A beamforming method, terminal and storage medium

Abstract

Embodiments of the present application disclose a beamforming method, a terminal and a storage medium. The terminal acquires antenna input data, wherein the antenna input data comprises target antenna data, a target antenna model and initial antenna weights. Simulation processing is performed according to the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold value, weight optimization processing is performed according to the antenna input data to obtain updated antenna data. Until an updated function value corresponding to the updated antenna data is less than the preset threshold value, target weight data is determined according to the updated antenna data to perform beamforming.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a beamforming method, a terminal, and a storage medium. Background Technology

[0002] Beamforming is a signal processing technique that uses sensor arrays to transmit and receive signals in a directional manner. In existing technologies, heuristic optimization algorithms are often used to achieve beamforming. For example, an alternating minimization algorithm can be used to optimize the beamforming vector. During the optimization process, the number of channels needs to be set. However, the setting of the number of channels will affect the beamforming accuracy. At the same time, if the channels are closed, the final gain of the waveform will also decrease. In addition, there are some methods where the main lobe width of the waveform is only related to the number of antenna elements during the calculation process. For an antenna with a certain number of elements, only one beamforming can be obtained after determining the azimuth angle.

[0003] This shows that existing technologies often suffer from poor beamforming performance. Summary of the Invention

[0004] This application provides a beamforming method, a terminal, and a storage medium that can effectively improve the beamforming effect.

[0005] The technical solution of this application embodiment is implemented as follows:

[0006] In a first aspect, embodiments of this application provide a beamforming method, the method comprising:

[0007] Acquire antenna input data; wherein, the antenna input data includes target antenna data, target antenna model, and initial antenna weights;

[0008] The initial antenna weights are used for simulation processing to obtain the initial simulation data corresponding to the initial antenna weights.

[0009] Calculate the initial function values ​​corresponding to the initial simulation data;

[0010] If the initial function value is greater than or equal to the preset threshold, then weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Then, target weight data is determined based on the updated antenna data for beamforming.

[0011] Secondly, embodiments of this application provide a terminal, which includes an acquisition unit, a calculation unit, and a processing unit.

[0012] The acquisition unit is used to acquire antenna input data; wherein the antenna input data includes target antenna data, target antenna model and initial antenna weights; and to perform simulation processing based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights;

[0013] The calculation unit is used to calculate the initial function value corresponding to the initial simulation data;

[0014] The processing unit is configured to perform weight optimization processing based on the antenna input data to obtain updated antenna data if the initial function value is greater than or equal to a preset threshold value, until the updated function value corresponding to the updated antenna data is less than the preset threshold value, and then determine target weight data based on the updated antenna data for beamforming.

[0015] Thirdly, embodiments of this application provide a terminal, which further includes a processor and a memory storing executable instructions of the processor. When the instructions are executed by the processor, the beamforming method described above is implemented.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a program applied in a terminal, wherein when the program is executed by a processor, it implements the beamforming method described above.

[0017] This application provides a beamforming method, a terminal, and a storage medium. The terminal acquires antenna input data, which includes target antenna data, target antenna model, and initial antenna weights. Simulation processing is performed based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Finally, target weight data is determined based on the updated antenna data for beamforming. Therefore, in this application, antenna input data, including target antenna data, target antenna model, and initial antenna weights, can be obtained first. Then, simulation processing can be performed based on the antenna input data to obtain initial simulation data corresponding to the initial antenna weights, and the initial function value corresponding to the initial simulation data can be calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing can continue based on the antenna input data until the target weight data is obtained. Beamforming can then be performed based on the target weight data. In other words, by calculating and obtaining the target weight data in the above manner, this application can effectively improve the beamforming effect and is conducive to the fine optimization of large-scale antenna technology. At the same time, by considering the target antenna model in the antenna input data, this application can make the beamforming method more versatile. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 1 ;

[0019] Figure 2 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 1 ;

[0020] Figure 3 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 2 ;

[0021] Figure 4 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 2 ;

[0022] Figure 5 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 3 ;

[0023] Figure 6 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 4 ;

[0024] Figure 7 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 5 ;

[0025] Figure 8 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 6 ;

[0026] Figure 9 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 7 ;

[0027] Figure 10 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 8 ;

[0028] Figure 11 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 9 ;

[0029] Figure 12 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 10 ;

[0030] Figure 13 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 3 ;

[0031] Figure 14 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 4 ;

[0032] Figure 15 This is a schematic diagram of the terminal structure proposed in the embodiments of this application. Figure 1 ;

[0033] Figure 16 This is a schematic diagram of the terminal structure proposed in the embodiments of this application. Figure 2 . Detailed Implementation

[0034] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the relevant application and not for limiting the application. Furthermore, it should be noted that, for ease of description, only the parts related to the relevant application are shown in the accompanying drawings.

[0035] In Massive MIMO, each cell can be configured with up to eight beams across eight time slots. The maximum gain of these eight beams in different directions is called the synthesized beam. The design of these eight beams directly determines the beamforming capability of the final synthesized beam. The more diverse the beam shape and the closer its shape and gain are to the expected values, the easier it is to design the synthesized beam, and the better its coverage and signal strength. Currently, most common beamforming methods employ analytical calculations or heuristic optimization. Common heuristic optimization algorithms include particle swarm optimization, genetic algorithms, and differential evolution algorithms. For example, an alternating minimization algorithm can be used to optimize the beamforming vector. During optimization, the number of channels needs to be set; however, setting the number of channels affects beamforming accuracy, and if channels are closed, the final gain of the waveform will also decrease. In addition, some methods only consider the main lobe width of the waveform to the number of antenna elements during calculation. For an antenna with a certain number of elements, only one beamforming can be obtained after determining the azimuth angle. Furthermore, existing technologies typically do not consider antenna types and lack versatility across multiple antenna types.

[0036] To address the problems existing in current beamforming methods, this application provides a beamforming method, a terminal, and a storage medium. The terminal acquires antenna input data, which includes target antenna data, target antenna model, and initial antenna weights. Simulation processing is performed based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing is performed based on the antenna input data to obtain target weight data for beamforming.

[0037] This solution enables simultaneous optimization in both the horizontal and vertical planes, facilitating fine-grained optimization of Massive MIMO. Furthermore, it is compatible with various antenna models and supports the optimization generation of beams with diverse bandwidths and directions. The generated beams can be combined and synthesized in various scenarios to meet different requirements.

[0038] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0039] Example 1

[0040] This application provides a beamforming method. Figure 1 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 1 ,like Figure 1 As shown, the beamforming method may include the following steps:

[0041] Step 101: Obtain antenna input data; wherein, the antenna input data includes target antenna data, target antenna model and initial antenna weights.

[0042] In the embodiments of this application, the terminal may first acquire antenna input data; wherein, the antenna input data includes target antenna data, target antenna model and initial antenna weights.

[0043] It should be noted that, in the embodiments of this application, at the beginning stage of beamforming, it is necessary to first acquire antenna input data, which includes target antenna data, target antenna model, and initial antenna weights.

[0044] It is understood that, in the embodiments of this application, target antenna data refers to the relevant data of the target beam to be obtained. Target antenna data may include horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, gain, and sidelobe suppression value, etc.

[0045] Furthermore, in the embodiments of this application, target weights can be calculated for any type of antenna to achieve beamforming. Therefore, this application can predetermine the target antenna model to achieve beamforming for this type of antenna, overcoming the deficiency of the prior art in not having the universality of multiple antennas.

[0046] Furthermore, in the embodiments of this application, the initial antenna weights are the first generation of antenna weights randomly generated during the parameter initialization stage; that is, at the beginning stage of beamforming, the initial antenna weights are first randomly generated, and then the target antenna weights are obtained based on the initial antenna weights; at the same time, at the beginning stage of beamforming, the number of antenna channels, the range of amplitude and phase values ​​on each channel, the mode of antenna channel shutdown, etc. can also be determined.

[0047] Furthermore, in the embodiments of this application, the initial antenna weights may include information such as the amplitude and phase of each channel.

[0048] Step 102: Perform simulation processing based on the initial antenna weights to obtain the initial simulation data corresponding to the initial antenna weights.

[0049] In the embodiments of this application, after the terminal acquires the antenna input data, it can perform simulation processing based on the initial antenna weights to obtain the initial simulation data corresponding to the initial antenna weights.

[0050] It should be noted that, in the embodiments of this application, the simulation processing is used to obtain beam-related information corresponding to the antenna weights, i.e., simulation data; therefore, after performing simulation processing on the initial antenna weights, the initial simulation data corresponding to the initial antenna weights can be obtained.

[0051] Specifically, in the embodiments of this application, the method for obtaining initial simulation data corresponding to the initial antenna weights through simulation processing may include: calculating field strength information and transmission power based on the initial antenna weights; then determining the antenna pattern based on the field strength information; and finally obtaining the initial simulation data based on the antenna pattern and transmission power to achieve simulation processing. The method for obtaining the initial simulation data based on the antenna pattern and transmission power may include determining the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern; thereby obtaining the initial simulation data based on the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power.

[0052] Step 103: Calculate the initial function values ​​corresponding to the initial simulation data.

[0053] In the embodiments of this application, after the terminal performs simulation processing based on the initial antenna weights to obtain the initial simulation data corresponding to the initial antenna weights, it can calculate the initial function value corresponding to the initial simulation data.

[0054] It should be noted that, in the embodiments of this application, the function value is calculated based on the objective function model. The function value can be used to measure the distance between the sub-beam corresponding to the currently obtained antenna weight and the target sub-beam corresponding to the target antenna data.

