Method, apparatus, and medium for beam selection

Abstract

The application discloses a beam selection method and device, electronic equipment and medium. Through the application of the technical solution, the base station device can automatically predict the signal strength values of all narrow beams in the direction of the terminal according to a preset network prediction model, and select multiple candidate beams from the adjacent transmission directions of the narrow beams with the highest signal strength values, so as to select the final target beam according to the signal strength values of each candidate beam. In this way, on the one hand, the problem of inaccurate selection accuracy that often occurs when selecting the optimal beam from a large number of narrow beams in a multi-antenna system is avoided. On the other hand, since a large amount of selection work is obtained by calculation and prediction, the number of measurement beams actually transmitted by the base station is greatly reduced, thereby achieving the purpose of reducing signaling overhead.

Description

TECHNICAL FIELD

[0001] The present application relates to signal data processing technology, and in particular to a beam selection method and device, electronic equipment and medium. BACKGROUND

[0002] In future mobile communication systems, high-frequency band communication represented by millimeter waves is considered as one of the key technologies of 5G due to its rich spectrum resources.

[0003] In related technologies, since millimeter waves have characteristics such as high carrier frequency and short wavelength, the signal is severely attenuated during transmission, which also causes large path loss during transmission. Therefore, it is usually combined with massive multiple antenna (Massive MIMO) technology.

[0004] However, compared with traditional multiple antenna technology, the massive multiple antenna system based on millimeter waves will form a larger number of narrower beams. Therefore, how to select a target beam from a large number of beams in an actual service scenario has become a problem to be solved by those skilled in the art. SUMMARY

[0005] The embodiments of the present application provide a beam selection method, device, electronic equipment and medium. Thus, the problem of inaccurate selection accuracy when selecting an optimal beam from a large number of narrow beams in a multiple antenna system in related technologies is solved.

[0006] According to an aspect of the embodiments of the present application, a beam selection method is provided, applied to a base station device, comprising:

[0007] predicting a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by a target terminal;

[0008] inputting the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of a plurality of narrow beams transmitted by the base station;

[0009] selecting a candidate beam matching a target narrow beam from all narrow beams corresponding to the base station, the target narrow beam being a narrow beam with the highest second signal strength value;

[0010] selecting a target beam from the candidate beam.

[0011] Optionally, in another embodiment based on the above-mentioned method of the present application, the step of predicting a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by a target terminal comprises:

[0012] predicting a signal strength value of each sounding beam transmitted by the base station according to the current channel state information and a preset sounding beam codebook matrix;

[0013] a vector composed of the signal strength values of each sounding beam is taken as the first signal strength value

[0014] Optionally, in another embodiment based on the above method of the present application, the selecting, from all the narrow beams corresponding to the base station, the candidate beams matching the target narrow beam comprises:

[0015] determining a transmission direction of the target narrow beam;

[0016] taking a region within a preset angle range from the transmission direction as a candidate region;

[0017] taking, from all the narrow beams transmitted by the base station, the narrow beams with transmission directions in the candidate region as the candidate beams.

[0018] Optionally, in another embodiment based on the above method of the present application, the selecting, from the candidate beams, the target beam comprises:

[0019] respectively transmitting each candidate beam to the target terminal and receiving a third signal strength value of each candidate beam fed back by the target terminal;

[0020] taking the candidate beam with the highest third signal strength value as the target beam.

[0021] Optionally, in another embodiment based on the above method of the present application, before the predicting the first signal strength values of the multiple sounding beams transmitted by the base station according to the current channel state information fed back by the target terminal, the method further comprises:

[0022] transmitting a training signal to a training terminal and receiving training channel state information fed back by the training terminal based on the training signal; and predicting first training signal strength values of multiple training sounding beams transmitted by the base station according to the training channel state information;

[0023] and,

[0024] transmitting multiple narrow beams to the training terminal and receiving second training signal strength values corresponding to the multiple narrow beams fed back by the training terminal;

[0025] training the network prediction model based on the first training signal strength values and the second training signal strength values.

[0026] Optionally, in another embodiment based on the above method of the present application, the training of the network prediction model based on the first training signal strength value and the second training signal strength value comprises:

[0027] normalizing the first training signal strength value to obtain training input data, and normalizing the second training signal strength value to obtain training output data;

[0028] iteratively training an initial network prediction model using the training input data and the training output data until the network prediction model is obtained.

[0029] According to yet another aspect of embodiments of the present application, a device for beam selection is provided, which is applied to a base station device and comprises:

[0030] a first prediction module configured to predict a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by a target terminal;

[0031] a second prediction module configured to input the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of a plurality of narrow beams transmitted by the base station;

[0032] a matching module configured to select, from all narrow beams corresponding to the base station, a candidate beam matching a target narrow beam, the target narrow beam being a narrow beam with the highest second signal strength value;

[0033] a selection module configured to select a target beam from the candidate beam.

