A Hybrid Precoding Full-Connection Structure and Its Beamforming Method

Abstract

The present invention proposes a hybrid precoding full connection structure and its beamforming method; the present invention proposes a new hybrid precoding structure, which can further utilize the advantages of the system structure, simplify the hardware structure, and compared with the relatively mature MO iterative optimization method, does not require complex matrix iterative optimization, and the algorithm is simple; the solution of the present inventor is not limited to the performance index of the system spectral efficiency, and can directly approximate the target vector, having more general versatility.

Description

Technical Field

[0001] The present invention relates to the field of millimeter-wave communication technologies, specifically to hybrid beamforming technologies, and more specifically, to a fully-connected structure for hybrid precoding and its beamforming method. Background Art

[0002] Millimeter-wave communication technology has become the most promising communication technology in the communication field at present. There are mainly two reasons: First, the short wavelength of millimeter waves allows a large number of antennas to be installed in a small-size space; Second, in a large-scale MIMO system, the antenna array can provide strong enough beamforming gain to compensate for the severe atmospheric absorption, rain attenuation, and low penetrability in the millimeter-wave band. In a digital precoding structure, each antenna is connected to an expensive radio frequency link, including a digital-to-analog converter, an analog-to-digital converter, an up (down) converter, a mixer, a local oscillator, etc. In the case of a large-scale antenna array, the hardware cost and operating power consumption have become the main obstacles to the development of millimeter-wave communication technology. To overcome this drawback, the prior art uses hybrid precoding technology to reduce the number of radio frequency links, reduce the system operating power consumption and hardware cost. Such a technology combines the advantages of digital beamforming technology and analog beamforming technology.

[0003] Currently, there is no unified system structure for hybrid beamforming technology, and this technology is still in the research and exploration stage. Currently, there are mainly two types of hybrid beamforming structures. One is a fully-connected structure, such as Figure 1 , each radio frequency link is connected to all antennas through phase shifters equal in number to the number of antennas; the other is a partially-connected structure, such as Figure 2 , a radio frequency link is connected to a fixed sub-array, which can greatly reduce the hardware complexity of the system. The fully-connected structure can make full use of the beamforming degrees of freedom provided by the radio frequency links, and its spectral efficiency is close to that of digital precoding under certain conditions of the number of radio frequency links. However, in the fully-connected structure, the original system structure needs to be changed greatly, that is, N RF N T phase shifters (N RF represents the number of radio frequency links, N T represents the number of antennas) and radio frequency adders need to be added, and the increase in radio frequency analog devices and links in practical applications will make it difficult to accurately control the amplitude and phase of the precoding vector. For the partially-connected structure, the phases of the antennas in the same sub-array can be regulated through radio frequency precoding, but the amplitude is provided by a single radio frequency link, and it is difficult to accurately control the precoding vector only through phase changes.

[0004] In this regard, in the Chinese invention patent with the publication date of December 10, 2021: Design Method of Hybrid Precoder and Combiner for 5G, a recursive algorithm based on matrix factorization is adopted to design the analog precoder and the analog combiner, that is, the singular value decomposition is performed on the channel matrix, and then the final analog precoder and the analog combiner are designed through a recursive idea. Then, based on the effective baseband signal associated with the obtained optimal analog precoder and combiner, the digital precoder and the digital combiner are calculated. However, in the derivation process of the spectral efficiency of this scheme, a large number of approximations are used, resulting in a certain performance loss. And in the experiment, two cases where the number of radio frequency links is equal to the number of data streams and the number of radio frequency links is twice the number of data streams are respectively selected, and there is still a large gap compared with the fully digital case. Although the iterative optimization algorithm is simplified to reduce the complexity and save resources, the goal of the spectral efficiency of the fully digital case is not fully achieved, so there are still certain limitations. Summary of the Invention

[0005] Aiming at the limitations of the existing technologies, the present invention proposes a fully connected structure for hybrid precoding. The technical solution adopted by the present invention is as follows:

[0006] A fully connected structure for hybrid precoding includes a transmitting end and a receiving end; at the transmitting end, it includes a precoding module F precoder and N T transmitting antennas; the precoding module F precoder sequentially includes a digital baseband precoder F of dimension BB , a number of radio frequency links at the transmitting end, a transmitting end switching network S of dimension WT , and an analog radio frequency precoder;

[0007] The digital baseband precoder F BB is divided into N S groups, and the dimensions of each group of digital baseband precoders are The radio frequency links at the transmitting end are divided into N S groups, and the numbers of each group of radio frequency links at the transmitting end are The transmitting end switching network is divided into N S groups, and the dimensions of each group of transmitting end switching networks are wherein, N S is the number of data streams of the input signal at the transmitting end.