[0055] Furthermore, in the embodiments of this application, the initial function value refers to the function value obtained by calculating the initial simulation data using the target function model; the initial function value can be used to measure the distance between the sub-beam generated corresponding to the initial antenna weight and the target sub-beam.

[0056] Specifically, in the embodiments of this application, the method for calculating the initial function value corresponding to the initial simulation data may include: performing distance calculation processing on the initial simulation data and target antenna data based on the objective function model to obtain the initial function value.

[0057] Furthermore, in the embodiments of this application, various objective function models can be designed according to requirements, and then any objective function model can be selected for calculation when using the objective function model.

[0058] Step 104: If the initial function value is greater than or equal to the preset threshold value, then perform weight optimization processing based on the antenna input data to obtain updated antenna data. Continue until the updated function value corresponding to the updated antenna data is less than the preset threshold value. Then, determine the target weight data based on the updated antenna data to perform beamforming.

[0059] In the embodiments of this application, after the terminal calculates the initial function value corresponding to the initial simulation data, if the initial function value is greater than or equal to a preset threshold value, it performs weight optimization processing based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold value. Then, the target weight data is determined based on the updated antenna data for beamforming.

[0060] It should be noted that in the embodiments of this application, the preset threshold value is selected based on the objective function model. That is, based on different objective function models, corresponding preset threshold values ​​can be set.

[0061] Furthermore, in the embodiments of this application, the preset threshold value can be used as a condition for determining whether to stop weight optimization; that is, if the function value corresponding to the currently obtained antenna weight value is greater than or equal to the preset threshold value, it can be determined that weight optimization needs to continue; while if the function value corresponding to the currently obtained antenna weight value is less than the preset threshold value, it can be determined that weight optimization does not need to continue, and the currently obtained antenna weight value can be used as the target weight value data, thereby performing beamforming based on the target weight value data to obtain the target beam.

[0062] It is understood that, in the embodiments of this application, the weight optimization process is used to calculate and update the weight data until the target weight data is obtained.

[0063] Specifically, in the embodiments of this application, the method for obtaining updated antenna data by performing weight optimization processing based on antenna input data may include: determining state data based on antenna input data; then using a first policy network to calculate and process the state data to obtain action data; and then determining updated antenna data based on the action data.

[0064] It is understood that, in the embodiments of this application, the target sub-beam can be obtained by performing beamforming based on the target weight data.

[0065] It should be noted that, in the embodiments of this application, in the process of weight optimization, in addition to judging according to the preset threshold value, it is also possible to determine whether to continue optimization or iteration according to the number of iteration steps. For example, a maximum step threshold value STEP_TH is preset. In the process of weight optimization, if the number of iteration steps is greater than STEP_TH, optimization or iteration can be stopped.

[0066] Furthermore, in embodiments of this application, the method for the terminal to obtain updated antenna data by performing weight optimization processing based on the antenna input data may include the following steps:

[0067] Step 201: Use a preset optimization model to perform weight optimization processing on the antenna input data to obtain updated antenna data; wherein, the preset optimization model includes a preset policy network and a preset evaluation network; the preset policy network includes a first policy network and a second policy network; the preset evaluation network includes a first evaluation network and a second evaluation network.

[0068] In the embodiments of this application, the terminal performs weight optimization processing on the antenna input data to obtain updated antenna data. Specifically, the terminal can use a preset optimization model to perform weight optimization processing on the antenna input data to obtain updated antenna data. The preset optimization model includes a preset policy network and a preset evaluation network. The preset policy network includes a first policy network and a second policy network. The preset evaluation network includes a first evaluation network and a second evaluation network.

[0069] It should be noted that, in the embodiments of this application, the preset optimization model is used to perform weight optimization processing; the preset optimization model includes a preset policy network and a preset evaluation network; the preset policy network includes a first policy network and a second policy network; the preset evaluation network includes a first evaluation network and a second evaluation network.

[0070] For example, in the embodiments of this application, the preset optimization model can be a deep reinforcement learning (Deep Deterministic Policy Gradient, DDPG) algorithm.

[0071] Furthermore, in the embodiments of this application, before performing weight optimization processing on the antenna input data using a preset optimization model, it is necessary to determine the relevant parameters of the preset optimization model; for example, if the preset optimization model is the DDPG algorithm, then before performing weight optimization processing using the DDPG algorithm, it is necessary to determine its hyperparameters, including parameters such as cache size, discount factor, batch size, learning rate, and network weight update factor.

[0072] Furthermore, in the embodiments of this application, the preset strategy network includes a first strategy network and a second strategy network, and the first strategy network and the second strategy network have the same structure; wherein, the preset strategy network can be set to different structures according to requirements.

[0073] For example, in the embodiments of this application, the preset policy network can be divided into two structures; Figure 2 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 1 ,like Figure 2 The diagram shows one structure of a preset policy network, which can be called the first structure. Based on this first structure, the input to the preset policy network is state data, and its output is a triple consisting of an antenna channel number, an amplitude, and a phase. This network structure includes a shared network layer and a distributed network, and may also include softmax and tanh functions to adjust the weights and phase of a channel. Another preset policy network structure... Figure 3 The structure shown has fewer branch networks for output antenna channel numbers, but more outputs are added to the branch networks for output amplitude and phase, which can be called the second structure. The output of the preset strategy network based on the second structure can be represented as a vector: (amplitude, phase) × number of channels, which can be used to adjust the amplitude and phase of all channels.

[0074] Furthermore, the method by which the terminal uses a preset optimization model to perform weight optimization processing on the antenna input data to obtain updated antenna data, i.e., the method proposed in step 201, may include the following steps:

[0075] Step 201a: Determine the status data based on the antenna input data.

[0076] In the embodiments of this application, the terminal uses a preset optimization model to perform weight optimization processing on the antenna input data to obtain updated antenna data. Specifically, the terminal can first determine the state data based on the antenna input data.

[0077] It should be noted that, in the embodiments of this application, the state data can be used as the input of the preset policy network; therefore, when using the preset optimization model to perform weight optimization processing on the antenna input data, it is first necessary to determine the state data based on the antenna input data, so that the state data can be input into the preset policy network to achieve weight optimization processing.

[0078] For example, in the embodiments of this application, state data is constituted based on the target antenna data, target antenna model, initial antenna weight, and initial simulation data corresponding to the initial antenna weight in the antenna input data.

[0079] Step 201b: Use the first policy network to process the state data to obtain action data.

[0080] In the embodiments of this application, after the terminal determines the state data based on the antenna input data, it can use the first policy network to calculate and process the state data to obtain action data.

[0081] It should be noted that, in the embodiments of this application, using the first policy network to calculate and process the state data is equivalent to inputting the state data into the first policy network, thereby the first policy network outputting the corresponding action data.

[0082] Step 201c: Determine the updated antenna data based on the action data.

[0083] In the embodiments of this application, the terminal uses the first policy network to calculate and process the state data, and after obtaining the action data, it can determine the updated antenna data based on the action data.

[0084] It should be noted that, in the embodiments of this application, the state data can be updated based on the action data to obtain the updated state data, which is the updated antenna data.

[0085] Furthermore, in the embodiments of this application, after updating the state data based on the action data obtained from the first policy network with different structures, the updated state data will also be different.

[0086] For example, in an embodiment of this application, based on the above example, if the structure of the preset policy network is a first structure, the state data is updated based on the action data obtained by the first policy network under the first structure, and the updated state data can be expressed as the following formula:

[0087]

[0088] Where next_state iThis refers to the updated state data, specifically the updated state data for the channel corresponding to the index value; i is the index value; state i This refers to the state data of the channel corresponding to the index value; A and P are the amplitude and phase calculated by the first policy network, i.e., the action data; and index is the index of the channel.

[0089] For example, in an embodiment of this application, based on the above example, if the structure of the preset policy network is a second structure, the state data is updated based on the action data obtained by the first policy network under the second structure, and the updated state data can be expressed as the following formula:

[0090] next_state = state + action (2)

[0091] Where next_state is the updated state data for all channels; state is the state data; and action is the action data.

[0092] Furthermore, after the terminal determines the updated antenna data based on the action data, i.e., after step 201c, it may also include the following steps:

[0093] Step 201d: Perform simulation processing based on the updated antenna data and state data to obtain simulation results; wherein, the simulation results include the first simulation data corresponding to the updated antenna data and the second simulation data corresponding to the state data.

[0094] In the embodiments of this application, after the terminal determines the updated antenna data based on the action data, it can perform simulation processing based on the updated antenna data and the state data to obtain simulation results; wherein, the simulation results include first simulation data corresponding to the updated antenna data and second simulation data corresponding to the state data.

[0095] It should be noted that, in the embodiments of this application, during the weight optimization process, simulation processing is required based on the state data obtained each time to obtain the corresponding simulation results. Only then can the corresponding function value be calculated based on the simulation results to determine whether the weight optimization process can be terminated.

[0096] Furthermore, in the embodiments of this application, the simulation results include first simulation data corresponding to the updated antenna data and second simulation data corresponding to the state data.

[0097] Step 201e: Calculate the first function value corresponding to the simulation result.

[0098] In the embodiments of this application, after the terminal performs simulation processing based on updated antenna data and status data and obtains the simulation results, it can calculate the first function value corresponding to the simulation results.

[0099] It should be noted that, in the embodiments of this application, since the simulation results include first simulation data corresponding to the updated antenna data and second simulation data corresponding to the state data, when calculating the first function value corresponding to the simulation results, it is necessary to calculate based on the first simulation data and the second simulation data to obtain the first function value.

