Ultra-large-scale MIMO near-field beam training method, device and electronic equipment

Abstract

The embodiments of the present invention provide a method, device, electronic device and computer-readable storage medium for ultra-large-scale MIMO near-field beam training. The method includes: obtaining a discrete Fourier transform codebook; performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set; determining a support angle interval according to the user received signal power set; determining user angle estimation information and user distance estimation information according to the support angle interval; determining a polar domain codeword according to the user angle estimation information and user distance estimation information. According to the solution of the embodiments of the present invention, the training overhead of beam training can be significantly reduced, and the achievable rate of the communication system can be improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a very large-scale MIMO near-field beam training method, device, electronic device and computer-readable storage medium. Background Art

[0002] Very large-scale arrays have become a very promising technology for the future sixth-generation wireless system. Very large-scale arrays use multiple-input multiple-output (MIMO) technology to achieve ultra-high spectrum and extremely high spatial resolution in high frequency bands, thus adapting to applications such as extended reality, holographic video, and autonomous driving. By increasing the number of antennas by another order of magnitude, very large-scale arrays are expected to bring fundamental changes to electromagnetic propagation modeling, that is, from traditional far-field channel modeling to new near-field channel modeling.

[0003] In particular, unlike the far-field channel modeling based on the plane wavefront assumption, the near-field channel needs to be accurately modeled based on the spherical wavefront. Unlike the traditional maximum ratio transmission far-field beamforming that directs the beam energy to a certain angle, the near-field beamforming based on the spherical wavefront characteristics allows the base station to concentrate the beam energy in a certain spatial area. This characteristic is called the near-field beam focusing effect, which causes the base station to produce an energy diffusion effect when using the traditional discrete Fourier transform-based far-field codebook for near-field channel beam scanning. This effect causes the beam energy pointing in a specific direction to spread to multiple angles, making it impossible to find the user angle by searching for the maximum received signal power at the receiving end. Therefore, directly applying the far-field beam training method based on the discrete Fourier transform codebook to the near-field communication system will result in a significant performance degradation.

[0004] To solve this problem, existing research work has focused on designing a new type of beam codebook and corresponding beam training scheme for near-field channels. Specifically, for beam codebook design, the related art proposes a new two-dimensional polar domain codebook for near-field beam training, which is uniformly sampled in the angle domain and non-uniformly sampled in the distance domain; based on this codebook, a simple and direct beam training method is to exhaustively search all candidate codewords in the two-dimensional polar domain. In addition, the related art also proposes an efficient two-stage hierarchical beam training method, which first searches for a rough angle based on a far-field discrete Fourier codebook, and then uses the designed hierarchical near-field codebook to search for a more accurate user angle range pair. However, the beam training overhead of these two beam training methods is too high, resulting in insufficient subsequent data transmission time and a reduced achievable rate of the communication system. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art.

[0006] To this end, the present invention proposes a very large-scale MIMO near-field beam training method, which can significantly reduce the training overhead of beam training and improve the achievable rate of the communication system.

[0007] The present invention also proposes a device applying the ultra-large-scale MIMO near-field beam training method.

[0008] The present invention also proposes an electronic device using the ultra-large-scale MIMO near-field beam training method.

[0009] The present invention also proposes a computer-readable storage medium that applies the above-mentioned ultra-large-scale MIMO near-field beam training method.

[0010] According to the ultra-large-scale MIMO near-field beam training method of the first aspect of the present invention, the method includes:

[0011] Obtain a discrete Fourier transform codebook;

[0012] Performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set;

[0013] Determining a support angle interval according to the user received signal power set;

[0014] Determining user angle estimation information and user distance estimation information according to the support angle interval;

[0015] A polar field codeword is determined according to the user angle estimation information and the user distance estimation information.

[0016] According to some embodiments of the present invention, performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set includes:

[0017] Performing beam direction adjustment processing on a plurality of training signals transmitted by a base station according to the discrete Fourier transform codebook;

[0018] Based on the plurality of training signals processed after receiving the beam direction adjustment at the user end, and determining the user received signal power from each of the training signals;

[0019] The multiple user received signal powers are determined as the user received signal power set.

[0020] According to some embodiments of the present invention, determining the support angle interval according to the user received signal power set includes:

[0021] Comparing the received signal power of each user in the received signal power set with a preset signal power threshold, wherein the signal power threshold is determined by a receiving beam mode of the user terminal;

[0022] Filtering out a subset of user received signal powers greater than a preset signal power threshold from the user received signal power set;

[0023] A codeword index set is determined according to the user received signal power subset, and the support angle interval is determined based on the codeword index set.

[0024] According to some embodiments of the present invention, determining user angle estimation information and user distance estimation information according to the support angle interval includes:

[0025] Determine the median value of the support angle interval as a user angle index;

[0026] Determine the user angle estimation information according to the user angle index and the discrete Fourier transform codebook;

[0027] The user distance estimation information is obtained by performing distance estimation processing according to the user angle estimation information and the support angle interval.

