Network optimization method and device, storage medium and program product

Abstract

The embodiment of the invention provides a network optimization method and device, a storage medium and a program product, and relates to the technical field of communication. The method comprises the following steps: acquiring a digital twin model of target network equipment; acquiring terminal data reported by a user terminal in a to-be-optimized signal range of the target network equipment; according to the terminal data, performing beam prediction on the target network equipment through the digital twin model to obtain target beam prediction data; and according to the target beam prediction data, carrying out optimization processing on the beam of the target network equipment. According to the embodiment of the invention, the target beam prediction data in the to-be-optimized signal range can be obtained only through a small amount of terminal data reported by the user terminal, the measurement data volume of the terminal data is shortened, and the utilization rate of measurement resources is higher; network optimization can be carried out on the network equipment under the condition of considering the beam utilization rate and the optimization efficiency.

Description

Technical Field

[0001] This application relates to, but is not limited to, the field of communication technology, and in particular to a method, apparatus, storage medium, and program product for network optimization. Background Technology

[0002] In the field of communication technology, network devices provide more stable and higher-quality communication services to user terminals by improving their network quality (such as capacity and coverage). Network quality is often related to the beamforming of network devices. However, in practical applications, network quality is often determined based on beamforming and management. Beamforming and management, however, rely heavily on a large amount of measurement information. In scenarios with many or mobile user terminals, frequent beamforming measurements are required, leading to low utilization of measurement resources and low optimization efficiency due to the large amount of measurement information. Therefore, how to optimize network devices while balancing measurement resource utilization and optimization efficiency is a pressing technical problem that needs to be solved. Summary of the Invention

[0003] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0004] This application provides a method, device, storage medium, and program product for network optimization, which can optimize network devices while taking into account both measurement resource utilization and optimization efficiency.

[0005] In a first aspect, a network optimization method according to an embodiment of this application includes: acquiring a digital twin model of a target network device; acquiring terminal data reported by user terminals located within the signal range to be optimized of the target network device; performing beam prediction on the target network device using the digital twin model based on the terminal data to obtain target beam prediction data; and optimizing the beam of the target network device based on the target beam prediction data.

[0006] In a second aspect, a network device according to an embodiment of this application includes at least one processor and at least one memory for storing at least one program; when at least one of the programs is executed by at least one of the processors, it implements the network optimization method as described in any one of the first aspects.

[0007] Thirdly, according to embodiments of the present application, a computer-readable storage medium stores computer-executable instructions for performing the network optimization method described in any of the first aspects.

[0008] Fourthly, a computer program product according to an embodiment of this application includes a computer program or computer instructions stored in a computer-readable storage medium, a processor of a network device reading the computer program or computer instructions from the computer-readable storage medium, and the processor executing the computer program or computer instructions to cause the network device to perform a network optimization method as described in any of the first aspects.

[0009] This application embodiment uses a digital twin model corresponding to the target network device to perform beam prediction on terminal data reported by user terminals within the signal range to be optimized, obtaining target beam prediction data. Based on the target beam prediction data, the beam of the target network device is optimized. In this way, only a small amount of terminal data reported by user terminals is needed to obtain the target beam prediction data within the signal range to be optimized, reducing the amount of terminal data measurement and improving measurement resource utilization. Therefore, this application embodiment can optimize network devices while balancing beam utilization and optimization efficiency. Attached Figure Description

[0010] Figure 1 A schematic diagram of the system framework of an embodiment of the network optimization system provided in this application;

[0011] Figure 2 A schematic diagram of the system framework of another embodiment of the network optimization system provided in this application;

[0012] Figure 3 A flowchart illustrating an embodiment of the network optimization method provided in this application;

[0013] Figure 4 A schematic diagram illustrating a scenario in which the network optimization method provided in this application is applied;

[0014] Figure 5 Another scenario illustration illustrating the application of the network optimization method provided in this application;

[0015] Figure 6 Another scenario illustration illustrating the application of the network optimization method provided in this application;