[0100] Specifically, in the embodiments of this application, a target function model is used to perform distance calculation processing on the first simulation data and the target antenna data to obtain a second function value; at the same time, a target function model is used to perform distance calculation processing on the second simulation data and the target antenna data to obtain a third function value; then the minimum function value among the second function value and the third function value is determined as the first function value.

[0101] Step 201f: If the first function value is greater than or equal to the preset threshold value, then perform subsequent processing based on the state data and the updated antenna data to obtain subsequent state data, until the subsequent function value corresponding to the subsequent state data is less than the preset threshold value, and then determine the target weight data based on the subsequent state data.

[0102] In the embodiments of this application, after the terminal calculates the first function value corresponding to the simulation result, if the first function value is greater than or equal to a preset threshold value, it performs subsequent processing based on the state data and the updated antenna data to obtain subsequent state data. The target weight data is determined based on the subsequent state data when the subsequent function value corresponding to the subsequent state data is less than the preset threshold value.

[0103] It is understood that in the embodiments of this application, if the first function value is greater than or equal to the preset threshold value, it means that the antenna weight data obtained in this round of optimization cannot be used as the target weight data. Therefore, it is necessary to continue to process the state data and the updated antenna data to obtain the subsequent state data until the subsequent function value corresponding to the subsequent state data is less than the preset threshold value. Then, the target weight data can be determined based on the subsequent state data.

[0104] Furthermore, in the embodiments of this application, the subsequent state data is the updated state data obtained by further processing based on the updated antenna data.

[0105] Figure 3 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 2 ,like Figure 3 As shown, the terminal performs subsequent processing based on the status data and updated antenna data. The method for obtaining subsequent status data may include the following steps:

[0106] Step 301: Use the first evaluation network to evaluate and process the state data and action data to obtain evaluation values.

[0107] In the embodiments of this application, the terminal performs subsequent processing based on the state data and updated antenna data to obtain subsequent state data. Specifically, the terminal can first use the first evaluation network to evaluate the state data and action data to obtain an evaluation value.

[0108] It should be noted that, in the embodiments of this application, the first evaluation network and the second evaluation network can be used to perform evaluation processing and obtain evaluation values.

[0109] For example, in the embodiments of this application, Figure 4 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 2 ,like Figure 4 The diagram shows the structure of the first evaluation network, which may include a concatenate function and a fully connected network. After inputting the state data and action data into the first evaluation network, the evaluation value q can be obtained.

[0110] Step 302: Use the second strategy network to process the updated antenna data to obtain the updated action data.

[0111] In the embodiments of this application, after the terminal uses the first evaluation network to evaluate and process the state data and action data and obtain the evaluation value, it can use the second policy network to calculate and process the updated antenna data to obtain the updated action data.

[0112] It is understood that, in the embodiments of this application, the process of using the second policy network to calculate and process the updated antenna data to obtain the updated action data refers to inputting the updated antenna data into the second policy network so that the updated action data can be output.

[0113] For example, in an embodiment of this application, the updated action data is represented as next_state. The next_state is input into the second policy network to obtain the updated action data next_action.

[0114] Step 303: Using the second evaluation network, evaluate the updated antenna data and the updated motion data to obtain the updated evaluation value.

[0115] In the embodiments of this application, after the terminal uses the second strategy network to calculate and process the updated antenna data to obtain the updated action data, it can use the second evaluation network to evaluate the updated antenna data and the updated action data to obtain the updated evaluation value.

[0116] It is understood that, in the embodiments of this application, the process of using the second evaluation network to evaluate the updated antenna data and the updated motion data to obtain the updated evaluation value refers to inputting the updated antenna data and the updated motion data into the second evaluation network, thereby obtaining the corresponding updated evaluation value.

[0117] For example, in an embodiment of this application, next_state and next_action are input into a second evaluation network to obtain the corresponding updated evaluation value next_q.

[0118] Step 304: Calculate the loss function value based on the evaluation value, the updated evaluation value, and the reward value.

[0119] In the embodiments of this application, after the terminal uses the second evaluation network to evaluate the updated antenna data and the updated action data and obtains the updated evaluation value, it can calculate the loss function value based on the evaluation value, the updated evaluation value and the reward value.

[0120] It should be noted that, in the embodiments of this application, the loss function value is calculated based on the evaluation value, the updated evaluation value, and the reward value.

[0121] For example, in an embodiment of this application, the loss function value can be calculated using the following formula:

[0122] critic_loss=MSE(q,q_target) (3)

[0123] Where `critic_loss` is the loss function value, `q` is the evaluated value, and `q_target` is the ideal evaluated value; it can be seen that the loss function value can be the mean squared error between the evaluated value and the ideal evaluated value; the ideal evaluated value can be calculated using the following formula:

[0124] q_target=r+γ×next_q (4)

[0125] Where r is the reward value, γ is the initial discount factor, and next_q is the updated evaluation value.

[0126] Step 305: Update the preset policy network and preset evaluation network according to the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network.

[0127] In the embodiments of this application, after the terminal calculates the loss function value based on the evaluation value, the updated evaluation value, and the reward value, it can update the preset policy network and the preset evaluation network based on the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network.

[0128] It should be noted that, in the embodiments of this application, the preset policy network and the preset evaluation network can be updated to obtain the updated policy network and the updated evaluation network, thereby enabling the acquisition of subsequent state data using the updated policy network and the updated evaluation network.

[0129] Specifically, in the embodiments of this application, the method of updating a preset policy network and a preset evaluation network based on loss function values ​​and evaluation values ​​to obtain updated policy networks and updated evaluation networks may include: updating a first evaluation network based on loss function values ​​to obtain an updated first evaluation network; simultaneously updating a first policy network based on evaluation values ​​to obtain an updated first policy network; and then updating a second policy network and a second evaluation network based on the updated first policy network and the updated first evaluation network to obtain updated second policy networks and updated second evaluation networks.

[0130] Step 306: Obtain subsequent state data based on the updated policy network, the updated evaluation network, and the updated antenna data.

[0131] In the embodiments of this application, after the terminal updates the preset policy network and the preset evaluation network according to the loss function value and the evaluation value, and obtains the updated policy network and the updated evaluation network, it can obtain subsequent state data based on the updated policy network, the updated evaluation network and the updated antenna data.

[0132] It is understood that, in the embodiments of this application, the process of obtaining subsequent state data based on the updated policy network, the updated evaluation network, and the updated antenna data is to repeat the aforementioned calculation, simulation, and evaluation processes using the updated first policy network and the updated second policy network, as well as the updated first evaluation network and the updated second evaluation network.

[0133] Figure 5 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 3 ,like Figure 5 As shown, before the terminal calculates the loss function value based on the evaluation value, the updated evaluation value, and the reward value, i.e., before step 304, the following steps may also be included:

[0134] Step 307: Calculate the reward value based on the simulation results.

[0135] In the embodiments of this application, before the terminal calculates the loss function value based on the evaluation value, the updated evaluation value, and the reward value, it can also calculate the reward value based on the simulation results.

[0136] It should be noted that, in the embodiments of this application, the reward value can be calculated based on the simulation results. Since the simulation results include the first simulation data corresponding to the updated antenna data and the second simulation data corresponding to the state data, the reward value also needs to be calculated based on the first simulation data and the second simulation data.

[0137] Specifically, in the embodiments of this application, the method for calculating the reward value based on the simulation results may include: calculating a first result value corresponding to the first simulation data based on the first simulation data and the target antenna data in the simulation results; simultaneously calculating a second result value corresponding to the second simulation data based on the second simulation data and the target antenna data in the simulation results; and then subtracting the first result value and the second result value to obtain the reward value.

[0138] Figure 6 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 4 ,like Figure 6 As shown, the method by which the terminal calculates the reward value based on the simulation results, i.e., the method proposed in step 307, may include the following steps:

[0139] Step 307a: Calculate the first result value corresponding to the first simulation data based on the first simulation data and the target antenna data in the simulation results.

[0140] In the embodiments of this application, the terminal calculates the reward value based on the simulation results. Specifically, the terminal can calculate the first result value corresponding to the first simulation data based on the first simulation data and the target antenna data in the simulation results.

[0141] It should be noted that, in the embodiments of this application, the distance calculation is performed on the first simulation data and the target antenna data based on the objective function model to obtain the first result value corresponding to the first simulation data.

[0142] For example, in an embodiment of this application, the first simulation data is represented as a2. Based on the objective function model, a2 and the target antenna data are processed to calculate the distance and obtain the first result value d2.

[0143] Step 307b: Calculate the second result value corresponding to the second simulation data based on the second simulation data and the target antenna data in the simulation results.

[0144] In the embodiments of this application, the terminal calculates the reward value based on the simulation results. Specifically, the terminal can calculate the second result value corresponding to the second simulation data based on the second simulation data and the target antenna data in the simulation results.

[0145] It should be noted that, in the embodiments of this application, similarly, the distance calculation process is performed on the second simulation data and the target antenna data based on the objective function model to obtain the second result value corresponding to the second simulation data.

[0146] For example, in an embodiment of this application, the second simulation data is represented as a1. Based on the objective function model, a1 and the target antenna data are processed to calculate the distance and obtain the second result value d1.

[0147] Step 307c: Subtract the first result value from the second result value to obtain the reward value.

[0148] In the embodiments of this application, after the terminal calculates the first result value corresponding to the first simulation data based on the first simulation data and the target antenna data in the simulation results, and calculates the second result value corresponding to the second simulation data based on the second simulation data and the target antenna data in the simulation results, it can subtract the first result value and the second result value to obtain the reward value.

[0149] For example, the reward value can be calculated using the following formula:

[0150] Reward = d2 - d1 (5)

[0151] Here, Reward is the reward value, d2 is the first result value, and d1 is the second result value.