[0028] According to some embodiments of the present invention, performing distance estimation processing according to the user angle estimation information and the support angle interval to obtain the user distance estimation information includes:

[0029] Determine the support angle interval width and the received power ratio according to the support angle interval; wherein the support angle interval width represents the difference between the maximum spatial angle and the minimum spatial angle in the support angle interval;

[0030] The user distance estimation information is obtained according to the received power ratio, the support angle interval width, the user angle estimation information and a preset distance calculation function.

[0031] According to some embodiments of the present invention, performing distance estimation processing according to the user angle estimation information and the support angle interval to obtain the user distance estimation information includes:

[0032] Constructing a received signal power ratio according to all received signal powers within the support angle interval and the received signal power corresponding to the user angle estimation information;

[0033] Constructing a power ratio mean square error solution optimization problem according to the received signal power ratio and a preset power ratio comparison value;

[0034] The power ratio mean square error solution optimization problem is solved based on a preset exhaustive search strategy to obtain the user estimation information.

[0035] According to some embodiments of the present invention, after determining the polar field codeword according to the user angle estimation information and the user distance estimation information, the method further includes:

[0036] Calculating an optimal beamforming vector according to the polar domain codeword;

[0037] A beamforming process is performed based on the optimal beamforming vector.

[0038] According to a second aspect of the present invention, a very large-scale MIMO near-field beam training device comprises:

[0039] A first processing module, used for obtaining a discrete Fourier transform codebook;

[0040] A second processing module, configured to perform beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set;

[0041] A third processing module, configured to determine a support angle interval according to the user received signal power set;

[0042] a fourth processing module, configured to determine user angle estimation information and user distance estimation information according to the support angle interval;

[0043] The fifth processing module is used to determine a polar field codeword according to the user angle estimation information and the user distance estimation information.

[0044] An electronic device according to an embodiment of the third aspect of the present invention includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the ultra-large-scale MIMO near-field beam training method as described above when executing the computer program.

[0045] According to a computer-readable storage medium of an embodiment of the fourth aspect of the present invention, the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a control processor, the ultra-large-scale MIMO near-field beam training method as described above is implemented.

[0046] The ultra-large-scale MIMO near-field beam training method according to an embodiment of the present invention has at least the following beneficial effects: in the process of near-field beam training, a discrete Fourier transform codebook is first obtained, and then a beam scanning process is performed according to the discrete Fourier transform codebook to obtain a user received signal power set; then a support angle interval is determined according to the user received signal power set; then, user angle estimation information and user distance estimation information are determined according to the support angle interval; finally, a polar domain codeword is determined according to the user angle estimation information and the user distance estimation information to complete the near-field beam training process; through the above technical solution, the training overhead of beam training can be significantly reduced, and the achievable rate of the communication system can be improved.

[0047] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are used to provide further understanding of the technical solution of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the technical solution of the present disclosure and do not constitute a limitation on the technical solution of the present disclosure.

[0049] Figure 1 It is a flow chart of a very large-scale MIMO near-field beam training method provided by an embodiment of the present invention;

[0050] Figure 2 It is a specific flow chart of beam scanning of a very large-scale MIMO near-field beam training method provided by an embodiment of the present invention;

[0051] Figure 3 It is a specific flow chart of determining the support angle interval of a very large-scale MIMO near-field beam training method provided by an embodiment of the present invention;

[0052] Figure 4 It is a specific flow chart of determining user angle estimation information and user distance estimation information of a super large-scale MIMO near-field beam training method provided by an embodiment of the present invention;

[0053] Figure 5 It is a specific flow chart of determining user distance estimation information of a super large-scale MIMO near-field beam training method provided by an embodiment of the present invention;

[0054] Figure 6 It is a specific flow chart of determining user distance estimation information of a super large-scale MIMO near-field beam training method provided by another embodiment of the present invention;

[0055] Figure 7 is a flow chart of a super-large-scale MIMO near-field beam training method provided by another embodiment of the present invention;

[0056] Figure 8 is a system schematic diagram of a narrowband ultra-large-scale array downlink communication system provided by an embodiment of the present invention;

[0057] Fig. 9 It is a schematic diagram of a receiving beam pattern for far-field beam scanning at different user angles and distances provided by an embodiment of the present invention;

[0058] Fig.10 is a schematic diagram of the relationship between angle estimation and reference signal-to-noise ratio in different solutions provided by an embodiment of the present invention;

[0059] Fig.11 is a schematic diagram of the relationship between distance estimation and reference signal-to-noise ratio in different solutions provided by an embodiment of the present invention;

[0060] Fig.12 is a schematic diagram of the relationship between the achievable rate and the reference signal-to-noise ratio of different solutions provided by an embodiment of the present invention;

[0061] Fig.13 It is a schematic diagram of the structure of a very large-scale MIMO near-field beam training device provided by an embodiment of the present invention;

[0062] Fig.14 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0065] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0066] The present invention provides a method, device, electronic device and computer-readable storage medium for ultra-large-scale MIMO near-field beam training. The method comprises: in the process of near-field beam training, firstly obtaining a discrete Fourier transform codebook, then performing beam scanning processing according to the discrete Fourier transform codebook to obtain a user received signal power set; then determining a support angle interval according to the user received signal power set; then determining user angle estimation information and user distance estimation information according to the support angle interval; finally determining a polar domain codeword according to the user angle estimation information and the user distance estimation information to complete near-field beam training processing; through the above technical scheme, the training overhead of beam training can be significantly reduced, and the achievable rate of the communication system can be improved.