[0034] According to yet another aspect of embodiments of the present application, an electronic device is provided, which comprises:

[0035] a memory configured to store executable instructions; and

[0036] a display configured to execute the executable instructions with the memory to complete the operations of any of the above-described methods for beam selection.

[0037] According to yet another aspect of embodiments of the present application, a computer readable storage medium is provided, which is configured to store computer readable instructions, the instructions being executed to perform the operations of any of the above-described methods for beam selection.

[0038] The technical solutions of the present application will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and, together with the description, serve to explain the principles of the application.

[0040] The application can be more clearly understood and appreciated from the following detailed description, taken in conjunction with the accompanying drawings of which:

[0041] Figure 1 A schematic diagram of a beam selection method according to the present application;

[0042] Figure 2 A schematic diagram of a beam selection method according to the present application;

[0043] Figure 3 A schematic diagram of a beam selection method according to the present application;

[0044] Figure 4 A schematic diagram of a beam selection method according to the present application;

[0045] Figure 5 A schematic diagram of a beam selection method according to the present application;

[0046] Figure 6 A schematic diagram of a beam selection method according to the present application;

[0047] Figure 7 A schematic diagram of a beam selection method according to the present application; DETAILED DESCRIPTION

[0048] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangements, numerical expressions, and numerical values of the components and steps set forth in these embodiments are not limiting to the scope of the present application unless otherwise specifically stated.

[0049] It should be understood, however, that the actual dimensions of the parts shown in the accompanying drawings can not be drawn to scale.

[0050] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way limiting to the scope of the application or its applications or uses.

[0051] Techniques, methods, and devices known to those of ordinary skill in the relevant art can not be discussed in detail herein. However, such techniques, methods, and devices should be construed as being a part of the specification to the full extent that they are used herein and described herein.

[0052] It should be noted that like reference numerals and characters refer to like elements throughout the following particular description and the appended drawings are not necessarily to scale, and that certain ones of the figures can be simplified for the sake of clarity.

[0053] In addition, the technical solutions among various embodiments of the present application can be combined with each other, but it must be based on the fact that a person skilled in the art can realize it, and when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist and is not within the protection scope required by the present application.

[0054] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative positional relationship, motion condition, etc. between components in a certain posture (as shown in the drawings), and if the certain posture changes, the directional indications will also change accordingly.

[0055] The method for performing beam selection according to the exemplary embodiments of the present application will be described below in conjunction with Figures 1-5 It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this respect. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0056] The present application also proposes a method and device for beam selection, an electronic device and a medium.

[0057] Figure 1 A flowchart of a method for beam selection according to an embodiment of the present application is schematically shown. As Figure 1 shown, the method is applied to a base station device and includes:

[0058] S101, predicting a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by a target terminal.

[0059] S102, inputting the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of a plurality of narrow beams transmitted by the base station.

[0060] S103, selecting a candidate beam matching a target narrow beam from all narrow beams corresponding to the base station, the target narrow beam being a narrow beam with the highest second signal strength value.

[0061] S104, selecting a target beam from the candidate beam.

[0062] In the related art, the beam alignment methods in the existing multi-antenna system can be roughly divided into the following categories:

[0063] First, beam alignment based on beam scanning, in this method, the base station transmits beams pointing to different angles in different time slots, the user receives the beams transmitted by the base station in different time slots respectively, and feeds back the signal strength, and the base station selects the beam with the highest strength for data transmission.

[0064] Second, beam alignment based on training signals, in this method, the base station side sends several known training signals to the user, the user side performs downlink channel estimation based on the received training signals, and feeds back the downlink channel state information to the base station, and the base station calculates the corresponding downlink beamforming vector according to the received downlink channel state information, so as to form a beam pointing to the user, and achieve the purpose of beam alignment.

[0065] Third, beam alignment based on position prediction, in this method, the position coordinates of the user are estimated and predicted through the positioning related reference signal, and the corresponding beamforming direction is obtained based on the geometric position relationship between the base station and the user, so as to realize beam alignment.

[0066] Fourth, beam alignment based on hierarchical search, in this method, the base station first activates a part of antenna elements to transmit several wide beams, and finds out the optimal wide beam in them, and then searches narrow beams within the spatial range of the wide beam.

[0067] However, the above methods have certain limitations when applied to millimeter wave Massive MIMO system. In the actual wireless propagation environment, there are often various obstructions, and after the beam is narrowed, the signal transmission is more easily affected by the obstruction; and the significant increase in the number of beams significantly increases the beam search amount in the beam alignment process, making it difficult to accurately select the optimal beam in a short time.