[0152] In summary, in the embodiments of this application, a weight optimization device or module can be set up to obtain the target antenna weight. First, a judgment can be made based on the randomly generated initial antenna weight, and then a weight optimization process can be performed based on the judgment result until the target weight data is obtained.

[0153] For example, in the embodiments of this application, Figure 7 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 5 ,like Figure 7As shown, the overall process described above can include: First, determining the target antenna data, which includes the target beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and gain. Then, parameter initialization is performed, and initial antenna weights are randomly generated. Next, based on the initial antenna weights, it is determined whether the termination condition is met. Specifically, based on the objective function model, the initial simulation results of the initial antenna weights and the target antenna data are used to calculate the distance, obtaining an initial function value. It is then determined whether the initial function value is less than a preset threshold. Further, if the initial antenna weights are determined to meet the termination condition, the relevant data can be saved. The relevant data includes the initial antenna weights and initial simulation results, which can be directly used to determine the target antenna weights. However, if the initial antenna weights do not meet the termination condition, weight optimization is performed to continue calculating the antenna weights and obtain updated antenna data. After obtaining the updated antenna data, simulation processing can be performed based on the updated antenna data, and the corresponding simulation results for the sub-beams can be collected. The simulation results are then input into the objective function model to obtain the corresponding function values. Based on the function values, it can be determined again whether the updated antenna data meets the termination condition, until the target weight data is obtained.

[0154] This application provides a beamforming method in which a terminal acquires antenna input data. The antenna input data includes target antenna data, target antenna model, and initial antenna weights. Simulation processing is performed based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Finally, target weight data is determined based on the updated antenna data for beamforming. Therefore, in this application, antenna input data, including target antenna data, target antenna model, and initial antenna weights, can be obtained first. Then, simulation processing can be performed based on the antenna input data to obtain initial simulation data corresponding to the initial antenna weights, and the initial function value corresponding to the initial simulation data can be calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing can continue based on the antenna input data until the target weight data is obtained. Beamforming can then be performed based on the target weight data. In other words, by calculating and obtaining the target weight data in the above manner, this application can effectively improve the beamforming effect and is conducive to the fine optimization of large-scale antenna technology. At the same time, by considering the target antenna model in the antenna input data, this application can make the beamforming method more versatile.

[0155] Example 2

[0156] Based on the above embodiments, in another embodiment of this application... Figure 8 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 6 ,like Figure 8 As shown, the method for calculating the initial function values ​​corresponding to the initial simulation data at the terminal, i.e., the method proposed in step 103, may include the following steps:

[0157] Step 103a: Based on the objective function model, perform distance calculation processing on the initial simulation data and the target antenna data to obtain the initial function value.

[0158] In the embodiments of this application, the terminal calculates the initial function value corresponding to the initial simulation data. Specifically, the terminal can perform distance calculation processing on the initial simulation data and the target antenna data based on the target function model to obtain the initial function value.

[0159] It should be noted that, in the embodiments of this application, the objective function model can be used to evaluate the distance between the sub-beam generated by the currently obtained weight data and the sub-beam generated by the target antenna data; the closer the wavelength width and pointing direction of the sub-beam generated by the currently obtained weight data are to the wavelength width and pointing direction of the sub-beam generated by the target antenna data, the better; the greater the gain, the better; and the greater the sidelobe suppression, the better. Based on this principle, various different objective function models can be set; for example, the weighted linear form of the objective function model can be expressed as: (ω1*|h sim -h design |)+ω2*|v sim -v design |+ω3*|azimuth sim -azimuth design |+ω4*|zenith sim -zenith design |+ω5*powerloss

[0160] For example, the objective function model in the form of a weighted sum of squares can be expressed as: (ω1*|h sim -h design | 2 )+ω2*|v sim -v design | 2 +ω3*|azimuth sim -azimuth design | 2 +ω4*|zenith sim -zenith design | 2 +ω5*powerloss 2

[0161] For example, the objective function model in the weighted exponential form can be expressed as:

[0162] In the above different forms of objective function models, ω1, ω2, ω3, ω4, and ω5 are coefficients that can be selected according to specific needs; items with the subscript "sim" represent data from the simulation results obtained after processing the weight data, and items with the subscript "design" represent target antenna data; h represents horizontal beamwidth, v represents vertical beamwidth, azimuth represents the azimuth angle, zenith represents the downtilt angle, and powerloss represents power loss; for example, h sim h represents the horizontal wavewidth in the simulation results. design This indicates the horizontal beamwidth in the target antenna data.

[0163] Figure 9 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 7 ,like Figure 9 As shown, after the terminal calculates the initial function value corresponding to the initial simulation data, i.e., after step 103, the following steps may also be included:

[0164] Step 105: If the initial function value is less than the preset threshold, the target weight data is determined based on the initial antenna weights in order to perform beamforming.

[0165] In the embodiments of this application, after the terminal calculates the initial function value corresponding to the initial simulation data, if the initial function value is less than a preset threshold value, the target weight data is determined based on the initial antenna weights in order to perform beamforming.

[0166] It is understood that in the embodiments of this application, if the initial function value is less than the preset threshold value, it means that the initial function value meets the termination condition and there is no need to perform weight optimization processing. Therefore, the target weight data can be determined based on the initial antenna weights for beamforming.

[0167] Furthermore, in the embodiments of this application, the method by which the terminal updates the preset policy network and the preset evaluation network based on the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network, i.e., the method proposed in step 305, may include the following steps:

[0168] Step 305a: Update the first evaluation network according to the loss function value to obtain the updated first evaluation network.

[0169] In the embodiments of this application, the terminal updates the preset policy network and the preset evaluation network according to the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network. Specifically, the terminal can update the first evaluation network according to the loss function value to obtain the updated first evaluation network.

[0170] It should be noted that, in the embodiments of this application, backpropagation can be used to update the parameters in the first evaluation network based on the loss function value, so as to realize the update processing of the first evaluation network.

[0171] Step 305b: Update the first policy network based on the evaluation value to obtain the updated first policy network.

[0172] In the embodiments of this application, the terminal updates the preset policy network and the preset evaluation network according to the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network. Specifically, the terminal can update the first policy network according to the evaluation value to obtain the updated first policy network.

[0173] It is understood that, in the embodiments of this application, the parameters in the first policy network can be updated using the evaluation value to achieve the update processing of the first policy network, thereby obtaining the updated first policy network.

[0174] Step 305c: Update the second policy network and the second evaluation network according to the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network.

[0175] In the embodiments of this application, the terminal updates the first evaluation network according to the loss function value to obtain the updated first evaluation network, updates the first policy network according to the evaluation value, and obtains the updated first policy network. Then, the terminal updates the second policy network and the second evaluation network according to the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network.

[0176] For example, in the embodiments of this application, a soft update method can be used at certain iteration intervals to update the updated first policy network and the updated first evaluation network into the second policy network and the second evaluation network, respectively, so as to realize the update processing of the second policy network and the second evaluation network and obtain the updated second policy network and the updated second evaluation network; soft update helps to improve the stability of training.

[0177] For example, in an embodiment of this application, the soft update method can be represented by the following formula:

[0178] w infer =w infer *ρ+w real_time *(1-ρ) (6)

[0179] Among them, w infer Let w represent the second policy network and the second evaluation network. real_time Let ρ represent the first policy network and the first evaluation network; the value of ρ ranges from (0, 1).

[0180] Furthermore, in embodiments of this application, the terminal updates the second policy network and the second evaluation network based on the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network. That is, the method proposed in step 305c may include the following steps:

[0181] Step 305c1: Determine the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network.

[0182] In the embodiments of this application, the terminal updates the second policy network and the second evaluation network according to the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network. Specifically, the terminal can first determine the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network.

[0183] It is understood that, in the embodiments of this application, when updating the second policy network and the second evaluation network, the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network can be determined first.

[0184] Step 305c2: Update the second policy network according to the updated policy parameters to obtain the updated second policy network.

[0185] In the embodiments of this application, after determining the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network, the terminal can update the second policy network according to the updated policy parameters to obtain the updated second policy network.

[0186] It is understood that, in the embodiments of this application, the updated policy parameters are used to update the second policy network, thereby obtaining the updated second policy network.

[0187] Step 305c3: Update the second evaluation network according to the updated evaluation parameters to obtain the updated second evaluation network.

[0188] In the embodiments of this application, after determining the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network, the terminal can update the second evaluation network according to the updated evaluation parameters to obtain the updated second evaluation network.

[0189] It is understood that, in the embodiments of this application, the updated evaluation parameters are used to update the second evaluation network, thereby obtaining the updated second evaluation network.

[0190] Figure 10 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 8 ,like Figure 10 As shown, the method for obtaining the initial simulation data corresponding to the initial antenna weights by performing simulation processing on the terminal based on the initial antenna weights, i.e., the method proposed in step 102, may include the following steps:

[0191] Step 102a: Calculate the field strength information and transmission power based on the initial antenna weights.

[0192] In the embodiments of this application, the terminal performs simulation processing based on the initial antenna weights to obtain the initial simulation data corresponding to the initial antenna weights. Specifically, the terminal can first calculate the field strength information and transmission power based on the initial antenna weights.

[0193] It should be noted that, in the embodiments of this application, a simulation module or simulation device may be provided to implement simulation processing.

[0194] Furthermore, in the embodiments of this application, before performing simulation processing, it is also necessary to initialize the simulation parameters, including using a simulation module or simulation device to receive the antenna element radiation pattern and basic antenna information, wherein the basic antenna information may include the target antenna model, the spacing between antenna elements, and the number of channels, etc.