[0067] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0068] like Figure 1 As shown, Figure 1 1 is a flow chart of a method for ultra-large-scale MIMO near-field beam training provided by an embodiment of the present invention. The method includes but is not limited to step S100, step S200, step S300, step S400 and step S500:

[0069] Step S100, obtaining a discrete Fourier transform codebook;

[0070] Step S200, performing beam scanning according to a discrete Fourier transform codebook to obtain a user received signal power set;

[0071] Step S300, determining a support angle interval according to a user received signal power set;

[0072] Step S400, determining user angle estimation information and user distance estimation information according to the support angle interval;

[0073] Step S500, determining a polar field codeword according to user angle estimation information and user distance estimation information.

[0074] It should be noted that in the process of near-field beam training, the discrete Fourier transform codebook is first obtained, and then beam scanning processing is performed according to the discrete Fourier transform codebook to obtain the user received signal power set; then the support angle interval is determined according to the user received signal power set; then the user angle estimation information and the user distance estimation information are determined according to the support angle interval; finally, the polar domain codeword is determined according to the user angle estimation information and the user distance estimation information to complete the near-field beam training processing; through the above technical solution, the training overhead of beam training can be significantly reduced, and the achievable rate of the communication system can be improved.

[0075] It is worth noting that the beam transmitted by the base station to the user end can be adjusted according to the discrete Fourier transform codebook, for example, the direction of the beam can be adjusted; wherein the user end can be a mobile phone. The user end can also scan the received beam to obtain the user received signal power set.

[0076] It should be noted that the support angle interval is determined according to the user received signal power set, where the support angle interval is a spatial angle region, and its corresponding normalized beam gain is higher than the peak normalized beam gain. Subsequently, the user angle estimation information and user distance estimation information can be determined according to the support angle interval, and the user angle and distance are determined based on the support angle interval, which greatly reduces the overhead of the entire beam training process.

[0077] It is worth noting that after determining the user angle estimation information and the user distance estimation information, the polar codeword can be determined based on the two to prepare for the subsequent beamforming.

[0078] In addition, in one embodiment, if Figure 2 As shown, the above step S200 may include but is not limited to step S210, step S220 and step S230.

[0079] Step S210, performing beam direction adjustment processing on multiple training signals transmitted by the base station according to a discrete Fourier transform codebook;

[0080] Step S220, receiving a plurality of training signals after the beam direction adjustment processing based on the user end, and determining the user received signal power from each training signal;

[0081] Step S230: determining the received signal powers of multiple users as a user received signal power set.

[0082] It should be noted that in the process of performing beam scanning according to the discrete Fourier transform codebook to obtain the user received signal power set, first, the beam direction adjustment processing is performed on the multiple training signals transmitted by the base station according to the discrete Fourier transform codebook; then, based on the multiple training signals after the user received beam direction adjustment processing, the user received signal power is determined from each training signal; finally, the multiple user received signal powers are determined as the user received signal power set.

[0083] It is worth noting that each training signal transmitted by the base station has a corresponding signal power. After the user end receives multiple training signals, it can determine multiple user received signal powers from the received multiple training signals; finally, the multiple user received signal powers are determined as a user received signal power set, in preparation for the subsequent support angle interval determination.

[0084] In addition, in one embodiment, if Figure 3 As shown, the above step S300 may include but is not limited to step S310, step S320 and step S330.

[0085] Step S310, comparing the received signal power of each user in the received signal power set with a preset signal power threshold, wherein the signal power threshold is determined by a receiving beam mode of the user terminal;

[0086] Step S320, filtering out a subset of user received signal powers greater than a preset signal power threshold from the user received signal power set;

[0087] Step S330: determine a codeword index set according to the user received signal power subset, and determine a support angle interval based on the codeword index set.

[0088] It should be noted that in the process of determining the support angle interval based on the user received signal power set, first, each user received signal power in the user received signal power set is compared with the preset signal power threshold; then, a user received signal power subset greater than the preset signal power threshold is screened out from the user received signal power set; then, a codeword index set is determined based on the user received signal power subset, and finally, the support angle interval is determined based on the codeword index set.

[0089] It is worth noting that the preset signal power threshold is determined by the receiving beam mode of the user end, and each user received signal power in the user received signal power set is compared with the signal power threshold to screen out the user received signal power that meets the comparison conditions; subsequently, all user received signal powers that meet the comparison conditions are taken as the user received signal power subset; subsequently, the corresponding codeword index set can be determined according to the user received signal power subset, and finally, the corresponding support angle interval can be determined based on the codeword index set.