[0008] Compared with the prior art, the present invention proposes a new hybrid beamforming structure, which can further utilize the advantages of the system structure. While simplifying the hardware structure, compared with the relatively mature MO iterative optimization method, it does not require complex matrix iterative optimization and has a simple algorithm. The solution of the present inventor is not limited to the performance index of the system spectral efficiency, and can directly approximate the target vector, having more general versatility.

[0009] As a preferred solution, the analog radio frequency precoder includes 2N T analog phase shifters and N T radio frequency adders to form the analog radio frequency precoder F RF-1 and the analog radio frequency precoder F RF-2 ; the analog radio frequency precoder F RF-1 and the analog radio frequency precoder F RF-2 each include N T analog phase shifters.

[0010] As a preferred solution, for the j-th path (j = 1, 2..., N S ) data stream s j in the input signal, after sequentially passing through digital baseband precoding the transmitter switching network and the analog radio frequency precoder and the analog radio frequency precoding for phase modulation, the obtained j-th path data stream s j the transmitted signal x after precoding i is expressed by the following formula:

[0011]

[0012] Furthermore, the transmitted signal x after precoding is expressed by the following formula:

[0013]

[0014] where as the transmitter precoding matrix, ||F precoder || 2 = N S ; the input signal satisfies

[0015] As a preferred solution, each radio frequency link is multiplexed at the transmitting antenna, that is, two identical amplitude vectors are selected as the basis to synthesize the full-digital beamforming vector of the corresponding antenna; the precoding matrix obtained by the digital baseband precoder for each path data stream in the input signal is a diagonal matrix, that is:

[0016]

[0017] Among them, base 1 is 1 / 2 of the maximum amplitude value of the j-th column of the optimal digital precoding of the input signal, that is Then base 1 The amplitude range that can be represented is (base 1 *cosθ*2, base 1 *2], and a total of m 1 amplitude values can be represented. Sort |F opt (:, j)| in descending order of magnitude as |F opt (:, j)| down , then there is And so on to obtain

[0018] Furthermore, the phase of each precoding vector in the j-th column F opt-j of the optimal digital precoding is:

[0019]

[0020] The included angle between the two bases corresponding to |F opt (i, j)| and θ opt(i,j) is:

[0021]

[0022] Among them, base n is the base corresponding to |F opt (i, j)|; then the analog RF precodings corresponding to each transmit antenna are respectively:

[0023]

[0024] Furthermore, at the receiving end, it includes N R transmit antennas and a decoder module W decoder ; The decoder module W decoder is sequentially provided with an analog RF decoder, an N-dimensional receiving-end switching network S WR , N receiving-end RF links, an N-dimensional digital baseband decoder W BB ;

[0025] The receiving-end switching network is divided into N S groups, and the dimensions of each group of receiving-end switching networks are respectively The receiving-end RF links are divided into N S groups, and the number of each group of receiving-end RF links is respectively The digital baseband decoder W BB is divided into N S groups, and the dimensions of each group of digital baseband decoders are respectively

[0026] Furthermore, the analog radio frequency decoder includes an analog radio frequency decoder W composed of 2N R analog phase shifters and N R radio frequency adders RF-1 and an analog radio frequency decoder W RF-2 ; The analog radio frequency decoder W RF-1 and the analog radio frequency decoder W RF-2 each include N R analog phase shifters.