[0068] In addition, at the same time, the beam alignment based on the beamforming vector depends on the estimation and feedback of the channel state information, and the influence of noise and interference in the actual system will cause channel estimation error, thereby reducing the accuracy of beam alignment.

[0069] Further, in order to solve the above problems, the present application proposes a beam selection method, which is designed to design a more reasonable millimeter wave multi-antenna system beam alignment method and process for the actual communication environment, aiming at the situation that the signal propagation is easily affected by obstacles (such as surrounding buildings, signs, vehicles, etc.) and noise, and aims to ensure the accuracy of beam alignment while reducing the time delay of beam search. To solve the low time delay and high robustness beam alignment problem of a large number of narrow beams in a high frequency large-scale multi-antenna system.

[0070] Further, the present application is applied to a base station device, and hereinafter Figure 2 The scheme will be described in detail:

[0071] Step 1, sending a training signal to a training terminal, and receiving training channel state information fed back by the training terminal based on the training signal.

[0072] Firstly, the embodiment of the present application can control the base station device to send a plurality of preset training signals P1 to the terminal side. So that the terminal side performs downlink channel estimation based on the received training signal P1, determines the downlink channel state information (i.e. training channel state information) according to the downlink channel estimation result, and feeds back the information to the base station device.

[0073] Step 2, predicting the first training signal strength value of the base station sending a plurality of training probe beams according to the training channel state information.

[0074] In one way, the embodiment of the present application can configure the number of training probe beams as N, and determine the activated adjacent antenna element interval and the preset probe beam codebook matrix according to N. For example, if N is 16, 16 antennas can be activated from all the antenna elements to emit training probe beams.

[0075] Wherein, N is less than or equal to the current number of all antenna elements.

[0076] In other words, the probe beam is emitted by the corresponding equidistant activated part or all antenna elements of the base station device. The base station side calculates the probe beam signal strength value received by the terminal device side when the base station device emits the probe beam according to the training channel state information obtained in step 1 and the probe beam codebook.

[0077] In one way, the training signal strength value can be represented by a vector value. For example, it can be recorded as: x = [x1, x2, …, x N ].

[0078] Step 3, sending a plurality of narrow beams to the training terminal, and receiving a plurality of second training signal strength values corresponding to the narrow beams fed back by the training terminal.

[0079] In one way, the embodiment of the present application can calculate the number of corresponding narrow beams when all the antenna elements of the base station side are activated according to the number of antenna elements, denoted as M (i.e. the narrow beam is generated by the base station activating all the antenna elements). The number is 1~M.

[0080] Further, the base station device emits the corresponding numbered narrow beam to the terminal side in M time slots (i.e. sequentially emits). So that the terminal side statistics the narrow beam signal strength received in each time slot to obtain the narrow beam signal strength probe vector (i.e. the second training signal strength value), denoted as: y = [y1, y2, …, y M ], and feeds back to the base station.

[0081] Step 4, normalizing the first training signal strength value to obtain training input data, and normalizing the second training signal strength value to obtain training output data.

[0082] Further, the embodiment of the present application can repeat the above steps for a large number of users in different positions, thereby obtaining a plurality of monitoring and statistical detection beam signal strength estimation vectors (i.e. the first training signal strength value) and narrow beam signal strength detection vectors (i.e. the second training signal strength value) of each user, which are used as training data sets for training the network prediction model.

[0083] Step 5, iteratively training the initial network prediction model using the training input data and the training output data until the network prediction model is obtained.

[0084] In one way, in combination with Figure 3 As shown in the figure, the embodiment of the present application can set the total amount of data samples required for training the network prediction model as S, and use the detection beam signal strength estimation vector received by the terminal side in the training data set as the input feature of the network prediction model, and use the narrow beam signal strength detection vector y received by the user in the training data set as the output feature of the network prediction model.

[0085] Further, the initial network prediction model can also be configured with weight parameters, bias parameters, the number of hidden layers, activation functions between layers, and other parameter data, and the fully connected neural network is trained based on the pre-set total amount of data samples S to obtain the corresponding relationship between the input and the output, thereby obtaining the network prediction model.

[0086] Step 6, predicting the signal strength value of each detection beam transmitted by the base station according to the current channel state information fed back by the target terminal and the pre-set detection beam codebook matrix.

[0087] In the specific implementation phase, for example, at a certain moment, the base station device can send a plurality of training signals to the target terminal, so that the target terminal performs downlink channel estimation based on the received training signals, thereby determining the current channel state information at the current moment according to the downlink channel estimation result, and feeding back the information to the base station device.

