrix of the wireless channel between two MIMO arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on the initial channel matrix. For detailed description, refer to the relevant description corresponding to the above method embodiment, and details will not be repeated here.

[0050] A determination module 802, configured to determine the region where non-zero elements in the angular domain channel matrix may appear according to the relative orientation angle between the two MIMO arrays, using geometric prior information, where the geometric prior information is the mapping relationship between the wave departure angle at the transmitter and the wave arrival angle at the receiver combined with the lobe width of array beamforming. For detailed description, refer to the relevant description corresponding to the above method embodiment, and details will not be repeated here.

[0051] A search module 803 is configured to search for the analog beamforming matrix at the transmitting end and the analog beamforming matrix at the receiving end in the region where non-zero elements may appear in the angular domain channel matrix by using the orthogonal matching pursuit algorithm. For detailed description, please refer to the relevant description corresponding to the above method embodiment, which will not be elaborated here.

[0052] A beamforming module 804 is configured to solve the digital beamforming matrix at the transmitting end and the digital beamforming matrix at the receiving end by a mathematical optimization method according to the analog beamforming matrix at the transmitting end, the analog beamforming matrix at the receiving end, and the known channel matrix, so as to obtain the shaped target beam. For detailed description, please refer to the relevant description corresponding to the above method embodiment, which will not be elaborated here.

[0053] Figure 9 An example of a schematic physical structure diagram of an electronic device is shown as Figure 9 As shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940. Among them, the processor 910, the communication interface 920, and the memory 930 complete communication with each other through the communication bus 940. The processor 910 may call the logical instructions in the memory 930 to execute a beamforming method for 5G multiple-input multiple-output technology based on geometric prior, and the method includes: obtaining an angular domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; determining, according to the relative orientation angle between the two multiple-input multiple-output arrays, a region where non-zero elements may appear in the angular domain channel matrix by using geometric prior information, where the geometric prior information is a mapping relationship between the wave departure angle at the transmitting end and the wave arrival angle at the receiving end combined with the lobe width of array beamforming; searching for the analog beamforming matrix at the transmitting end and the analog beamforming matrix at the receiving end in the region where non-zero elements may appear in the angular domain channel matrix by using the orthogonal matching pursuit algorithm; and solving the digital beamforming matrix at the transmitting end and the digital beamforming matrix at the receiving end by a mathematical optimization method according to the analog beamforming matrix at the transmitting end, the analog beamforming matrix at the receiving end, and the known channel matrix, so as to obtain the shaped target beam.

[0054] In addition, when the logical instructions in the above-mentioned memory 930 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0055] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior provided by the above-mentioned various methods. The method includes: obtaining an angular domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; according to the relative orientation angle between the two multiple-input multiple-output arrays, using geometric prior information to determine the area where non-zero elements in the angular domain channel matrix may appear, where the geometric prior information is the mapping relationship between the wave departure angle at the transmitting end and the wave arrival angle at the receiving end combined with the lobe width of the array beamforming; using the orthogonal matching pursuit algorithm to search for the analog beamforming matrix at the transmitting end and the analog beamforming matrix at the receiving end in the area where non-zero elements in the angular domain channel matrix may appear; according to the analog beamforming matrix at the transmitting end, the analog beamforming matrix at the receiving end, and the known channel matrix, solving for the digital beamforming matrix at the transmitting end and the digital beamforming matrix at the receiving end through a mathematical optimization method to obtain the shaped target beam.

[0056] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements a beamforming method for 5G multiple-input multiple-output technology based on geometric prior, and the method includes: obtaining an angular domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, where the angular domain channel matrix is obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; according to the relative orientation angle between the two multiple-input multiple-output arrays, using geometric prior information to determine the region where non-zero elements in the angular domain channel matrix may appear, where the geometric prior information is the mapping relationship between the wave departure angle at the transmitting end and the wave arrival angle at the receiving end combined with the lobe width of array beamforming; using the orthogonal matching pursuit algorithm to search for the analog beamforming matrix at the transmitting end and the analog beamforming matrix at the receiving end in the region where non-zero elements in the angular domain channel matrix may appear; according to the analog beamforming matrix at the transmitting end, the analog beamforming matrix at the receiving end, and the known channel matrix, solving for the digital beamforming matrix at the transmitting end and the digital beamforming matrix at the receiving end through a mathematical optimization method to obtain the shaped target beam.