[0027] Furthermore,

[0028] The present invention also includes the following content:

[0029] A beamforming method implemented based on the foregoing hybrid precoding full connection structure, comprising the following steps:

[0030] Obtain the optimal digital precoding F of the input signal opt ;

[0031] Arrange the modulus values of F opt (:, j) in descending order to obtain |F opt (:, j)| down , and find the elements of the digital baseband precoding diagonal matrix

[0032] According to the formula Calculate the phase θ of F opt (i, j) opt(i,j) ;

[0033] According to the formula Find the deflection angle θ of the base relative to the target vector; respectively obtain two groups of analog radio frequency precodings BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is the existing full connection structure;

[0035] Figure 2 is the existing partial connection structure;

[0036] Figure 3 Schematic diagram of the hybrid precoding full connection structure provided in Embodiment 1 of the present invention;

[0037] Figure 4It is a decomposition structure diagram of the optimal digital precoding vector;

[0038] Figure 5 It is a simplified structure schematic diagram of the hybrid precoding full-connection structure provided in Embodiment 2 of the present invention;

[0039] Figure 6 It is a schematic diagram of the relationship between the base deflection angle and the power error;

[0040] Figure 7 It is a schematic diagram of the relationship between the base deflection angle and the angle error;

[0041] Figure 8 It is the experimental result of the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 3;

[0042] Figure 9 It is the experimental result of the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 4;

[0043] Figure 10 It is the experimental result of the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 5;

[0044] Figure 11 It is the experimental result of the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 6;

[0045] Figure 12 It is the experimental result of the relationship between the number of RF links and the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 3;

[0046] Figure 13 It is the experimental result of the relationship between the number of RF links and the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 4;

[0047] Figure 14 It is the experimental result of the relationship between the number of RF links and the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 5;

[0048] Figure 15 It is the experimental result of the relationship between the number of RF links and the spectral efficiency of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 6;

[0049] Figure 16 It is the experimental result of the relative error of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 3;

[0050] Figure 17 It is the experimental result of the relative error of the simulation experiment in Embodiment 3 of the present invention when the phase shifter quantization bit number Bit = 4;

[0051] Figure 18 This is the experimental result of the relative error when the quantization bit number Bit of the phase shifter in the simulation experiment of Embodiment 3 of the present invention is 5;

[0052] Figure 19 This is the experimental result of the relative error when the quantization bit number Bit of the phase shifter in the simulation experiment of Embodiment 3 of the present invention is 6. Detailed implementation manners

[0053] The accompanying drawings are only for illustrative purposes and should not be construed as limiting the present patent;

[0054] It should be clear that the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the embodiments of the present application.

[0055] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. The singular forms "a", "the" and "said" used in the embodiments of the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term " / and" used in the present invention refers to and includes any or all possible combinations of one or more of the associated listed items.

[0056] When the following description relates to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all the implementation manners consistent with the present application. On the contrary, they are only examples of the devices and methods consistent with some aspects of the present application as detailed in the appended claims. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0057] In addition, in the description of the present application, unless otherwise specified, "a plurality of" means two or more. " / and" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after. The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0058] To address the limitations of the prior art, this embodiment provides a technical solution. The technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0059] Embodiment 1

[0060] A hybrid precoding fully-connected structure. Please refer to Figure 3 , a hybrid precoding fully-connected structure, including a transmitting end and a receiving end; at the transmitting end, it includes a precoding module F precoder and N T transmitting antennas; the precoding module F precoder sequentially has a D-dimensional digital baseband precoder F BB , M transmitting-end RF links, an M-dimensional transmitting-end switching network S WT , and an analog RF precoder;

[0061] The digital baseband precoder F BB is divided into N S groups, and the dimensions of each group of digital baseband precoders are The transmitting-end RF links are divided into N S groups, and the number of each group of transmitting-end RF links is The transmitting-end switching network is divided into N S groups, and the dimensions of each group of transmitting-end switching networks are wherein, N S is the number of data streams of the input signal at the transmitting end.

[0062] Compared with the prior art, the present invention proposes a new hybrid beamforming structure, which can further utilize the advantages of the system structure. While simplifying the hardware structure, compared with the relatively mature MO iterative optimization method, it does not require complex matrix iterative optimization, and the algorithm is simple; the solution of the present inventor is not limited to the performance index of the system spectral efficiency, and can directly approach the target vector, having more general versatility.