[0088] Further, the base station predicts and calculates the current estimation vector of the plurality of detection beam signal strengths received by the terminal side (i.e. the first signal strength value) when the base station transmits a plurality of detection beams according to the detection beam codebook obtained in the beam training phase and the current channel state information. For example, it can be recorded as

[0089] It should be noted that the base station device in the embodiment of the present application does not need to actually send the probe beam to the target terminal. Instead, the base station device predicts the signal strength value of the terminal receiving the probe beam if the base station device sends the probe beam to the target terminal according to the current channel state information and the preset probe beam codebook (which records the correlation between the channel state and the probe beam signal strength), that is, the base station side estimates the signal strength of the probe beam received by the terminal side.

[0090] Step 7, the vector composed of the signal strength values of each probe beam is taken as the first signal strength value.

[0091] Step 8, input the first signal strength value into the pre-trained network prediction model to predict the second signal strength value of the base station sending multiple narrow beams.

[0092] In the embodiment of the present application, the narrow beam refers to the beam formed by activating all antenna elements, and the narrow beam corresponding to the maximum value in the narrow beam signal strength prediction vector output by the network prediction model is defined as the target narrow beam. The beam obtained by locally searching the target narrow beam is the target beam of the present application. As an example, the network prediction model is a deep neural network with multiple hidden layers.

[0093] In one way, the base station side can train the deep neural network by configuring the parameters such as weight, bias, number of hidden layers, and activation function, based on a large number of probe beam signal strength estimation vectors and narrow beam signal strength detection vectors, so that the neural network prediction value gradually converges to the narrow beam signal strength detection value.

[0094] In one way, the network prediction model in the embodiment of the present application is deployed on the base station device side, and is used to realize the prediction function of the narrow beam received signal strength by the method of artificial intelligence model training.

[0095] Further, the embodiment of the present application can take the first signal strength value x * As the input of the network prediction model, the corresponding relationship Y=f(X) between the input and output of the network prediction model obtained in the beam training stage is used to predict the received signal strength of all narrow beams formed by the base station, and output the current narrow beam signal strength prediction vector (i.e. the second signal strength value). For example, Y * , that is, Y * =f(X * ).

[0096] Step 9, the base station selects the candidate beam matching the target narrow beam from all the narrow beams corresponding to the base station.

[0097] Among them, the target narrow beam is the narrow beam with the highest second signal strength value.

[0098] In one mode, the embodiment of the present application needs to select the narrow beam with the maximum received signal strength value in the output result as the target narrow beam.

[0099] Further, the present application can select the spatial region with an included angle less than or equal to θ with the direction pointed by the target narrow beam as the candidate region, and determine the narrow beam in the candidate region as the candidate beam.

[0100] For example, taking θ as 40 degrees, after determining the target narrow beam, the present application can first determine the transmission direction of the target narrow beam, and offset 20 degrees (i.e. the included angle between the two sides is 40 degrees) from the transmission direction.

[0101] It can be understood that the spatial region of 40 degrees is the candidate region, and further, the beam with all the transmission points located in the candidate region is the candidate beam.

[0102] Step 10, respectively send each candidate beam to the target terminal, and receive the third signal strength value of each candidate beam fed back by the target terminal.

[0103] Step 11, select the candidate beam with the highest third signal strength value as the target beam.

[0104] Further, the base station device needs to respectively transmit the above candidate beams to the terminal side in turn in several time slots (for example, transmit candidate beam 1, candidate beam 2, candidate beam 3 in turn). So that the terminal side statistics all the received signal strength of the candidate beams, select the beam with the maximum received signal strength (i.e. the highest third signal strength value) as the target beam, and report the number of the optimal beam to the base station, complete the beam selection.

[0105] It can be understood that the technical solution proposed by the present application is divided into a beam training phase (used to obtain a network prediction model) and a beam selection phase (used to select a target beam).

[0106] In the beam training phase, a large number of user samples at different positions can be used to train the network prediction model. Specifically, the embodiment of the present application can obtain the network prediction model through offline learning. This module establishes the performance correspondence between all the probe beams and the narrow beams.

[0107] In one mode, the input feature during training is the signal strength value of the probe beam calculated by the base station side according to the channel state information and the probe beam codebook. And the output feature is the signal strength value of the narrow beam obtained by the base station sending the narrow beam to the user according to the time slot.

[0108] In addition, in the beam selection phase, the base station device can perform prediction through the network prediction model obtained in the beam training phase, and complete the beam alignment by combining local search.

[0109] It can be understood that the technical scheme provided in the application can not only calculate the signal strength of the probe beam by using the estimated channel information, but also avoid generating a large amount of time delay by probing all beams, and master rough beam information in the initial stage.

[0110] In addition, the embodiment of the application considers the clustering characteristics of the millimeter wave channel, and the performance of the surrounding beams of the optimal beam is often better. The local search plays the advantage of the signal alignment technology in accuracy, and the beam around the target narrow beam is actually searched, which can further improve the beam alignment accuracy. At the same time, the application only actually searches a part of the beams, which can also achieve the purpose of reducing the time delay.