[0195] Furthermore, in the embodiments of this application, when calculating the field strength information and transmission power based on the initial antenna weights, the initial antenna weights can be represented as a weight matrix. For example, the weight matrix is ​​represented as w(m, n), where m and n represent the number of elements in the horizontal and vertical directions, respectively, and one w(m, n) contains an amplitude and a phase.

[0196] It should be noted that, in the embodiments of this application, the method for calculating the field strength information can be expressed as the following formula:

[0197]

[0198] in, Here, w(m,n) represents the field strength information, and w(m,n) represents the antenna weights. For displacement vectors, Antenna element radiation pattern; displacement vector The calculation method can be expressed as the following formula:

[0199]

[0200] Where m and n are the channel numbers, For different angles, d h For the horizontal spacing of the array, d v The vertical spacing of the array.

[0201] Furthermore, in the embodiments of this application, the method for calculating the transmission power based on the initial antenna weights may include: determining the amplitude information corresponding to the channel based on the initial antenna weights; and then calculating the transmission power based on the amplitude information corresponding to the channel.

[0202] Step 102b: Determine the antenna pattern based on the field strength information.

[0203] In the embodiments of this application, after the terminal calculates the field strength information and transmission power based on the initial antenna weights, it can determine the antenna pattern based on the field strength information.

[0204] It should be noted that, in the embodiments of this application, the method for determining the antenna pattern can be expressed as the following formula:

[0205]

[0206] in, This refers to the antenna radiation pattern.

[0207] Step 102c: Based on the antenna pattern and transmission power, obtain initial simulation data to perform simulation processing.

[0208] In the embodiments of this application, after the terminal determines the antenna pattern based on the field strength information, it can obtain initial simulation data based on the antenna pattern and transmission power to achieve simulation processing.

[0209] Specifically, in the embodiments of this application, the method for obtaining initial simulation data based on the antenna pattern and transmission power may include: determining the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value according to the antenna pattern; and then obtaining the initial simulation data based on the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power.

[0210] Figure 11 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 9 ,like Figure 11 As shown, the method for the terminal to obtain initial simulation data based on the antenna pattern and transmission power may include the following steps:

[0211] Step 401: Determine the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern.

[0212] In the embodiments of this application, the terminal obtains initial simulation data based on the antenna pattern and transmission power. Specifically, the terminal can determine the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern.

[0213] Specifically, in the embodiments of this application, the method for determining the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern may include: determining the maximum beam gain and sidelobe suppression value based on the antenna pattern; then determining the corresponding maximum azimuth angle and maximum downtilt angle based on the maximum beam gain; and finally determining the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and maximum downtilt angle.

[0214] Step 402: Obtain initial simulation data based on the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power.

[0215] In the embodiments of this application, after the terminal determines the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern, it can obtain initial simulation data based on the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power.

[0216] It is understood that, in the embodiments of this application, any simulation data obtained based on antenna weights may include the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power; the initial simulation data obtained based on the initial antenna weights also includes the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power.

[0217] Furthermore, in the embodiments of this application, the method by which the terminal determines the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern, i.e., the method proposed in step 401, may include the following steps:

[0218] Step 401a: Determine the maximum beam gain and sidelobe suppression value based on the antenna pattern.

[0219] In the embodiments of this application, the terminal determines the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern. Specifically, the terminal can first determine the maximum beam gain and sidelobe suppression value based on the antenna pattern.

[0220] It should be noted that, in the embodiments of this application, the maximum beam gain is obtained by looking up the antenna pattern.

[0221] Furthermore, in the embodiments of this application, the method for calculating the sidelobe suppression value can be expressed as the following formula:

[0222]

[0223] Where G is the sidelobe suppression value, and max(gain(φ,θ)) is the global maximum value in the antenna pattern; the calculation method of gain_sub can be expressed as the following formula:

[0224]

[0225] It can be seen that gain_sub is in The maximum value in the antenna pattern within the range of |θ-θf|>20.

[0226] Step 401b: Determine the corresponding maximum azimuth angle and maximum downtilt angle based on the maximum beam gain.

[0227] In the embodiments of this application, after the terminal determines the maximum beam gain and sidelobe suppression value based on the antenna pattern, it can determine the corresponding maximum azimuth angle and maximum downtilt angle based on the maximum beam gain.

[0228] It should be noted that, in the embodiments of this application, after the maximum beam gain is determined by searching in the antenna pattern, the corresponding azimuth angle and downtilt angle can be recorded based on the maximum beam gain. The azimuth angle and downtilt angle are the maximum azimuth angle and the maximum downtilt angle.

[0229] For example, in an embodiment of this application, in The maximum beam gain is located and determined, and the maximum azimuth angle is determined based on the maximum beam gain. and maximum downtilt angle θ max .

[0230] Step 401c: Determine the horizontal wavewidth, vertical wavewidth, direction angle, and downtilt angle based on the maximum heading angle and maximum downtilt angle.

[0231] In the embodiments of this application, after the terminal determines the corresponding maximum azimuth angle and maximum downtilt angle based on the maximum beam gain, it can determine the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and maximum downtilt angle.

[0232] Specifically, in the embodiments of this application, the method for determining the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and the maximum downtilt angle may include: calculating the vertical beamwidth and downtilt angle based on the maximum azimuth angle; and calculating the horizontal beamwidth and azimuth angle based on the maximum downtilt angle.

[0233] Furthermore, in the embodiments of this application, the method by which the terminal determines the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and the maximum downtilt angle, i.e., the method proposed in step 401c, may include the following steps:

[0234] Step 401c1: Calculate the vertical beamwidth and downtilt angle based on the maximum azimuth angle.

[0235] In the embodiments of this application, the terminal determines the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and the maximum downtilt angle. Specifically, the terminal can calculate the vertical beamwidth and downtilt angle based on the maximum azimuth angle.

[0236] Specifically, in the embodiments of this application, the method for calculating the vertical beamwidth and downtilt angle based on the maximum azimuth angle may include: obtaining the gain value corresponding to the downtilt angle on the plane corresponding to the maximum azimuth angle; determining a first gain value among the gain values ​​corresponding to the downtilt angle; wherein the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle; performing normalization processing based on the first gain value to obtain a first result; performing search processing on the first result according to a preset standard to determine a first value and a second value; obtaining the first absolute value of the subtraction result of the first value and the second value, and determining the first absolute value as the vertical beamwidth; obtaining the first mean of the first value and the second value, and determining the downtilt angle based on the first mean.

[0237] Step 401c2: Calculate the horizontal wavewidth and direction angle based on the maximum downtilt angle.

[0238] In the embodiments of this application, the terminal determines the horizontal beamwidth, vertical beamwidth, azimuth angle, and downtilt angle based on the maximum azimuth angle and the maximum downtilt angle. Specifically, the terminal can calculate the horizontal beamwidth and azimuth angle based on the maximum downtilt angle.

[0239] Specifically, in the embodiments of this application, the method for calculating the horizontal beamwidth and azimuth angle based on the maximum downtilt angle may include: obtaining the gain value corresponding to the azimuth angle on the plane corresponding to the maximum downtilt angle; determining a second gain value among the gain values ​​corresponding to the azimuth angle; wherein the second gain value is the largest gain value among the gain values ​​corresponding to the azimuth angle; performing normalization processing based on the second gain value to obtain a second result; performing search processing on the second result according to a preset standard to determine a third value and a fourth value; obtaining the second absolute value of the subtraction result of the third value and the fourth value, and determining the second absolute value as the horizontal beamwidth; obtaining the second mean of the third value and the fourth value, and determining the azimuth angle based on the second mean.

[0240] Furthermore, in the embodiments of this application, the method by which the terminal calculates the vertical beamwidth and downtilt angle based on the maximum azimuth angle, i.e., the method proposed in step 401c1, may include the following steps:

[0241] Step 501: Obtain the gain value corresponding to the downtilt angle on the plane corresponding to the maximum direction angle.

[0242] In the embodiments of this application, the terminal calculates the vertical beamwidth and downtilt angle based on the maximum azimuth angle. Specifically, the terminal can first obtain the gain value corresponding to the downtilt angle on the plane corresponding to the maximum azimuth angle.

[0243] For example, in an embodiment of this application, the maximum orientation angle is Pick For a plane, determine the gain value f(θ) corresponding to each point on the plane with a downtilt angle θ = (0, 180).

[0244] Step 502: Determine the first gain value among the gain values ​​corresponding to the downtilt angle; wherein, the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle.

[0245] In the embodiments of this application, after obtaining the gain value corresponding to the downtilt angle on the plane corresponding to the maximum azimuth angle, the terminal can determine the first gain value among the gain values ​​corresponding to the downtilt angle; wherein, the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle.

[0246] It should be noted that, in the embodiments of this application, the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle.

[0247] For example, in an embodiment of this application, a lookup process is performed on the gain value f(θ) corresponding to the downtilt angle to determine that the maximum value is f(θ). max , f(θ) max As the first gain value.

[0248] Step 503: Normalize the first gain value to obtain the first result.

[0249] In the embodiments of this application, after determining the first gain value among the gain values ​​corresponding to the downtilt angle, the terminal can perform normalization processing based on the first gain value to obtain the first result.

[0250] For example, in an embodiment of this application, the method for normalizing based on the first gain value can be expressed as the following formula:

[0251] F = f(θ) - f(θ) max (12)

[0252] Here, F is the first result.

[0253] Step 504: Perform search processing on the first result according to the preset standard to determine the first value and the second value.