[0063] As a preferred embodiment, the analog RF precoder includes an analog RF precoder F T constituted by 2N T analog phase shifters and N RF-1 RF adders, and an analog RF precoder F RF-2 ; the analog RF precoder F RF-1 and the analog RF precoder F RF-2 each include N T analog phase shifters.

[0064] Specifically, at the transmitting end, the input signal sequentially passes through the digital baseband precoder F BB , the radio frequency link, and the transmitting end switching network S WT , and then is split into two paths, respectively passing through the analog radio frequency precoder F RF-1 and the analog radio frequency precoder F RF-2 . The analog phase shifters of the analog radio frequency precoders perform phase modulation, and then are synthesized through the radio frequency adder to form the precoded transmitted signal.

[0065] Furthermore, at the receiving end, it includes N R transmitting antennas and the decoder module W decoder ; The decoder module W decoder is sequentially provided with an analog radio frequency decoder, the N-dimensional receiving end switching network S WR , the N receiving end radio frequency links, the N-dimensional digital baseband decoder W BB ;

[0066] The receiving end switching network is divided into N S groups, and the dimensions of each group of receiving end switching networks are The receiving end radio frequency links are divided into N S groups, and the numbers of each group of receiving end radio frequency links are The digital baseband decoder W BB is divided into N S groups, and the dimensions of each group of digital baseband decoders are

[0067] Even further, the analog radio frequency decoder includes an analog radio frequency decoder W R composed of 2N analog phase shifters and N radio frequency adders R and the analog radio frequency decoder W RF-1 ; The analog radio frequency decoder W RF-2 and the analog radio frequency decoder W RF-1 respectively include N RF-2 analog phase shifters. R

[0068] Specifically, the large-scale MIMO system considered in this embodiment is a point-to-point system.

[0069] In the structure provided in this embodiment, and can both be adjusted according to the actual situation. For example, when N S = 3, the following three situations may exist:

[0070] ​For the convenience of discussion, in the following derivation and experiment of the principle of this embodiment, the number of data streams at both the transmitter and receiver ends will be selected as N s , and it satisfies N S << N T , N S << N R , and

[0071] As a preferred embodiment, the j-th (j = 1, 2..., N S ) data stream s j in the input signal sequentially passes through digital baseband precoding the transmitter switch network and the analog RF precoder After phase modulation with the analog RF precoding , the j-th data stream s j of the transmitted signal x after precoding is represented by the following formula: i

[0072]

[0073] Furthermore, the transmitted signal x after precoding is represented by the following formula:

[0074]

[0075] where is the precoding matrix at the transmitter end. For the normalized power limit of the entire system, i.e., ||F precoder || 2 = N S ; the input signal satisfies

[0076] The structure of the receiver is similar to that of the transmitter. At the receiver, and S WR-j respectively represent the digital baseband decoder, two sets of analog RF decoders, and the receiver switch network for the j-th data stream of the received signal. The signal after decoding the received signal is:

[0077] y = (W decoder ) H HF precoder s + (W decoder ) H n;

[0078] where is the total decoding matrix; ρ represents the average transmit power, is the transmission matrix of the channel. n is the additive white Gaussian noise (AWGN) that is independent and identically distributed (i.i.d.) Analog radio frequency precoding and analog radio frequency decoding control the phase of the signal by phase shifters. Therefore, in F RF-i-1 , F RF-i-2 , W RF-i-1 and W RF-i-2 all non-zero elements should satisfy the transverse mode constraint, i.e., |(F RF-i-1 ) i,j | = 1, |(F RF-i-2 ) i,j | = 1, |(W RF-i-1 ) i,j | = 1 and |(W RF-i-2 ) i,j | = 1.