[0111] Furthermore, the application introduces a network prediction model, and finds the target narrow beam through data driving and model training, so that many steps can be directly calculated on the base station side to reduce a large amount of signaling overhead.

[0112] Again, the input parameter of the network prediction model in the application during training is an estimated signal strength vector of the probe beam, which is an estimated value calculated by the channel matrix with channel estimation error, and the output parameter is a narrow beam signal strength detection vector, which is a real value actually detected. This can weaken the influence of channel estimation error on beam alignment to a certain extent, so that the scheme still has high beam alignment accuracy when the signal-to-noise ratio of the system environment is low.

[0113] Finally, the received signal strength of the probe beam in the application can also play a role in detecting the surrounding channel geometry environment, which can provide a better reference for the selection of the narrow beam when there are obstacles in the communication environment.

[0114] In the application, the base station can predict first signal strength values of multiple probe beams transmitted by the base station according to the current channel state information fed back by the target terminal; input the first signal strength values into a pre-trained network prediction model to predict second signal strength values of multiple narrow beams transmitted by the base station; select candidate beams matching the target narrow beam from all narrow beams corresponding to the base station, the target narrow beam being the narrow beam with the highest second signal strength value. The target beam is selected from the candidate beams. By applying the technical scheme of the application, the base station device can automatically predict the signal strength values of all narrow beams in the direction of the terminal according to the preset network prediction model, and select multiple candidate beams from the narrow beams adjacent to the narrow beam with the highest signal strength value, so as to select the final target beam according to the signal strength value of each candidate beam. In this way, on the one hand, the problem of inaccurate selection of the optimal beam from a large number of narrow beams in a multi-antenna system is avoided. On the other hand, a large amount of selection work is obtained by calculation and prediction, which also leads to a significant reduction in the actual measurement beams transmitted by the base station, thereby achieving the purpose of reducing the signaling overhead.

[0115] Optionally, in another embodiment based on the above method of the present application, the step of predicting the first signal strength value of the plurality of sounding beams transmitted by the base station according to the current channel state information fed back by the target terminal comprises:

[0116] predicting the signal strength value of each sounding beam transmitted by the base station according to the current channel state information and a preset sounding beam codebook matrix;

[0117] taking the vector composed of the signal strength value of each sounding beam as the first signal strength value.

[0118] Optionally, in another embodiment based on the above method of the present application, the step of selecting the candidate beam matching the target narrow beam from all the narrow beams corresponding to the base station comprises:

[0119] determining the transmission direction of the target narrow beam;

[0120] taking the area within the preset angle range from the transmission direction as the candidate area;

[0121] taking the narrow beam whose transmission direction is in the candidate area from all the narrow beams transmitted by the base station as the candidate beam.

[0122] Optionally, in another embodiment based on the above method of the present application, the step of selecting the target beam from the candidate beam comprises:

[0123] respectively transmitting each candidate beam to the target terminal and receiving the third signal strength value of each candidate beam fed back by the target terminal;

[0124] taking the candidate beam with the highest third signal strength value as the target beam.

[0125] Optionally, in another embodiment based on the above method of the present application, before the step of predicting the first signal strength value of the plurality of sounding beams transmitted by the base station according to the current channel state information fed back by the target terminal, the method further comprises:

[0126] transmitting a training signal to a training terminal and receiving the training channel state information fed back by the training terminal based on the training signal; and predicting the first training signal strength value of the plurality of training sounding beams transmitted by the base station according to the training channel state information;

[0127] and,

[0128] transmitting a plurality of narrow beams to the training terminal and receiving the second training signal strength value corresponding to the plurality of narrow beams fed back by the training terminal;

[0129] Based on the first training signal strength value and the second training signal strength value, the network prediction model is trained.

[0130] Optionally, in another embodiment of the method described above, based on the first training signal strength value and the second training signal strength value, the network prediction model is trained, comprising:

[0131] The first training signal strength value is normalized to obtain training input data, and the second training signal strength value is normalized to obtain training output data.

[0132] The initial network prediction model is iteratively trained using the training input data and the training output data until the network prediction model is obtained.

[0133] Hereinafter, a beam selection method according to the present application is described in detail:

[0134] For example, the base station device side in the present application can use a 4x8 uniform planar antenna array, and the target terminal side uses a single antenna. Moreover, the number of probe beams is configured to be 8, and the DFT codebook matrix v = [v1, v2,..., v8] of 8 probe beams generated by calculating the equal-interval active 8 antenna elements is calculated.

[0135] As an example, the calculation formula of the codebook matrix is as follows:

[0136] The codebook matrix of the UPA can be obtained from the codebook matrices of two ULAs, i.e.