[0090] In addition, in one embodiment, if Figure 4 As shown, the above step S400 may include but is not limited to step S410, step S420 and step S430.

[0091] Step S410, determining the median value of the support angle interval as the user angle index;

[0092] Step S420, determining user angle estimation information according to the user angle index and the discrete Fourier transform codebook;

[0093] Step S430: performing distance estimation processing according to the user angle estimation information and the support angle interval to obtain user distance estimation information.

[0094] It should be noted that in the process of determining the user angle estimation information and the user distance estimation information based on the support angle interval, the median value of the support angle interval is first determined as the user angle index, and then the user angle estimation information can be determined based on the user angle index and the discrete Fourier transform codebook; finally, the user distance estimation information can be obtained by performing distance estimation processing based on the user angle estimation information and the support angle interval, which greatly reduces the overhead of beam training.

[0095] It is worth noting that the median value of the support angle interval is approximately equal to the user angle, so the median value of the support angle interval can be determined as the user angle index, and the user angle estimation information can be subsequently determined based on the user angle index and the discrete Fourier transform codebook.

[0096] In addition, in one embodiment, if Figure 5 As shown, the above step S430 may include but is not limited to step S431 and step S432.

[0097] Step S431, determining a support angle interval width and a receiving power ratio according to the support angle interval; wherein the support angle interval width represents a difference between a maximum spatial angle and a minimum spatial angle in the support angle interval;

[0098] Step S432, obtaining user distance estimation information according to the received power ratio, the support angle interval width, the user angle estimation information and a preset distance calculation function.

[0099] It should be noted that in the process of performing distance estimation processing based on the user angle estimation information and the support angle interval to obtain the user distance estimation information, the support angle interval width and the receiving power ratio are first determined based on the support angle interval, wherein the support angle interval width represents the difference between the maximum spatial angle and the minimum spatial angle within the support angle interval; then the user distance estimation information is obtained based on the receiving power ratio, the support angle interval width, the user angle estimation information and a preset distance calculation function; the user angle estimation information is determined based on the support angle interval, and the user distance estimation information is determined based on the support angle interval width.

[0100] It is worth noting that the user distance estimation information and the support angle interval width satisfy the distance calculation function relationship. Therefore, the received power ratio, support angle interval width and user angle estimation information can be input into the distance calculation function to obtain the user distance estimation information and reduce the beam training overhead.

[0101] In addition, in one embodiment, if Figure 6 As shown, the above step S430 may include but is not limited to step S433, step S434 and step S435.

[0102] Step S433, constructing a received signal power ratio according to all received signal powers within the support angle interval and the received signal power corresponding to the user angle estimation information;

[0103] Step S434, constructing a power ratio mean square error solution optimization problem according to the received signal power ratio and a preset power ratio comparison value;

[0104] Step S435 , solving the power ratio mean square error optimization problem based on a preset exhaustive search strategy to obtain user estimation information.

[0105] It should be noted that in the process of performing distance estimation processing based on user angle estimation information and support angle interval to obtain user distance estimation information, firstly, a received signal power ratio is constructed based on all received signal powers within the support angle interval and the received signal power corresponding to the user angle estimation information; then, a power ratio mean square error solution optimization problem is constructed based on the received signal power ratio and a preset power ratio comparison value; finally, the power ratio mean square error solution optimization problem can be solved based on a preset exhaustive search strategy to obtain user estimation information. Through the above technical solution, the overhead of beam training can also be reduced, and compared with step S431 to step S432, the determination of user distance can be more accurate.

[0106] It is worth noting that estimating the user distance based on the power of all received signals within the support angle interval makes the confirmation of the user distance more accurate, because in the process of range estimation based on the width of the support angle interval, inaccuracies are easily caused under low signal-to-noise ratio conditions.

[0107] In addition, in one embodiment, if Figure 7 As shown, after executing the above step S500, it may also include but not limited to step S610 and step S620.

[0108] Step S610, calculating and obtaining an optimal beamforming vector according to the polar domain codeword;

[0109] Step S620: Perform beamforming processing based on the optimal beamforming vector.

[0110] It should be noted that after determining the polar codeword according to the user angle estimation information and the user distance estimation information, the optimal beamforming vector can also be calculated according to the polar codeword; then beamforming processing is performed based on the optimal beamforming vector to prepare for subsequent data transmission.

[0111] In order to more clearly illustrate the specific process of the ultra-large-scale MIMO near-field beam training method provided by an embodiment of the present invention, a specific example is used below for illustration.

[0112] like Figure 8As shown in FIG, the downlink beam training of a narrowband ultra-large-scale array XL-array communication system is considered, where the BS is equipped with a uniform linear array (ULA) with N antennas, the user has a single antenna, and there are scatterers in the environment.