[0079] The spectral efficiency in point-to-point MIMO communication can be expressed as the following expression:

[0080]

[0081] The objective of this embodiment is:

[0082]

[0083] The millimeter-wave channel follows the Rayleigh fading channel. The channel model in this embodiment is the SV channel model. For the millimeter-wave channel, H can be expressed as the following expression:

[0084]

[0085] where, N cl and N ray represent the number of clusters and the number of path channels contained in each cluster, α il represents the l-th propagation path in the i-th propagation cluster; it is assumed that α il is independently and identically distributed according to the complex Gaussian distribution and the normalization coefficient satisfies In addition and represent the receive and transmit array response vectors respectively, and represent the azimuth and elevation angles of arrival and departure respectively. This embodiment considers a uniform square array (USPA) with antenna elements. Therefore, the response vector of the l-th path in the i-th cluster of the array can be written as:

[0086]

[0087] where d and λ represent the spacing of the antennas and the wavelength of the signal, and and are indices of the two-dimensional plane of the antenna. Although this channel model will be used in subsequent simulation experiments, this is only an experimental condition set for the experiment. In fact, the structure and method of the present invention can be used in more general scenarios. In subsequent experiments, it will be assumed that perfect channel state information (CSI) can be obtained. For the convenience of verification, a narrowband fading channel will be selected in subsequent experiments.

[0088] Maximizing the objective function will approximately result in the maximum spectral efficiency:

[0089]

[0090] The unconstrained optimal digital precoding is F opt and the digital decoding is W opt which are the first N s columns of V and U, where both V and U are unitary matrices from the singular value decomposition (SVD) of the channel matrix, i.e., H = UΣV H . At the transmitter, the transverse mode constraint of the RF precoding and the limitation of the transmit power need to be satisfied, and at the receiver, only the transverse mode constraint of the RF precoding needs to be satisfied.

[0091] In the structure of this embodiment, the optimal digital precoding vector and digital receiving vector F opt and W opt are N T ×N S and N R ×N S vector matrices. Since the structure of the system is simplified, complex optimization methods do not need to be adopted. Because there is a mathematical closed-form solution under this structure, the result must satisfy the requirements regarding the transverse mode constraint and the transmit power proposed in the above formula. This embodiment is derived with the transmitter as an example, and the receiver is similar. The basic theoretical derivation is as follows:

[0092] The basic theorem of plane vectors: Taking two non-collinear vectors and (both non-zero vectors) in the plane as the basis, any vector in the plane has a unique pair of real numbers, λ and μ, such that holds.

[0093] As a preferred embodiment, each RF link is multiplexed at the transmitting antenna, that is, two identical amplitude vectors are selected as the basis to synthesize the full-digital beamforming vector corresponding to the antenna; the precoding matrix obtained by the digital baseband precoder for each data stream in the input signal is a diagonal matrix, i.e.:

[0094]

[0095] Among them, base 1 is 1 / 2 of the maximum amplitude of the j-th column of the optimal digital precoding of the input signal, that is Then base 1 can represent an amplitude range of (base 1 *cosθ*2, base 1 *2], and can represent a total of m 1 amplitudes. Sort |F opt (:,j)| in descending order of magnitude as |F opt (:,j)| down , then there is And so on to get

[0096] Furthermore, the phase of each precoding vector in the j-th column F opt-j of the optimal digital precoding is:

[0097]

[0098] The angle between the two bases corresponding to |F opt (i,j)| and θ opt(i,j) is:

[0099]

[0100] Among them, base n is the base corresponding to |F opt (i,j)|; then the analog radio frequency precodings corresponding to each transmit antenna are respectively:

[0101]

[0102] Thus, the decomposition of the optimal digital precoding into two bases with equal lengths can be achieved, and the vector representation is shown in Figure 4 .

[0103] Under this structure, the optimal digital precoding vector can be exactly the same as the sum of the two decomposed vectors, that is, the effect of the optimal digital precoding can be achieved.

[0104] Embodiment 2

[0105] Embodiment 2 can be regarded as a further improved, refined or supplementary scheme based on Embodiment 1, specifically:

[0106] When Figure 3 the number of radio frequency links in hour, Figure 3 The system structure will degenerate into Figure 5 structure.