[0137]

[0138] wherein, and ULA codebook matrices of horizontal direction antennas and vertical direction antennas, respectively, and the calculation formula is as follows:

[0139]

[0140] wherein

[0141] In one way, the process of training the network prediction model is described as follows:

[0142] Step a, the target base station sends a pre-set Zad-off Chu sequence as a training signal to the target terminal in the cell, and the target terminal performs downlink channel estimation based on the received training signal with the MMSE algorithm to obtain training channel state information, denoted as H, and feeds back H to the target base station.

[0143] Step b, set the target base station transmit power PT = 40W, the target base station side calculates the received signal strength value of the target terminal according to the formula The received signal strength value of the target terminal when transmitting the probe beam i (i = 1, 2, …, 8) is calculated, and the first training signal strength value x = [x1, x2, …, x8] of the probe beam is obtained.

[0144] Step c, set the narrow beam number to 1-32, the target base station transmits the corresponding narrow beam to the target terminal in the next 32 time slots, and the target terminal statistics the received signal strength of the 32 narrow beams, and obtains the second training signal strength value y = [y1, y2, …, y32] corresponding to the narrow beam, and feeds back the strength value to the target base station. 32

[0145] Step d, for the target terminal in the cell, repeat the above steps, monitor and statistics the first training signal strength value and the second training signal strength value of each target terminal at different positions, and take them as the training data set of the network prediction model.

[0146] Step e, set the total amount of data samples required for training the network prediction model to 50000, normalize the 50000 first training signal strength values and second training signal strength values obtained in step d, as the input and output of the network prediction model, build a neural network, and train the network prediction model by configuring the weight parameters, bias parameters, number of hidden layers, and activation function between layers.

[0147] In another way, the process of selecting the target beam is described as follows:

[0148] Step f, the target base station sends the pre-set Gold sequence to the target terminal in the cell as the training signal, the target terminal performs downlink channel estimation based on the received training signal, determines the current channel state information according to the downlink channel estimation result, obtains the channel matrix H, and feeds back H to the target base station.

[0149] Step g, the target base station side calculates the received signal strength of the target terminal when transmitting the probe beam i (i = 1, 2, …, 8) according to the formula The first signal strength value x = [x1, x2, …, x8] corresponding to the probe beam is obtained.

[0150] Step h, input the normalized result x * to the network prediction model obtained in the beam training stage to obtain the second signal strength value y * predicted by the current narrow beam signal strength, and select the narrow beam corresponding to the maximum value in y * as the target narrow beam.

[0151] ​Step i, configure a preset angle range θ, for example θ = 40°, select a conical space region with an included angle less than or equal to 40° with the direction in which the target narrow beam (for example, No. 17 beam) points as a candidate region, and determine the narrow beams in the candidate region as candidate beams.

[0152] For example, among all the transmitting beams of the base station device, the beams whose transmitting positions fall within the candidate region are No. 9, 10, 17, 18, 25, and 26 beams, and these beams are determined as candidate beams.

[0153] Step g, the target base station transmits the above candidate beams (i.e., No. 9, 10, 17, 18, 25, and 26 beams) to the target terminal in six consecutive time slots, and the target terminal counts the third signal strength values of all candidate beams.

[0154] For example, the candidate beam with the largest received signal strength is the third one, i.e., No. 17 beam, and No. 17 beam is determined as the target beam, and its number 17 is reported to the target base station, thereby completing beam alignment.

[0155] For example, the selection process of the above optimal beam is as shown in Figure 5 . The grid region of the first step represents the target narrow beam; the grid region of the second step is the candidate region, and the beams in the grid region are polled and searched, and the finally obtained grid region represents the target beam.

[0156] By applying the technical solution of the present application, the base station device can automatically predict the signal strength values of all narrow beams in the direction of the terminal according to a preset network prediction model, and select multiple candidate beams from the narrow beams adjacent to the narrow beam with the highest signal strength value, so as to select the final target beam according to the signal strength value of each candidate beam. In this way, on the one hand, the problem of inaccurate selection accuracy often occurs when selecting the optimal beam from a large number of narrow beams in a multi-antenna system is avoided. On the other hand, since a large amount of selection work is obtained by calculation and prediction, the number of measurement beams actually transmitted by the base station is greatly reduced, thereby achieving the purpose of reducing signaling overhead.

[0157] Optionally, in another embodiment of the present application, as shown in Figure 6 , the present application also provides a beam selection device. The device comprises:

[0158] A first prediction module 201 configured to predict the first signal strength values of the plurality of probe beams transmitted by the base station according to the current channel state information fed back by the target terminal;

[0159] The second prediction module 202 is configured to input the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of the plurality of narrow beams transmitted by the base station.