[0113] The XL-array is placed on the y-axis, and the coordinates of its nth antenna are (0,δ n d), where Based on the spherical wave transmission model, the line-of-sight (LoS) transmission channel from the BS to the user can be expressed as

[0114]

[0115] in, is the complex channel gain, r u represents the distance from the center of the BS to the user, β represents the channel gain per unit distance of 1m, and b H (θ u ,r u ) represents the near-field channel steering vector, defined as

[0116]

[0117] in, represents the spatial angle of the spatial departure angle (angle-of-departure, AoD) φ0, and Represents the distance from the nth antenna of the XL-array at the BS to the user. Similarly, r1 and θ1 are defined as the distance and angle between the BS and the near-field scatterer in the non-line-of-sight (NLoS) path. Based on the channel between the BS and the user in the above figure, the user's received signal is

[0118]

[0119] Where x represents the BS’s transmitted signal, v represents the BS’s transmit beamforming vector, and z represents the variance σ 2 If perfect channel state information (CSI) (θ u ,r u ), we can conclude that the optimal beamforming vector is v * =b(θ u ,r u ). Then, given the transmission power P, the maximum achievable rate of the communication system can be expressed as

[0120]

[0121] First, we use the far-field beamforming vector to perform beam scanning and observe and analyze the received beam pattern received at the user. Interestingly, it is shown that the received beam pattern received at the user in the near-field channel contains the user's angle and distance information. Let represents the far-field beamforming vector, where Ω∈[-1,1] represents the spatial angle in the entire angle domain. First, the following two definitions are given:

[0122] Definition 1: Given a channel steering vector b H (θ u ,r u ), define f(w,θ u ,r u )@|b H (θ u ,r u )w| is the normalized beam gain of the beamforming vector w.

[0123] Definition 2: Given a channel steering vector b H (θ u ,r u ), when the far-field beamforming vector When used for beam scanning, define is the support angle of βdB. Specifically, characterizes a spatial angular region whose corresponding normalized beam gain is above κf peak ,in represents the gain factor, f peak represents the peak normalized beam gain. Mathematically,

[0124]

[0125] Next, its support angle width is defined as

[0126] Γ βdB (θ u ,r u )=Ω right -Ω left ,

[0127] Among them, Ω left and Ω right Angle support The minimum and maximum spatial angles in , whose values ​​are in the range [-1,1].

[0128] like Fig. 9Receive beam patterns for far-field beam scanning at different user angles and distances, where the number of antennas is N=256 and the carrier frequency is f=30 GHz. The actual user spatial angle is marked by the vertical black dashed line.

[0129] exist Fig. 9 In the figure, different user angles and distances {(θ u ,r u )}, far-field beamforming The relationship between the normalized beam gain and the spatial angle Ω. This figure shows the change of beam gain with the change of spatial angle, and there is an important observation.

[0130] Observation 1:

[0131] 1) Relationship between user angle and support angle: The actual user space angle is approximately in the middle of the 3dB support angle, that is,

[0132]

[0133] Here Med means taking the median value of the variable.

[0134] 2) Relationship between user distance and support angle width: Given a user angle θ u , support angle width Γ 3dB (θ u ,r u ) As the user distance r u In addition, given the user distance, the support angle width Γ 3dB (θ u ,r u ) decreases with the increase of the absolute value of the spatial angle.

[0135] Observation 1 shows that the user angle can be estimated by determining the middle of the support angles. Furthermore, based on the estimated user angle, the user distance can be inferred from the support angle width, since the support angle width decreases monotonically with increasing user distance.

[0136] Solution 1: Angle Support Width based Joint Angle and Range Estimation (ASW-JE)

[0137] Theoretically, it can be proved that the angle support width is a function of the user's angle and distance. Therefore, this motivates us to first estimate the user's angle and then use the support angle width to solve for the user's distance. We call this scheme the joint angle and distance estimation scheme based on the support angle width.

[0138] Phase 1: Beam Scanning

[0139] The base station sends N training symbols in sequence and uses the predefined DFT codebook Dynamically adjust the beam direction (specified by the beam codeword). The signal power received by the user each time can be expressed as

[0140]

[0141] Among them, x represents the transmitted signal, z n is additive white Gaussian noise.

[0142] Stage 2: Jointly estimate angle and distance

[0143] Phase 1 can obtain the received signal power Then the user angle and distance can be estimated according to the support angle in the following order:

[0144] Angle estimation:

[0145] Known received signal power The user can determine the received signal power interval (i.e., support angle) to obtain the corresponding codeword index set, whose value is

[0146]

[0147] Among them, κ 2 ≈0.5, and its specific value depends on the receive beam pattern. Then, the index of the estimated user angle in the DFT codebook is here means rounding up, then the corresponding user angle estimate is

[0148]

[0149] Distance estimation:

[0150] Given an estimated user perspective First, we introduce a receiving power such as

[0151]