[0107] Assuming that the target amplitude of a single array element is: base*cosθ*2, due to the use of a low-resolution phase shifter, the actual amplitude is: base*cos(θ+Δ)*2 (the reason why cos(θ+Δ) is selected here instead of cos(θ-Δ) is that: as the angle decreases, the change in the cosine value is smaller than the error), then the power attenuation of the corresponding array element is:

[0108]

[0109] Under the premise of maximum power error, the maximum amplitude deviation in the horizontal direction is:

[0110]

[0111] The vertical magnitude of the final target vector:

[0112] base×cos(θ+Δ)×2;

[0113] The final maximum angle deviation is:

[0114]

[0115] The image of the error result is Figure 6 , Figure 7 shown.

[0116] Although in theoretical analysis, Figure 3 The structure can achieve the same effect as the full digital precoding, but based on the high-resolution phase shifter, when a low-resolution phase shifter is used in the system, direct quantization of the phase of the analog RF precoding in the above text will bring about a large performance loss, and a RF link performs baseband precoding on a data stream, which will lose a lot of energy and reduce the stability of the system. Therefore, the structure of the present invention can effectively solve the problems of system performance loss, energy consumption and system stability.

[0117] It can be concluded from the experimental results that when the deflection angle of the substrate is between 0° and 60°, the maximum error of the final vector angle is basically the same as the quantization error of the phase shifter, but the power error is quite different and increases with the increase of the deflection angle. Therefore, in order to improve the amplitude accuracy of the system, the deflection angle is controlled in a smaller range as much as possible. By increasing the number of radio frequency links, the deflection angle of the system can be further reduced. For this purpose, the present invention proposed Figure 3 The structure in .

[0118] In this embodiment, the relative error of hybrid precoding is defined as a parameter for system measurement:

[0119]

[0120] F precoder represents the precoding vector matrix under the condition of the same-structure algorithm. For the MO-Alt method and the SDR-Alt method [Alternating Minimization Algorithms for Hybrid Precoding in MillimeterWave MIMO Systems]: F precoder = F RF F BB ; For the structure proposed by the present invention:

[0121]

[0122] When the number of radio frequency links matching each data stream in the structure of the present invention is 1, that is, the structure of FIG. 5, then there is F precoder = F RF F BB .

[0123] Embodiment 3

[0124] A beamforming method implemented based on the hybrid precoding full-connection structure of Embodiment 1 or 2, comprising the following steps:

[0125] Obtain the optimal digital precoding F of the input signal opt ;

[0126] Arrange the modulus values of F opt (:, j) in descending order to obtain |F opt (:, j)| down , and obtain the elements of the digital baseband precoding diagonal matrix

[0127] According to the formula Calculate the phase θ of F opt (i, j) opt(i,j) ;

[0128] According to the formula Find out the deflection angle θ of the base relative to the target vector; respectively obtain two groups of analog radio frequency precodings

[0129] After performing the above steps on the data streams from 1 to Ns in the input signal respectively, the precoded waveform can be obtained.

[0130] The method of this embodiment is based on the condition that it can be applied to a phase shifter with low resolution. The main reason is that in the application of the millimeter-wave spectrum, a high-resolution phase shifter will significantly increase power consumption and hardware complexity. Therefore, it is more practical to use a low-resolution phase shifter in actual applications. In actual engineering applications, 6-bit and 8-bit phase shifters are more commonly used, and 10-bit high-resolution phase shifters are also used.

[0131] Next, this embodiment will show simulation results to prove the effectiveness of the algorithm we proposed. In the MIMO system, the number of transmit and receive antennas N T = 144, N R = 36, both the transmitter and receiver are USPA, and the channel parameters are set as N cl = 5 clusters, N ray = 10, the average power of each cluster The antenna spacing is λ / 2, where λ represents the wavelength. The azimuth and elevation angles of AOD and AOA both follow the Laplace distribution [0, 2π), and the angle spread is 10 degrees.

[0132] All simulation results are achieved with more than 1000 channel data.

[0133] Since in the structure of the present invention, the deflection angle between the target vector and the base vector is between 0 - 90°, the minimum quantization bit number of the phase shifter selected in the experiment of this embodiment is 3 bits. In the experiment of this embodiment, due to the poor spectral efficiency performance of the SDR-Alt algorithm with a partial connection structure, its phase quantization is not further compared. In the experiment of this embodiment, the novel structure and the fully connected MO-Alt algorithm proposed by the present invention are both the spectral efficiency obtained directly after quantization on the premise of a certain quantization bit number of the phase shifter.