[0160] The matching module 203 is configured to select a candidate beam matching the target narrow beam from all the narrow beams corresponding to the base station, the target narrow beam being the narrow beam with the highest second signal strength value.

[0161] The selection module 204 is configured to select a target beam from the candidate beams.

[0162] By applying the technical solution of the present application, the base station device can automatically predict the signal strength values of all the narrow beams in the direction of the terminal according to the pre-set network prediction model, and select a plurality of candidate beams from the narrow beams adjacent to the narrow beam with the highest signal strength value, so as to select a final target beam according to the signal strength value of each candidate beam. In this way, on the one hand, the problem of inaccurate selection precision in the related art when selecting an optimal beam from a large number of narrow beams in a multi-antenna system is avoided. On the other hand, since a large amount of selection work is obtained by calculation and prediction, the number of measurement beams actually transmitted by the base station is greatly reduced, thereby achieving the purpose of reducing signaling overhead.

[0163] In another embodiment of the present application, the matching module 203 is configured to:

[0164] predict a signal strength value of each sounding beam transmitted by the base station according to the current channel state information and a pre-set sounding beam codebook matrix;

[0165] compose a vector of the signal strength values of each sounding beam as the first signal strength value.

[0166] In another embodiment of the present application, the matching module 203 is configured to:

[0167] select a region within a pre-set angle range from the transmission direction as a candidate region;

[0168] select, from all the narrow beams transmitted by the base station, a narrow beam with a transmission direction in the candidate region as the candidate beam.

[0169] In another embodiment of the present application, the matching module 203 is configured to:

[0170] transmit each candidate beam to the target terminal respectively and receive a third signal strength value of each candidate beam fed back by the target terminal;

[0171] The third signal strength value of the candidate beam with the highest signal strength is taken as the target beam.

[0172] In another embodiment of the present application, the matching module 203 is configured to:

[0173] sending a training signal to a training terminal and receiving training channel state information fed back by the training terminal based on the training signal; and predicting first training signal strength values of a plurality of training probe beams sent by the base station based on the training channel state information;

[0174] and,

[0175] sending a plurality of narrow beams to the training terminal and receiving second training signal strength values corresponding to the plurality of narrow beams fed back by the training terminal;

[0176] training the network prediction model based on the first training signal strength values and the second training signal strength values.

[0177] In another embodiment of the present application, the matching module 203 is configured to:

[0178] normalizing the first training signal strength values to obtain training input data; and normalizing the second training signal strength values to obtain training output data;

[0179] iteratively training an initial network prediction model using the training input data and the training output data until the network prediction model is obtained.

[0180] Figure 7 is a logical structure block diagram of an electronic device according to an exemplary embodiment. For example, the electronic device 300 can be an electronic device.

[0181] In the example embodiment, a non-transitory computer readable storage medium, such as a memory including instructions, is also provided. The instructions can be executed by a processor of an electronic device to perform the method of beam selection, which includes: predicting a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by the target terminal; inputting the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of a plurality of narrow beams transmitted by the base station; selecting a candidate beam matching a target narrow beam from all narrow beams corresponding to the base station, the target narrow beam being the narrow beam with the highest second signal strength value; and selecting a target beam from the candidate beam. Optionally, the instructions can also be executed by the processor of the electronic device to perform other steps involved in the example embodiment. For example, the non-transitory computer readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0182] In the example embodiment, an application program / computer program product is also provided, which includes one or more instructions that can be executed by a processor of an electronic device to perform the method of beam selection, which includes: predicting a first signal strength value of a plurality of probe beams transmitted by the base station according to current channel state information fed back by the target terminal; inputting the first signal strength value into a pre-trained network prediction model to predict a second signal strength value of a plurality of narrow beams transmitted by the base station; selecting a candidate beam matching a target narrow beam from all narrow beams corresponding to the base station, the target narrow beam being the narrow beam with the highest second signal strength value; and selecting a target beam from the candidate beam. Optionally, the instructions can also be executed by the processor of the electronic device to perform other steps involved in the example embodiment.

[0183] Figure 7 An example diagram of an electronic device 300 is shown. Those skilled in the art can understand that the diagram is only an example of the electronic device 300 and does not limit the electronic device 300, which can include more or fewer components than shown, or combine some components, or include different components, for example, the electronic device 300 can also include an input / output device, a network access device, a bus, etc. Figure 7 The diagram is only an example of the electronic device 300 and does not limit the electronic device 300, which can include more or fewer components than shown, or combine some components, or include different components, for example, the electronic device 300 can also include an input / output device, a network access device, a bus, etc.

[0184] The processor 302 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, or the like. The general-purpose processor can be a microprocessor or the processor 302 can also be any conventional processor. The processor 302 is a control center of the electronic device 300, and is connected to various parts of the electronic device 300 through various interfaces and lines.