[0254] In the embodiments of this application, after the terminal performs normalization processing based on the first gain value to obtain the first result, it can perform search processing on the first result according to a preset standard to determine the first value and the second value.

[0255] For example, in the embodiments of this application, the preset standard is to satisfy F ≤ -3; from f(θ) max For the corresponding angle, look to both sides for the first result that satisfies F≤-3, and determine it as the first value θ1 and the second value θ2.

[0256] Step 505: Obtain the first absolute value of the subtraction result of the first value and the second value, and determine the first absolute value as the vertical wavewidth.

[0257] In the embodiments of this application, after the terminal performs a search process on the first result according to a preset standard to determine the first value and the second value, it can obtain the first absolute value of the subtraction result of the first value and the second value, and determine the first absolute value as the vertical wavewidth.

[0258] For example, in an embodiment of this application, the first value θ1 and the second value θ2 are subtracted and their absolute values ​​are taken. The first absolute value can be expressed as |θ2-θ1|, and the result of |θ2-θ1| can be determined as the horizontal wavewidth.

[0259] Step 506: Obtain the first mean of the first and second values, and determine the downslope angle based on the first mean.

[0260] In the embodiments of this application, after the terminal performs search processing on the first result according to the preset standard and determines the first value and the second value, it can obtain the first average of the first value and the second value, and determine the downtilt angle based on the first average.

[0261] For example, in an embodiment of this application, the method for calculating the downtilt angle can be expressed as the following formula:

[0262]

[0263] Where, θ t Let θ be the downslope angle. It can be seen that the first mean is obtained by taking the first value θ1 and the second value θ2, and then subtracting 90 from the first mean to calculate the downslope angle.

[0264] Furthermore, in the embodiments of this application, the method by which the terminal calculates the horizontal beamwidth and azimuth angle based on the maximum downtilt angle, i.e., the method proposed in step 401c2, may include the following steps:

[0265] Step 601: Obtain the gain value corresponding to the direction angle on the plane corresponding to the maximum downtilt angle.

[0266] In the embodiments of this application, the terminal calculates the horizontal beamwidth and azimuth angle based on the maximum downtilt angle. Specifically, the terminal can first obtain the gain value corresponding to the azimuth angle on the plane corresponding to the maximum downtilt angle.

[0267] For example, in an embodiment of this application, the maximum downtilt angle is θ. max Let θ = θ max Plane, determine the direction angle on that plane. Gain values ​​at each point

[0268] Step 602: Determine the second gain value among the gain values ​​corresponding to the direction angle; wherein, the second gain value is the largest gain value among the gain values ​​corresponding to the direction angle.

[0269] In the embodiments of this application, after obtaining the gain value corresponding to the azimuth angle on the plane corresponding to the maximum tilt angle, the terminal can determine the second gain value among the gain values ​​corresponding to the azimuth angle; wherein, the second gain value is the largest gain value among the gain values ​​corresponding to the azimuth angle.

[0270] It should be noted that, in the embodiments of this application, the second gain value is the largest gain value among the gain values ​​corresponding to the direction angle.

[0271] For example, in an embodiment of this application, the gain value corresponding to the direction angle... Perform a search to determine the maximum value. Will As the second gain value.

[0272] Step 603: Normalize the second gain value to obtain the second result.

[0273] In the embodiments of this application, after determining the second gain value among the gain values ​​corresponding to the azimuth angle, the terminal can perform normalization processing based on the second gain value to obtain the second result.

[0274] For example, in an embodiment of this application, the method for normalizing based on the second gain value can be expressed as the following formula:

[0275]

[0276] Here, F is the second result.

[0277] Step 604: Perform search processing on the second result according to the preset standard to determine the third and fourth values.

[0278] In the embodiments of this application, after the terminal performs normalization processing based on the second gain value to obtain the second result, it can perform search processing on the second result according to a preset standard to determine the third and fourth values.

[0279] For example, in an embodiment of this application, the preset standard is to satisfy F ≤ -3; from For the corresponding angle, look to both sides for a second result that satisfies F ≤ -3, and determine it as the third value. and the fourth value

[0280] Step 605: Obtain the second absolute value of the subtraction result between the third and fourth values, and determine the second absolute value as the horizontal wave width.

[0281] In the embodiments of this application, after the terminal performs a search process on the second result according to a preset standard to determine the third and fourth values, it can obtain the second absolute value of the subtraction result of the third and fourth values ​​and determine the second absolute value as the horizontal wavewidth.

[0282] For example, in an embodiment of this application, the third numerical value... and the fourth value Subtracting the two values ​​and taking their absolute values, the second absolute value can be expressed as: Thus can The result was determined to be the horizontal wavewidth.

[0283] Step 606: Obtain the second mean of the third and fourth values, and determine the direction angle based on the second mean.

[0284] In the embodiments of this application, after the terminal performs search processing on the second result according to the preset standard to determine the third and fourth values, it can obtain the second average of the third and fourth values ​​and determine the direction angle based on the second average.

[0285] For example, in an embodiment of this application, the method for calculating the direction angle can be expressed as the following formula:

[0286]

[0287] in, This is the direction angle.

[0288] Furthermore, in embodiments of this application, the method for the terminal to calculate the transmission power based on the initial antenna weights may include the following steps:

[0289] Step 701: Determine the amplitude information corresponding to the channel based on the initial antenna weights.

[0290] In the embodiments of this application, the terminal calculates the transmission power based on the initial antenna weights. Specifically, the terminal can first determine the amplitude information corresponding to the channel based on the initial antenna weights.

[0291] It should be noted that, in the embodiments of this application, the amplitude information corresponding to the channel can be determined based on the initial antenna weights; for example, the amplitude information corresponding to the channel is represented as a1, a2, ... a32.

[0292] Step 702: Calculate the transmission power based on the amplitude information corresponding to the channel.

[0293] In the embodiments of this application, after the terminal determines the amplitude information corresponding to the channel based on the initial antenna weights, it can calculate the transmission power based on the amplitude information corresponding to the channel.

[0294] For example, in an embodiment of this application, calculating the transmission power based on the amplitude information corresponding to the channel can be expressed as: (a1) 2 +a2 2 +...a32 2 ) / m*n; where m and n are the channel numbers.

[0295] In summary, in the embodiments of this application, a beam simulation device or module can be set up to implement simulation processing; for example, Figure 12 This is a schematic diagram of the implementation process of the beamforming method proposed in the embodiments of this application. Figure 10 ,like Figure 12As shown, firstly, the simulation parameters can be initialized; then, the antenna weights sent by the weight optimization device can be received; the weight matrix can be extracted and relevant information can be determined; then, the field strength information, transmission power, and sidelobe suppression value can be calculated; the maximum beam gain can be calculated, and the corresponding maximum azimuth angle and maximum downtilt angle can be determined; then, the vertical beamwidth, downtilt angle, horizontal beamwidth, and azimuth angle can be calculated based on the maximum azimuth angle and maximum downtilt angle; finally, the above-mentioned maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power are sent to the weight optimization device as simulation data.

[0296] For example, in the embodiments of this application, Figure 13 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 3 ,like Figure 13 As shown, this application can set up a weight optimization device. The weight optimization device can obtain antenna input data to calculate and determine target weight data. It mainly includes calculation processing based on the objective function model, weight optimization processing, and saving the target weight data. Figure 14 This is a schematic diagram illustrating the implementation of the beamforming method proposed in the embodiments of this application. Figure 4 ,like Figure 14 As shown, this application can include a beam simulation device, which can be used to calculate the beam corresponding to the weight data output by the weight optimization device. The simulation mainly involves processing the weight data sent by the weight optimization device, including calculating the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, sidelobe suppression value, and transmission power, thereby obtaining the simulation results. Therefore, this application can achieve the acquisition of target weight data by setting up a weight optimization device and a beam simulation device. In practical applications, even active antenna units (AAUs) from different manufacturers and of different models can obtain the same target beam after determining the corresponding target antenna weights, achieving cross-manufacturer compatibility.

[0297] This application provides a beamforming method in which a terminal acquires antenna input data. The antenna input data includes target antenna data, target antenna model, and initial antenna weights. Simulation processing is performed based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Finally, target weight data is determined based on the updated antenna data for beamforming. Therefore, in this application, antenna input data, including target antenna data, target antenna model, and initial antenna weights, can be obtained first. Then, simulation processing can be performed based on the antenna input data to obtain initial simulation data corresponding to the initial antenna weights, and the initial function value corresponding to the initial simulation data can be calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing can continue based on the antenna input data until the target weight data is obtained. Beamforming can then be performed based on the target weight data. In other words, by calculating and obtaining the target weight data in the above manner, this application can effectively improve the beamforming effect and is conducive to the fine optimization of large-scale antenna technology. At the same time, by considering the target antenna model in the antenna input data, this application can make the beamforming method more versatile.

[0298] Example 3

[0299] Based on the above embodiments, in another embodiment of this application... Figure 15 This is a schematic diagram of the terminal structure proposed in the embodiments of this application. Figure 1 ,like Figure 15 As shown, the terminal 10 proposed in this application embodiment may include an acquisition unit 11, a calculation unit 12, a processing unit 13, and a determination unit 14.

[0300] The acquisition unit 11 is used to acquire antenna input data; wherein the antenna input data includes target antenna data, target antenna model and initial antenna weight; and to perform simulation processing based on the initial antenna weight to obtain initial simulation data corresponding to the initial antenna weight.

[0301] The calculation unit 12 is used to calculate the initial function value corresponding to the initial simulation data.

[0302] The processing unit 13 is configured to perform weight optimization processing based on the antenna input data to obtain updated antenna data if the initial function value is greater than or equal to a preset threshold value, until the updated function value corresponding to the updated antenna data is less than the preset threshold value, and then determine target weight data based on the updated antenna data for beamforming.