[0134] In Figures 8 - 11 In the experimental results, when the number of RF links is the same as the number of data streams, that is When, the novel structure proposed by the present invention has significantly better spectral efficiency performance than the fully connected structure when the quantization bit number of the phase shifter is greater than 3, and is basically similar to the spectral efficiency of the fully connected structure when N S = 6. When Under the condition of a phase shifter with a low quantization bit number, only when the quantization bit number is 3, its spectral efficiency is slightly lower than that of the fully connected structure. When the quantization bit number is 4 or higher, its spectral efficiency is better than that of the fully connected structure, and when the quantization bit number is greater than 5, it can be seen that the novel structure of the present invention has N S= 3, and it is also more spectrally efficient than the fully connected structure. In practical engineering, 3, 4, 5, and 6-bit phase shifters are relatively common in engineering applications. Therefore, the experimental results prove that the novel structure proposed by the present invention can achieve higher spectral efficiency under the condition of low-resolution phase shifters, and further reduces the number of RF links and the complexity of the optimization algorithm.

[0135] In Figures 12 - 15 's experimental results, it is under the condition of signal-to-noise ratio SNR = 0 and the number of data streams N S = 3 that the relationship between the number of RF links and the system spectral efficiency is compared under different quantization bit numbers of the phase shifter. It can be seen that when the number of RF links is less than twice the number of data streams, the structure proposed by the present invention has obvious advantages in spectral efficiency. Even under the condition of a low-precision phase shifter with a quantization bit number of 3, when the number of RF links gradually increases, under the condition of low quantization bit numbers, the fully connected structure is slightly better than the structure proposed by the present invention. However, when the quantization bit number increases, when bit = 6, the spectral efficiency achieved by the structure proposed by the present invention is better than that of the fully connected structure under the same conditions. Therefore, whether it is to reduce the number of RF links or further improve the spectral efficiency, the structure proposed by the present invention has obvious advantages.

[0136] In Figures 16 - 19 's experimental results, the present implementation evaluates the performance of the system according to the relative error defined in the previous text. On the premise that the number of RF links is greater than or equal to the number of data streams, the relative error between the structure proposed in the present invention and the optimal digital precoding vector is the lowest under the conditions of phase shifters with different quantization bit numbers. The relative error of the partially connected structure is always at a relatively high level, and the relative error of the fully connected structure is always higher than that of the structure proposed by the present invention, especially when the number of RF links is between 1 and 2 times the number of data streams. The relative error defined by the present invention can simultaneously measure the errors of the amplitude and phase with the optimal all-digital precoding vector. From the experimental results of the relative error, the effectiveness of the structure proposed by the present invention can be seen.

[0137] Conclusion:

[0138] The effectiveness of the structure proposed by the present invention is proved by various simulation experimental results. Compared with the traditional fully connected structure, the number of RF links in the system is further reduced. The structure proposed by the present invention can further utilize the advantages of the system structure, simplify the hardware structure while, and compared with the relatively mature MO iterative optimization algorithm, there is no complex matrix iterative optimization, and the algorithm is simple. And it is not limited to the performance index of the system spectral efficiency, and can directly approach the target vector, with more general versatility.

[0139] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, rather than limitations on the implementation manners of the present invention. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation manners here. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the claims of the present invention.