[0152] in, When z n = 0 and When θ n =θ u When , the approximation in (c2) holds. Therefore, in distance estimation, can be regarded as the power ratio G(θ u ,r u ,θ n ). Then, the user uses φ = θm To construct the support angle interval, that is

[0153]

[0154] Next, given the function G(θ u ,r u ,φ), which is defined as follows:

[0155]

[0156] Where C(·) and S(·) represent the Fresnel integrals and parameter

[0157]

[0158] Then, the user's distance can be estimated by the following formula:

[0159]

[0160] Among them, μ0 satisfies and

[0161] Phase 3: Beam Decision

[0162] After using the DFT codebook to jointly estimate the angle and distance, the user will Feedback to the BS. Then, the following method is applied to select the beamforming vector. First, it is assumed that the distance of the polar domain codeword can be flexibly controlled, that is, the angle of the polar domain codeword is obtained based on grid division (on-grid), while the distance domain is obtained based on the off-grid method. Therefore, given It is easy to obtain the optimal beamforming vector for data transmission as

[0163] In order to further improve the accuracy of user angle and range estimation, K>1 candidate user angles are often selected among the support angles, which can reduce the error of angle estimation more than selecting only one estimated angle. Then, for all candidate user angles and distances Another round of beam scanning is needed to select the best candidate user angle and range pair.

[0164] Solution 2: Power-Ratio MSE based Joint Angle and Range Estimation (prMSE-JE)

[0165] The performance of the ASW-JE scheme depends critically on the estimation accuracy of the support angle width, which may suffer considerable losses in low SNR situations. To address this problem, an optimization scheme is proposed. The scheme is called joint angle and distance estimation based on power ratio mean square error (prMSE-JE), which can greatly improve the accuracy of distance estimation. The key idea is to use the high SNR received signal power within the support angle interval to estimate the user distance by minimizing the MSE of the total power ratio, thereby effectively suppressing the influence of received noise to achieve more accurate distance estimation accuracy.

[0166] Phase 1: Beam Scanning

[0167] The base station uses the traditional DFT codebook Perform beam scanning and transmit N signals in sequence. The receiving end receives N signals accordingly, and the power of the corresponding nth signal is

[0168] Stage 2: Jointly estimate angle and distance

[0169] Based on the power of the received signal First, find the support angle range as follows Among them, κ 2 =0.5, which is the power threshold of the support angle interval.

[0170] Angle estimation:

[0171] Therefore, the user perspective can be expressed as The approximate value is given by express.

[0172] Distance estimation:

[0173] In order to solve the user distance, first define the received signal power ratio:

[0174]

[0175] in, It indicates the power of the signal received by the user at the receiving end. It is worth pointing out that Both are determined only by the user's distance. Therefore, the user distance is transformed into solving the following optimization problem:

[0176]

[0177] Z Fre ≤r u ≤Z Rayl ,

[0178] in, and represent the Fresnel distance and Rayleigh distance respectively. Then, the user distance is estimated by exhaustive search to solve problem (P1), and its value is expressed as express.

[0179] Beam Decision:

[0180] Similar to the ASW-JE scheme, after joint angle and range estimation using the DFT codebook, the off-grid polar codeword is used to select the best beam codeword for data transmission. In this case, the designed XL array beamforming vector is

[0181] Finally, the performance of the two proposed near-field beam training schemes and existing baseline schemes are compared in terms of training overhead and estimation accuracy.

[0182] For training cost:

[0183] Both proposed schemes require beam scanning based on DFT codebook and a small number of training symbols for medium K angle estimation and training. (1) =T (2) = N + K training symbols. This is faster than the two-stage beam training scheme T (2P) =N+KS and exhaustive beam search scheme T (ex) =NS is much smaller. For example, consider the case where N = 256, K = 3 and S = 5. Then T (1) =T (2) =259, which is much smaller than T (ex) =1280,T (2P) =271.

[0184] For estimated accuracy:

[0185] For angle estimation, the proposed two schemes achieve similar estimation accuracy as the two-stage near-field beam training scheme. For distance estimation, both proposed schemes can achieve more accurate distance estimation. In addition, compared with the ASW-JE scheme, the prMSE-JE scheme can achieve more accurate distance estimation. This is because the prMSE-JE scheme uses all received signal powers in the support angle interval to minimize the power ratio MSE, while the ASW-JE scheme only uses the support angle width.

[0186] Simulation analysis:

[0187] Normalized mean square error (NMSE) of angle estimation

[0188] like Fig.10As shown, this figure shows the relationship between the angle estimation NMSE and the reference SNR of different schemes.

[0189] First, under different K, the angle estimation NMSE of the two proposed schemes decreases with the increase of the reference signal-to-noise ratio (SNR) and approaches the performance of the two-stage beam training scheme, while the angle estimation NMSE of the exhaustive search beam training scheme remains unchanged. This is because under the DFT codebook, as the SNR increases, the received beam pattern at the user usually becomes more stable, and thus the angle estimation error decreases. In contrast, the exhaustive search beam training exploits the beam focusing property and selects the best polarization codeword that produces the maximum received signal power at the user, so its angle estimation NMSE is robust to the SNR.