[0303] Further, the processing unit 13 is specifically used to perform weight optimization processing on the antenna input data using a preset optimization model to obtain the updated antenna data; wherein, the preset optimization model includes a preset policy network and a preset evaluation network; the preset policy network includes a first policy network and a second policy network; the preset evaluation network includes a first evaluation network and a second evaluation network.

[0304] Furthermore, the processing unit 13 is specifically configured to determine state data based on the antenna input data; perform calculation processing on the state data using the first policy network to obtain action data; and determine the updated antenna data based on the action data.

[0305] Furthermore, the processing unit 13 is also configured to, after determining the updated antenna data based on the action data, perform simulation processing based on the updated antenna data and the state data to obtain simulation results; wherein, the simulation results include first simulation data corresponding to the updated antenna data and second simulation data corresponding to the state data; calculate a first function value corresponding to the simulation results; if the first function value is greater than or equal to the preset threshold value, then perform subsequent processing based on the state data and the updated antenna data to obtain subsequent state data, until the subsequent function value corresponding to the subsequent state data is less than the preset threshold value, and determine the target weight data based on the subsequent state data.

[0306] Furthermore, the processing unit 13 is specifically configured to: use the first evaluation network to evaluate the state data and the action data to obtain an evaluation value; use the second policy network to calculate the updated antenna data to obtain updated action data; use the second evaluation network to evaluate the updated antenna data and the updated action data to obtain an updated evaluation value; calculate a loss function value based on the evaluation value, the updated evaluation value, and the reward value; update the preset policy network and the preset evaluation network based on the loss function value and the evaluation value to obtain an updated policy network and an updated evaluation network; and obtain the subsequent state data based on the updated policy network, the updated evaluation network, and the updated antenna data.

[0307] Furthermore, the calculation unit 12 is also configured to calculate the reward value based on the simulation results before the processing unit 13 calculates the loss function value based on the evaluation value, the updated evaluation value, and the reward value.

[0308] Further, the calculation unit 12 is specifically used to calculate a first result value corresponding to the first simulation data based on the first simulation data and the target antenna data in the simulation results; and to calculate a second result value corresponding to the second simulation data based on the second simulation data and the target antenna data in the simulation results; and to subtract the first result value and the second result value to obtain the reward value.

[0309] Furthermore, the calculation unit 12 is specifically used to perform distance calculation processing on the initial simulation data and the target antenna data based on the objective function model to obtain the initial function value.

[0310] Furthermore, the determining unit 14 is configured to, after the calculation unit 12 calculates the initial function value corresponding to the initial simulation data, if the initial function value is less than the preset threshold value, determine the target weight data based on the initial antenna weights for beamforming.

[0311] Furthermore, the processing unit 13 is specifically configured to update the first evaluation network according to the loss function value to obtain an updated first evaluation network; update the first policy network according to the evaluation value to obtain an updated first policy network; and update the second policy network and the second evaluation network according to the updated first policy network and the updated first evaluation network to obtain an updated second policy network and an updated second evaluation network.

[0312] Furthermore, the processing unit 13 is specifically configured to determine the updated policy parameters corresponding to the updated first policy network and the updated evaluation parameters corresponding to the updated first evaluation network; and to update the second policy network according to the updated policy parameters to obtain the updated second policy network; and to update the second evaluation network according to the updated evaluation parameters to obtain the updated second evaluation network.

[0313] Furthermore, the acquisition unit 11 is specifically used to calculate the field strength information and transmission power based on the initial antenna weights; and to determine the antenna pattern based on the field strength information; and to obtain the initial simulation data based on the antenna pattern and the transmission power, so as to realize the simulation processing.

[0314] Furthermore, the acquisition unit 11 is specifically used to determine the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern; and to obtain the initial simulation data based on the maximum beam gain, the horizontal beamwidth, the vertical beamwidth, the azimuth angle, the downtilt angle, the sidelobe suppression value, and the transmission power.

[0315] Furthermore, the acquisition unit 11 is specifically configured to determine the maximum beam gain and the sidelobe suppression value based on the antenna pattern; and to determine the corresponding maximum azimuth angle and maximum downtilt angle based on the maximum beam gain; and to determine the horizontal beamwidth, the vertical beamwidth, the azimuth angle, and the downtilt angle based on the maximum azimuth angle and the maximum downtilt angle.

[0316] Furthermore, the acquisition unit 11 is specifically used to calculate the vertical wavewidth and the downtilt angle based on the maximum azimuth angle; and to calculate the horizontal wavewidth and the azimuth angle based on the maximum downtilt angle.

[0317] Furthermore, the acquisition unit 11 is specifically configured to acquire the gain value corresponding to the downtilt angle on the plane corresponding to the maximum azimuth angle; and determine a first gain value among the gain values ​​corresponding to the downtilt angle; wherein the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle; and perform normalization processing based on the first gain value to obtain a first result; and perform search processing on the first result according to a preset standard to determine a first value and a second value; and acquire the first absolute value of the subtraction result of the first value and the second value, and determine the first absolute value as the vertical wavewidth; and acquire the first mean of the first value and the second value, and determine the downtilt angle based on the first mean.

[0318] Furthermore, the acquisition unit 11 is specifically configured to acquire the gain value corresponding to the azimuth angle on the plane corresponding to the maximum downtilt angle; and determine a second gain value among the gain values ​​corresponding to the azimuth angle; wherein the second gain value is the largest gain value among the gain values ​​corresponding to the azimuth angle; and perform normalization processing based on the second gain value to obtain a second result; and perform search processing on the second result according to a preset standard to determine a third value and a fourth value; and acquire the second absolute value of the subtraction result of the third value and the fourth value, and determine the horizontal wavewidth based on the second absolute value; and acquire the second mean of the third value and the fourth value, and determine the azimuth angle based on the second mean.

[0319] Furthermore, the acquisition unit 11 is specifically used to determine the amplitude information corresponding to the channel based on the initial antenna weights; and to calculate the transmission power based on the amplitude information corresponding to the channel.

[0320] Figure 16 This is a schematic diagram of the terminal structure proposed in the embodiments of this application. Figure 2 ,like Figure 16 As shown, the terminal 10 proposed in this application embodiment may further include a processor 15, a memory 16 storing instructions executable by the processor 15, and further, the terminal 10 may also include a communication interface 17 and a bus 18 for connecting the processor 15, the memory 16 and the communication interface 17.

[0321] In the embodiments of this application, the processor 15 can be at least one of the following: Application-Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), Central Processing Unit (CPU), Controller, Microcontroller, and Microprocessor. It is understood that for different devices, the electronic device used to implement the above-mentioned processor function can also be other, and this application embodiment does not specifically limit it. The processor 15 may also include a memory 16, which can be connected to the processor 15. The memory 16 is used to store executable program code, which includes computer operation instructions. The memory 16 may include high-speed RAM memory and may also include non-volatile memory, such as at least two disk drives.

[0322] In embodiments of this application, bus 18 is used to connect communication interface 17, processor 15, and memory 16, as well as the mutual communication between these devices.

[0323] In embodiments of this application, memory 16 is used to store instructions and data.

[0324] Furthermore, in the embodiments of this application, the processor 15 is used to acquire antenna input data; wherein, the antenna input data includes target antenna data, target antenna model, and initial antenna weights;

[0325] The initial antenna weights are used for simulation processing to obtain the initial simulation data corresponding to the initial antenna weights.

[0326] Calculate the initial function values ​​corresponding to the initial simulation data;

[0327] If the initial function value is greater than or equal to the preset threshold value, then weight optimization processing is performed based on the antenna input data to obtain updated antenna data for beamforming.

[0328] In practical applications, the aforementioned memory 16 can be volatile memory, such as random-access memory (RAM); or non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); or a combination of the above types of memory, and provide instructions and data to the processor 15.

[0329] Furthermore, in this embodiment, the functional modules can be integrated into one analysis unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional module.

[0330] If the integrated unit is implemented as a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the method of this embodiment. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0331] This application provides a terminal that acquires antenna input data, including target antenna data, target antenna model, and initial antenna weights. Simulation processing is performed based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights. An initial function value corresponding to the initial simulation data is calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Finally, target weight data is determined based on the updated antenna data for beamforming. Therefore, in this application, antenna input data, including target antenna data, target antenna model, and initial antenna weights, can be obtained first. Then, simulation processing can be performed based on the antenna input data to obtain initial simulation data corresponding to the initial antenna weights, and the initial function value corresponding to the initial simulation data can be calculated. If the initial function value is greater than or equal to a preset threshold, weight optimization processing can continue based on the antenna input data until the target weight data is obtained. Beamforming can then be performed based on the target weight data. In other words, by calculating and obtaining the target weight data in the above manner, this application can effectively improve the beamforming effect and is conducive to the fine optimization of large-scale antenna technology. At the same time, by considering the target antenna model in the antenna input data, this application can make the beamforming method more versatile.

[0332] Specifically, the program instructions corresponding to a beamforming method in this embodiment can be stored on storage media such as optical discs, hard disks, and USB flash drives; when the program instructions corresponding to a beamforming method in the storage media are read or executed by an electronic device, the following steps are included:

[0333] Acquire antenna input data; wherein, the antenna input data includes target antenna data, target antenna model, and initial antenna weights;

[0334] The initial antenna weights are used for simulation processing to obtain the initial simulation data corresponding to the initial antenna weights.

[0335] Calculate the initial function values ​​corresponding to the initial simulation data;

[0336] If the initial function value is greater than or equal to the preset threshold, then weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Then, target weight data is determined based on the updated antenna data for beamforming.