[0057] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0058] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A beamforming method for 5G multiple-input multiple-output technology based on geometric prior, characterized in that: include: Acquire an angle domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, wherein the angle domain channel matrix is ​​obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; According to the relative orientation angle of the two MIMO arrays, the region where non-zero elements in the angle domain channel matrix may appear is determined using geometric prior information, wherein the geometric prior information is a mapping relationship between a wave departure angle at a transmitting end and a wave arrival angle at a receiving end combined with a lobe width of array beamforming; An orthogonal matching pursuit algorithm is used to search for an analog beamforming matrix of a transmitting end and an analog beamforming matrix of a receiving end in a region where non-zero elements in the angle domain channel matrix may appear; According to the analog beamforming matrix of the transmitting end and the analog beamforming matrix of the receiving end and a known channel matrix, a digital beamforming matrix of the transmitting end and a digital beamforming matrix of the receiving end are solved by a mathematical optimization method to obtain a shaped target beam.

2. The method according to claim 1, characterized in that The method of determining, according to the relative orientation angle of the two MIMO arrays, using geometric prior information, an area where non-zero elements in the angle domain channel matrix may appear includes: Determine a closed-form expression for geometric prior information according to the relative orientation angle of the two MIMO arrays; The region where non-zero elements in the angle domain channel matrix may appear is determined according to the closed-form expression.

3. The method according to claim 2, characterized in that The method of using an orthogonal matching pursuit algorithm to search for an analog beamforming matrix of a transmitting end and an analog beamforming matrix of a receiving end in an area where non-zero elements in the angle domain channel matrix may appear includes: Initialize the residual matrix; In each cycle, the product modulus of the residual matrix and each column of the dictionary matrix of the transmitting end is calculated, and a column with the largest product modulus is selected to be added to the codeword set to be used; The residual matrix and the analog beamforming matrix of the transmitting end are updated based on the existing codeword set until the number of used codewords reaches the number of radio frequency links of the transmitting end.

4. The method according to claim 3, characterized in that The step of solving the digital beamforming matrix of the transmitting end and the digital beamforming matrix of the receiving end by a mathematical optimization method according to the analog beamforming matrix of the transmitting end and the analog beamforming matrix of the receiving end and a known channel matrix to obtain a shaped target beam includes: Estimate the angle between the two MIMO array normals using a convex optimization algorithm according to the simulated beamforming matrix of the transmitting end, the simulated beamforming matrix of the receiving end, and a known channel matrix; Based on the angle between the two MIMO array normals, the digital beamforming matrix of the transmitting end and the digital beamforming matrix of the receiving end are optimized to obtain a shaped target beam.

5. The method according to claim 4, characterized in that The estimating the angle between the two MIMO array normals using a convex optimization algorithm according to the simulated beamforming matrix of the transmitting end, the simulated beamforming matrix of the receiving end and a known channel matrix comprises: Defining a residual function according to a known pilot signal, the beamforming matrix, noise, and transmit power; A convex optimization method is used to solve the angle value that minimizes the residual function, and the angle value that minimizes the residual function is used as the angle between the two MIMO array normals.

6. The method according to claim 1, characterized in that The method further comprises: When searching for the wave arrival angle at the receiving end, the geometric prior information is used to narrow the search range, and the square modulus of the product of the row satisfying the specific condition and the residual matrix is ​​calculated.

7. A beamforming device based on geometric prior 5G multiple-input multiple-output technology, characterized in that: include: An acquisition module, used for acquiring an angle domain channel matrix of a wireless channel between two multiple-input multiple-output arrays, wherein the angle domain channel matrix is ​​obtained by performing a two-dimensional discrete Fourier transform on an initial channel matrix; A determination module, configured to determine, according to the relative orientation angle of the two MIMO arrays, an area where non-zero elements in the angle domain channel matrix may appear using geometric prior information, wherein the geometric prior information is a mapping relationship between a wave departure angle at a transmitting end and a wave arrival angle at a receiving end combined with a lobe width of array beamforming; A search module, configured to search for an analog beamforming matrix of a transmitting end and an analog beamforming matrix of a receiving end in an area where non-zero elements in the angle domain channel matrix may appear, using an orthogonal matching pursuit algorithm; The beamforming module is used to solve the digital beamforming matrix of the transmitting end and the digital beamforming matrix of the receiving end through a mathematical optimization method according to the analog beamforming matrix of the transmitting end, the analog beamforming matrix of the receiving end and the known channel matrix, so as to obtain a shaped target beam.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, it implements the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, it implements the beamforming method of the 5G multiple-input multiple-output technology based on geometric prior as described in any one of claims 1 to 6.

Images