[0185] The memory 301 can be used to store computer readable instructions 303. The processor 302 implements various functions of the electronic device 300 by running or executing the computer readable instructions or modules stored in the memory 301, and calling data stored in the memory 301. The memory 301 can mainly include a program storage area and a data storage area. The program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), and the like. The data storage area can store data created according to the use of the electronic device 300, and the like. In addition, the memory 301 can include a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, a read-only memory (ROM), a random access memory (RAM) or other non-volatile / volatile storage devices.

[0186] The modules integrated in the electronic device 300, if implemented in the form of software function modules and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiments of the method can also be implemented through computer readable instructions instructing related hardware. The computer readable instructions can be stored in a computer readable storage medium, and when executed by a processor, can implement the steps of the above-mentioned various method embodiments.

[0187] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the application being indicated by the following claims.

[0188] It is to be understood that the application is not limited to the precise construction herein disclosed and shown in the drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the claims that follow.

Claims

1. A method for beam selection, characterized in that, Applied to base station equipment, including: Based on the current channel state information fed back by the target terminal, the first signal strength value of the multiple probe beams sent by the base station is predicted; The first signal strength value is input into the pre-trained network prediction model to predict the second signal strength value of the base station transmitting multiple narrow beams; Among all the narrow beams corresponding to the base station, a candidate beam that matches the target narrow beam is selected, wherein the target narrow beam is the narrow beam with the highest second signal strength value; Select the target beam from the candidate beams; The step of selecting a candidate beam that matches the target narrow beam from all narrow beams corresponding to the base station includes: Determine the transmission direction of the target narrow beam; The region within a preset angle range from the launch direction is selected as the candidate region; The narrow beam whose transmission direction is in the candidate area among all the narrow beams transmitted by the base station is selected as the candidate beam; Selecting the target beam from the candidate beams includes: Each alternative beam is sent to the target terminal, and the third signal strength value of each alternative beam is received from the target terminal. The candidate beam with the highest third signal strength value is selected as the target beam.

2. The method as described in claim 1, characterized in that, The step of predicting the first signal strength value of the multiple probe beams transmitted by the base station based on the current channel state information fed back by the target terminal includes: Based on the current channel state information and the preset probe beam codebook matrix, the signal strength value of each probe beam transmitted by the base station is predicted; The vector composed of the signal strength values ​​of each probe beam is used as the first signal strength value.

3. The method as described in claim 1, characterized in that, Before predicting the first signal strength value of the base station transmitting multiple probe beams based on the current channel state information fed back by the target terminal, the method further includes: The system sends training signals to the training terminal and receives training channel state information fed back by the training terminal based on the training signals; and predicts the first training signal strength value of the multiple training probe beams sent by the base station based on the training channel state information. as well as, Send multiple narrow beams to the training terminal and receive the second training signal strength values ​​corresponding to the multiple narrow beams fed back by the training terminal; The network prediction model is trained based on the first training signal strength value and the second training signal strength value.

4. The method as described in claim 3, characterized in that, The step of training the network prediction model based on the first training signal intensity value and the second training signal intensity value includes: The first training signal intensity value is normalized to obtain training input data; and the second training signal intensity value is normalized to obtain training output data. The initial network prediction model is iteratively trained using the training input data and the training output data until the network prediction model is obtained.

5. A beam selection device, characterized in that, Applied to base station equipment, including: The first prediction module is configured to predict the first signal strength value of the multiple probe beams transmitted by the base station based on the current channel state information fed back by the target terminal. The second prediction module is configured to input the first signal strength value into a pre-trained network prediction model to predict the second signal strength value of the base station transmitting multiple narrow beams. The matching module is configured to select a candidate beam that matches the target narrow beam from all narrow beams corresponding to the base station, wherein the target narrow beam is the narrow beam with the highest second signal strength value; The matching module is specifically configured as follows: Determine the transmission direction of the target narrow beam; The region within a preset angle range from the launch direction is selected as the candidate region; The narrow beam whose transmission direction is in the candidate area among all the narrow beams transmitted by the base station is selected as the candidate beam; The selection module is configured to select a target beam from the candidate beams; The selection module is specifically configured as follows: Each alternative beam is sent to the target terminal, and the third signal strength value of each alternative beam is received from the target terminal. The candidate beam with the highest third signal strength value is selected as the target beam.

6. An electronic device, characterized in that, include: Memory, used to store executable instructions; as well as, A processor, configured to execute the executable instructions with the memory to perform the operation of the beam selection method of any one of claims 1-4.

7. A computer-readable storage medium for storing computer-readable instructions, characterized in that, When the instruction is executed, it performs the operation of the beam selection method according to any one of claims 1-4.

Images