Claims

1. A hybrid precoding full-connection structure, Characterized in that, It includes a transmitting end and a receiving end; at the transmitting end, it includes a precoding module F precoder and N T transmitting antennas; the precoding module F precoder is sequentially provided with a D-dimensional digital baseband precoder F BB , T transmitting-end RF links, a T-dimensional transmitting-end switching network S WT and N S groups of analog RF precoders; One set of the analog RF pre - encoders includes 2N T analog phase shifters and N T RF adders to form the analog RF pre - encoder F RF-1 and the analog RF pre - encoder F RF-2 ; the analog RF pre - encoder F RF-1 and the analog RF pre - encoder F RF-2 each include N T analog phase shifters; At the transmitting end, the input signal sequentially passes through the digital baseband precoder F BB , the radio frequency link, and the transmitting end switching network S WT , and then is split into two paths, and respectively passes through the analog radio frequency precoder F RF-1 and the analog radio frequency precoder F RF-2 's analog phase shifters for phase modulation, and then is synthesized through the radio frequency adder to form the precoded transmitted signal; The digital baseband precoder F BB is divided into N S groups, and the dimensions of the digital baseband precoders in each group are The transmit - end radio - frequency links are divided into N S groups, and the numbers of the transmit - end radio - frequency links in each group are The transmit - end switching network is divided into N S groups, and the dimensions of the transmit - end switching networks in each group are where N S is the number of data streams of the input signal at the transmit end.

2. The hybrid precoding full-connection structure according to claim 1, Characterized in that, The j-th path in the input signal, where j = 1, 2..., N S , data stream s j sequentially passes through digital baseband precoding the transmitter switching network and the analog RF precoder After phase modulation with the analog RF precoding , the j-th data stream s obtained j The transmitted signal x after precoding i is expressed by the following formula:

3. The hybrid precoding full-connection structure according to claim 2, Characterized in that, The transmitted signal x after precoding is expressed by the following formula: Among them, As the transmit - end precoding matrix, ||F precoder || 2 = N S ; The input signal satisfies 4. The hybrid precoding full-connection structure according to claim 1, Characterized in that, For each radio frequency link, multiplexing is achieved at the transmitting antenna, that is, two identical amplitude vectors are selected as the basis to synthesize the full-digital beamforming vector of the corresponding antenna; the precoding matrix obtained by the digital baseband precoder for each data stream in the input signal is a diagonal matrix, that is: Among them, base 1 is 1 / 2 of the maximum amplitude value of the j-th column of the optimal digital precoding of the input signal, that is Then base 1 The amplitude range that can be represented is (base 1 * cosθ * 2, base 1 * 2], and a total of m 1 amplitude values can be represented. Sort |F opt (:, j)| in descending order of magnitude as |F opt (:, j)| down , then there is base 2 = |F opt (:, j)| down(m1+1) , and so on to obtain 5. The hybrid precoding full-connection structure according to claim 4, Characterized in that, The phase of each precoding vector in the optimal digital precoding, column j of F opt-j is as follows: |F opt (i,j)|The included angle between the corresponding two bases and θ opt(i,j) is as follows: Among them, base n is |F opt (i, j)| corresponding basis; then the analog radio frequency precodings corresponding to each transmitting antenna are respectively:

6. The hybrid precoding full-connection structure according to claim 1, Characterized in that, At the receiving end, it includes N R transmitting antennas and a decoder module W decoder ; The decoder module W decoder sequentially has N S groups of analog RF decoders, a receiving-end switching network S WR , receiving-end RF links, a digital baseband decoder W BB ; The receiving - end switching network is divided into N S groups, and the dimensions of the receiving - end switching networks of each group are The receiving - end radio - frequency link is divided into N S groups, and the numbers of the receiving - end radio - frequency links of each group are The digital base - band decoder W BB is divided into N S groups, and the dimensions of the digital base - band decoders of each group are 7. The hybrid precoding full-connection structure according to claim 6, Characterized in that, One set of the analog RF decoders includes 2N R analog phase shifters and N R RF adders to form the analog RF decoder W RF-1 and the analog RF decoder W RF-2 ; the analog RF decoder W RF-1 and the analog RF decoder W RF-2 each include N R analog phase shifters.

8. A beamforming method implemented based on the hybrid precoding full-connection structure according to any one of claims 4 to 5, Characterized in that, Comprising the following steps: Obtain the optimal digital precoding F of the input signal opt ; Arrange the modulus values of F opt (:, j) in descending order to obtain |F opt (:, j)| down , and find the elements of the digital baseband precoding diagonal matrix According to the formula calculate F opt the phase θ of (i, j) opt(i,j) ; According to the formula calculate the deflection angle θ of the base relative to the target vector; Two sets of simulated RF precoding are obtained respectively

Images