[0337] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0338] This application is described with reference to schematic and / or block diagrams of implementations of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block of the schematic and / or block diagrams can be implemented by computer program instructions, and combinations of blocks in the schematic and / or block diagrams can be implemented. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the schematic and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0339] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in the implementation flow diagram. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0340] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0341] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A beamforming method, characterized in that, The method includes: Acquire antenna input data; wherein, the antenna input data includes target antenna data, target antenna model, and initial antenna weights; The initial antenna weights are used for simulation processing to obtain the initial simulation data corresponding to the initial antenna weights. Calculate the initial function values ​​corresponding to the initial simulation data; If the initial function value is greater than or equal to a preset threshold, then weight optimization processing is performed based on the antenna input data to obtain updated antenna data. This process continues until the updated function value corresponding to the updated antenna data is less than the preset threshold. Then, target weight data is determined based on the updated antenna data for beamforming. The function value is used to measure the distance between the sub-beam corresponding to the currently obtained antenna weight and the target sub-beam corresponding to the target antenna data. The step of performing weight optimization processing based on the antenna input data to obtain updated antenna data includes: The antenna input data is processed by weight optimization using a preset optimization model to obtain the updated antenna data; wherein, the preset optimization model is a deep reinforcement learning algorithm; the preset optimization model includes a preset policy network and a preset evaluation network; the preset policy network includes a first policy network and a second policy network; the preset evaluation network includes a first evaluation network and a second evaluation network.

2. The method according to claim 1, characterized in that, The step of performing weight optimization processing based on the antenna input data to obtain updated antenna data includes: Based on the antenna input data, determine the status data; The state data is processed using the first policy network to obtain action data; Based on the action data, the updated antenna data is determined.

3. The method according to claim 2, characterized in that, After determining the updated antenna data based on the action data, the method further includes: Simulation processing is performed based on the updated antenna data and the state data to obtain simulation results; wherein, the simulation results include first simulation data corresponding to the updated antenna data and second simulation data corresponding to the state data; Calculate the first function value corresponding to the simulation result; If the first function value is greater than or equal to the preset threshold value, then subsequent processing is performed based on the state data and the updated antenna data to obtain subsequent state data, until the subsequent function value corresponding to the subsequent state data is less than the preset threshold value, and then the target weight data is determined based on the subsequent state data.

4. The method according to claim 3, characterized in that, The subsequent processing based on the state data and the updated antenna data to obtain subsequent state data includes: The first evaluation network is used to evaluate the state data and the action data to obtain an evaluation value. The updated antenna data is processed using the second strategy network to obtain updated action data. The updated antenna data and the updated motion data are evaluated using the second evaluation network to obtain an updated evaluation value. Calculate the loss function value based on the evaluation value, the updated evaluation value, and the reward value; The preset policy network and the preset evaluation network are updated based on the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network. The subsequent state data is obtained based on the updated policy network, the updated evaluation network, and the updated antenna data.

5. The method according to claim 4, characterized in that, Before calculating the loss function value based on the evaluation value, the updated evaluation value, and the reward value, the method further includes: The reward value is calculated based on the simulation results.

6. The method according to claim 5, characterized in that, The step of calculating the reward value based on the simulation results includes: Based on the first simulation data and the target antenna data in the simulation results, calculate the first result value corresponding to the first simulation data; Based on the second simulation data and the target antenna data in the simulation results, calculate the second result value corresponding to the second simulation data; The reward value is obtained by subtracting the first result value from the second result value.

7. The method according to claim 1 or 3, characterized in that, The calculation of the initial function value corresponding to the initial simulation data includes: Based on the objective function model, the initial simulation data and the target antenna data are processed to calculate the distance and obtain the initial function value.

8. The method according to claim 1 or 3, characterized in that, After calculating the initial function value corresponding to the initial simulation data, the method further includes: If the initial function value is less than the preset threshold value, the target weight data is determined based on the initial antenna weights for beamforming.

9. The method according to claim 4, characterized in that, The step of updating the preset policy network and the preset evaluation network based on the loss function value and the evaluation value to obtain the updated policy network and the updated evaluation network includes: The first evaluation network is updated based on the loss function value to obtain the updated first evaluation network; The first policy network is updated based on the evaluation value to obtain the updated first policy network; The second policy network and the second evaluation network are updated based on the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network.

10. The method according to claim 9, characterized in that, The step of updating the second policy network and the second evaluation network based on the updated first policy network and the updated first evaluation network to obtain the updated second policy network and the updated second evaluation network includes: Determine the updated policy parameters corresponding to the updated first policy network, and the updated evaluation parameters corresponding to the updated first evaluation network; The second policy network is updated according to the updated policy parameters to obtain the updated second policy network; The second evaluation network is updated according to the updated evaluation parameters to obtain the updated second evaluation network.

11. The method according to claim 1 or 3, characterized in that, The step of performing simulation processing based on the initial antenna weights to obtain the initial simulation data corresponding to the initial antenna weights includes: Calculate the field strength information and transmission power based on the initial antenna weights; Based on the field strength information, determine the antenna pattern; Based on the antenna pattern and the transmission power, the initial simulation data is obtained to implement the simulation processing.

12. The method according to claim 11, characterized in that, The process of obtaining the initial simulation data based on the antenna pattern and the transmission power includes: The maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value are determined based on the antenna pattern. The initial simulation data is obtained based on the maximum beam gain, the horizontal beamwidth, the vertical beamwidth, the azimuth angle, the downtilt angle, the sidelobe suppression value, and the transmission power.

13. The method according to claim 12, characterized in that, The determination of the maximum beam gain, horizontal beamwidth, vertical beamwidth, azimuth angle, downtilt angle, and sidelobe suppression value based on the antenna pattern includes: The maximum beam gain and the sidelobe suppression value are determined based on the antenna pattern. The maximum azimuth angle and maximum downtilt angle are determined based on the maximum beam gain. The horizontal wavewidth, the vertical wavewidth, the direction angle, and the downtilt angle are determined based on the maximum azimuth angle and the maximum downtilt angle.

14. The method according to claim 12, characterized in that, The step of determining the horizontal wavewidth, the vertical wavewidth, the azimuth angle, and the downtilt angle based on the maximum azimuth angle and the maximum downtilt angle includes: Based on the maximum azimuth angle, calculate the vertical beamwidth and the downtilt angle; Based on the maximum downtilt angle, the horizontal wavewidth and the azimuth angle are calculated.

15. The method according to claim 14, characterized in that, The calculation of the vertical beamwidth and the downtilt angle based on the maximum azimuth angle includes: Obtain the gain value corresponding to the downtilt angle on the plane corresponding to the maximum direction angle; Determine the first gain value among the gain values ​​corresponding to the downtilt angle; wherein the first gain value is the largest gain value among the gain values ​​corresponding to the downtilt angle. The first gain value is normalized to obtain the first result; The first result is searched and processed according to a preset standard to determine the first value and the second value. Obtain the first absolute value of the subtraction result between the first value and the second value, and determine the first absolute value as the vertical wavewidth; Obtain the first average of the first value and the second value, and determine the downtilt angle based on the first average.

16. The method according to claim 14, characterized in that, The calculation of the horizontal wavewidth and the azimuth angle based on the maximum downtilt angle includes: Obtain the gain value corresponding to the orientation angle on the plane corresponding to the maximum downtilt angle; Determine the second gain value among the gain values ​​corresponding to the direction angle; wherein, the second gain value is the largest gain value among the gain values ​​corresponding to the direction angle; The second result is obtained by normalizing the second gain value. The second result is searched and processed according to the preset criteria to determine the third and fourth values; Obtain the second absolute value of the subtraction result between the third value and the fourth value, and determine the second absolute value as the horizontal wave width; Obtain the second average of the third and fourth values, and determine the direction angle based on the second average.

17. The method according to claim 11, characterized in that, The method includes: The amplitude information corresponding to the channel is determined based on the initial antenna weights; The transmission power is calculated based on the amplitude information corresponding to the channel.

18. A terminal, characterized in that, The terminal includes an acquisition unit, a calculation unit, and a processing unit. The acquisition unit is used to acquire antenna input data; wherein the antenna input data includes target antenna data, target antenna model and initial antenna weights; and to perform simulation processing based on the initial antenna weights to obtain initial simulation data corresponding to the initial antenna weights; The calculation unit is used to calculate the initial function value corresponding to the initial simulation data; The processing unit is configured to perform weight optimization processing based on the antenna input data to obtain updated antenna data if the initial function value is greater than or equal to a preset threshold value, until the updated function value corresponding to the updated antenna data is less than the preset threshold value, and then determine target weight data based on the updated antenna data for beamforming; the function value is used to measure the distance between the sub-beam corresponding to the currently obtained antenna weight value and the target sub-beam corresponding to the target antenna data; The step of performing weight optimization processing based on the antenna input data to obtain updated antenna data includes: The antenna input data is processed by weight optimization using a preset optimization model to obtain the updated antenna data; wherein, the preset optimization model is a deep reinforcement learning algorithm; the preset optimization model includes a preset policy network and a preset evaluation network; the preset policy network includes a first policy network and a second policy network; the preset evaluation network includes a first evaluation network and a second evaluation network.

19. A terminal, characterized in that, The terminal further includes a processor and a memory storing processor-executable instructions, which, when executed by the processor, implement the method as described in any one of claims 1-17.

20. A computer-readable storage medium having a program stored thereon, applicable to a terminal, wherein when the program is executed by a processor, it implements the method as described in any one of claims 1-17.

Images