[0190] Second, in the high SNR case, by slightly increasing the number of candidate angles to K = 3, the proposed ASW-JE and prMSE-JE schemes obtain much lower angle estimation NMSE and eventually achieve the same NMSE performance as the exhaustive search beam training. This is because selecting 3 candidate angles can further reduce the probability of angle estimation errors.

[0191] like Fig.11 As shown in the figure, the relationship between the distance estimation NMSE and the reference SNR of different schemes is shown. This figure is a curve of the relationship between the distance estimation NMSE and the reference SNR. First, the distance estimation NMSE of the two proposed schemes decreases with the increase of the reference SNR, and they are significantly better than the two benchmark beam training schemes in the high SNR case. This is because the two benchmark beam training schemes use predefined distance codewords to estimate the user range; therefore, their accuracy is determined by the number of sampled distances S. However, the two proposed schemes apply off-grid distance estimation, thereby achieving better accuracy under high SNR conditions.

[0192] Secondly, the proposed prMSE-JE and ASW-JE (K = 3) have higher distance accuracy than K = 1. This is because more candidate angles can obtain more accurate angles.

[0193] Finally, it is observed that the proposed prMSE-JE scheme performs better than the ASW-JE scheme. This is because the prMSE-JE scheme utilizes the power of all received signals within the support angle interval, while the ASW-JE scheme only utilizes the support angle width for range estimation, which may cause inaccurate distance estimation in low SNR conditions.

[0194] For achievable rate:

[0195] like Fig.12As shown in the figure, the relationship between the achievable rate and the reference SNR of different schemes is shown. In the figure, the given user distance r is compared. u =16m and the user angles are randomly distributed in the area [-1,1] for different schemes. First, it can be seen that the proposed ASW-JE and prMSE-JE schemes with K=3 outperform the exhaustive search beam training scheme in the high SNR range (e.g., 28-38dB). This is because both proposed schemes estimate the user distance in a grid-based manner.

[0196] Second, in low SNR conditions, the proposed ASW-JE and prMSE-JE schemes suffer a slight rate loss compared to the exhaustive search beam training scheme. This is because in low SNR conditions, the received beam pattern fluctuates violently, which may lead to errors in the angle estimation.

[0197] Finally, the proposed prMSE-JE scheme outperforms the proposed ASW-JE scheme due to the more accurate estimated user distances.

[0198] Based on the above technical solution, the embodiment of the present application first proposes two important observations, that is, when the traditional DFT codebook is used for beam scanning in the near-field channel, the beam receiving mode of the receiving end contains both user angle and distance information. This inspires the design of two new and efficient beam training schemes for jointly estimating user angles and distances. Specifically, first, in order to describe the received signal with relatively high power in the beam receiving mode, the support angle is defined. Secondly, both methods use the middle of the support angle interval to estimate the user angle. Then, the first method uses the width of the interval to estimate the user distance, while the second method uses the power of all signals in the interval to construct an optimization problem, and then finds the user's distance through an exhaustive search. It is worth noting that it shows that compared with the existing exhaustive beam search beam training scheme and the two-stage beam training scheme, the proposed new joint angle and distance estimation The two beam training schemes can solve the user distance in an off-grid manner, so the user distance estimation accuracy is higher. At the same time, the scheme can greatly reduce the beam training overhead and obtain better beamforming performance.

[0199] In some embodiments of the present invention, Fig.13 As shown, an embodiment of the present invention further provides a very large-scale MIMO near-field beam training device 10, the device comprising:

[0200] The first processing module 100 is used to obtain a discrete Fourier transform codebook;

[0201] The second processing module 200 is used to perform beam scanning according to a discrete Fourier transform codebook to obtain a user received signal power set;

[0202] The third processing module 300 is used to determine the support angle interval according to the user received signal power set;

[0203] A fourth processing module 400, configured to determine user angle estimation information and user distance estimation information according to the support angle interval;

[0204] The fifth processing module 500 is used to determine a polar field codeword according to the user angle estimation information and the user distance estimation information.

[0205] The specific implementation of the near-field beam training device 10 is basically the same as the specific implementation of the above-mentioned ultra-large-scale MIMO near-field beam training method, and will not be repeated here.

[0206] In some embodiments of the present invention, Fig.14 As shown, an embodiment of the present invention further provides an electronic device 700, including: a memory 720, a processor 710, and a computer program stored in the memory 720 and executable on the processor 710, wherein the processor 710 implements the ultra-large-scale MIMO near-field beam training method in the above embodiment when executing the computer program, for example, executing the above described Figure 1 Steps S100 to S500 of the method, Figure 2 Steps S210 to S230 of the method, Figure 3 Steps S310 to S330 of the method, Figure 4 Steps S410 to S430 of the method, Figure 5 Steps S431 to S432 of the method, Figure 6 Steps S433 to S435 and Figure 7 It should be noted that the electronic device in the embodiment of the present invention may be a base station.

[0207] In some embodiments of the present invention, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are executed by a processor or controller, for example, by a processor in the above-mentioned device embodiment, so that the above-mentioned processor can execute the ultra-large-scale MIMO near-field beam training method in the above-mentioned embodiment, for example, to execute the above-described Figure 1 Steps S100 to S500 of the method, Figure 2 Steps S210 to S230 of the method, Figure 3 Steps S310 to S330 of the method, Figure 4 Steps S410 to S430 of the method, Figure 5 Steps S431 to S432 of the method, Figure 6 Steps S433 to S435 and Figure 7 Method steps S610 to S620 in .

[0208] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0209] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present invention. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A method for ultra-large-scale MIMO near-field beam training, characterized in that: The method comprises: Obtain a discrete Fourier transform codebook; Performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set; Determining a support angle interval according to the user received signal power set; Determining user angle estimation information and user distance estimation information according to the support angle interval; Determine a polar field codeword according to the user angle estimation information and the user distance estimation information; The step of performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set includes: Performing beam direction adjustment processing on a plurality of training signals transmitted by a base station according to the discrete Fourier transform codebook; Based on the plurality of training signals processed after receiving the beam direction adjustment at the user end, and determining the user received signal power from each of the training signals; Determining the plurality of user received signal powers as the user received signal power set; The determining the support angle interval according to the user received signal power set includes: Comparing the received signal power of each user in the received signal power set with a preset signal power threshold, wherein the signal power threshold is determined by a receiving beam mode of the user terminal; Filtering out a subset of user received signal powers greater than a preset signal power threshold from the user received signal power set; A codeword index set is determined according to the user received signal power subset, and the support angle interval is determined based on the codeword index set.

2. The ultra-large-scale MIMO near-field beam training method according to claim 1, characterized in that: The determining the user angle estimation information and the user distance estimation information according to the support angle interval includes: Determine the median value of the support angle interval as a user angle index; Determine the user angle estimation information according to the user angle index and the discrete Fourier transform codebook; The user distance estimation information is obtained by performing distance estimation processing according to the user angle estimation information and the support angle interval.

3. The ultra-large-scale MIMO near-field beam training method according to claim 2, characterized in that: The performing distance estimation processing according to the user angle estimation information and the support angle interval to obtain the user distance estimation information includes: Determine the support angle interval width and the received power ratio according to the support angle interval; wherein the support angle interval width represents the difference between the maximum spatial angle and the minimum spatial angle in the support angle interval; The user distance estimation information is obtained according to the received power ratio, the support angle interval width, the user angle estimation information and a preset distance calculation function.

4. The ultra-large-scale MIMO near-field beam training method according to claim 2, characterized in that: The performing distance estimation processing according to the user angle estimation information and the support angle interval to obtain the user distance estimation information includes: Constructing a received signal power ratio according to all received signal powers within the support angle interval and the received signal power corresponding to the user angle estimation information; Constructing a power ratio mean square error solution optimization problem according to the received signal power ratio and a preset power ratio comparison value; The power ratio mean square error solution optimization problem is solved based on a preset exhaustive search strategy to obtain the user distance estimation information.

5. The ultra-large-scale MIMO near-field beam training method according to claim 1, characterized in that: After determining the polar field codeword according to the user angle estimation information and the user distance estimation information, the method further includes: Calculating an optimal beamforming vector according to the polar domain codeword; A beamforming process is performed based on the optimal beamforming vector.

6. A very large-scale MIMO near-field beam training device, characterized in that: include: A first processing module, used for obtaining a discrete Fourier transform codebook; A second processing module, configured to perform beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set; A third processing module, configured to determine a support angle interval according to the user received signal power set; a fourth processing module, configured to determine user angle estimation information and user distance estimation information according to the support angle interval; A fifth processing module, configured to determine a polar field codeword according to the user angle estimation information and the user distance estimation information; The step of performing beam scanning according to the discrete Fourier transform codebook to obtain a user received signal power set includes: Performing beam direction adjustment processing on a plurality of training signals transmitted by a base station according to the discrete Fourier transform codebook; Based on the plurality of training signals processed after receiving the beam direction adjustment at the user end, and determining the user received signal power from each of the training signals; Determining the plurality of user received signal powers as the user received signal power set; The determining the support angle interval according to the user received signal power set includes: Comparing the received signal power of each user in the received signal power set with a preset signal power threshold, wherein the signal power threshold is determined by a receiving beam mode of the user terminal; Filtering out a subset of user received signal powers greater than a preset signal power threshold from the user received signal power set; A codeword index set is determined according to the user received signal power subset, and the support angle interval is determined based on the codeword index set.

7. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the ultra-large-scale MIMO near-field beam training method as described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium storing computer-executable instructions, characterized in that: When the computer executable instructions are executed by the control processor, the ultra-large-scale MIMO near-field beam training method as described in any one of claims 1 to 5 is implemented.

Images