[0016] Figure 7 A schematic diagram of the device hardware structure corresponding to the device management method provided in this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0018] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0019] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0020] In the field of communication technology, beam management is a key technology in 5G NR (New Radio) networks. It dynamically selects the beam direction for communication between each user equipment and network equipment based on channel measurement results. In practical applications, network equipment improves its network quality (such as capacity and coverage) to provide more stable and higher-quality communication services to user terminals. Network quality is often related to the beam of network equipment, especially for millimeter-wave systems. Due to significant propagation loss and attenuation, large-scale antenna arrays are typically installed at the receiver / transmitter to compensate for path loss, i.e., multiple-input multiple-output (MIMO) is introduced to improve signal gain, and beamforming technology is used to make the beam pointing more precise. Beam management and beamforming both rely on a large amount of measurement information. When there are many users or users are in mobile scenarios, beam information needs to be measured more frequently, leading to significant overhead and latency. Therefore, how to optimize network equipment while balancing measurement resource utilization and efficiency is a pressing technical problem that needs to be solved. Based on this, embodiments of this application provide a method, device, storage medium, and program product for network optimization, which can optimize network devices while taking into account both measurement resource utilization and optimization efficiency.

[0021] Reference Figure 1 As shown, a network optimization system according to an embodiment of this application includes a target network device and a user terminal. The target network device and the user terminal are communicatively connected. The target network device performs the following steps: obtaining a digital twin model of the target network device; obtaining terminal data reported by the user terminal located within the signal range to be optimized of the target network device; performing beam prediction on the target network device based on the terminal data and the digital twin model to obtain target beam prediction data; and optimizing the beam of the target network device based on the target beam prediction data.

[0022] Reference Figure 2 As shown, a network optimization system according to an embodiment of this application includes a third-party device, a target network device, and a user terminal. The target network device is communicatively connected to the user terminal and the third-party device. The third-party device acquires a digital twin model of the target network device; acquires terminal data reported by the user terminal located within the signal range to be optimized of the target network device; performs beam prediction on the target network device based on the terminal data and the digital twin model to obtain target beam prediction data; and optimizes the beam of the target network device based on the target beam prediction data.

[0023] The target network device can be an access network device or core network device such as NR or LTE, or other network devices such as vehicle networking or satellite communication; the user terminal can be a mobile phone, computer, tablet, etc.; the third-party device is relative to the target network device. One third-party device can be configured to correspond to multiple target network devices, that is, multiple digital twin models of target network devices can be deployed on one third-party device. The third-party device can be a general server, cloud server, or other network devices other than the target network device.

[0024] It should be noted that, for Figure 1 In this context, the digital twin model is deployed in the target network device. The target network device calls the deployed digital twin model and performs beam prediction based on the terminal data reported by user terminals within the signal range to be optimized, thus obtaining the target beam prediction data. Figure 2 In this case, the digital model is deployed on a third-party device. When the third-party device performs beam prediction on the target network device, it calls the digital twin model of the target network device and performs beam prediction based on the terminal data reported by the user terminal within the signal range to be optimized, thus obtaining the target beam prediction data.

[0025] It should be noted that the digital twin model is trained from the digital twin physical model built based on the digital twin technology of the target network device. Therefore, due to the differences in configuration and physical environment of each target network device, the digital twin model is set up in a one-to-one correspondence with the target network device.

[0026] It should be noted that after the antenna deployment of the target network device is completed, its coverage area is also clearly defined. This application is based on the optimization of the target device's beam within the coverage area. The signal range to be optimized can represent the entire signal coverage area of ​​the network device, or it can be a specific area within the entire signal coverage area of ​​the network device. Therefore, this application does not impose specific limitations on the signal range to be optimized. It should also be noted that the target beam prediction data is used to record and reflect the beam performance of the target network device. In some embodiments, the target beam prediction data includes the direction, signal strength, signal quality, and corresponding configuration parameters of each beam within the target device's coverage area. In other embodiments, the target beam prediction data includes the configuration parameters of the beam with the best performance within the coverage area. Configuration parameters include antenna element amplitude and phase. The target beam prediction data includes multiple sets of beamforming data distributed over time within the prediction duration (wherein, beamforming data includes beam direction, beam signal strength, signal quality, etc.).

[0027] The embodiments of this application are applicable to beam-based communication systems, such as NR base stations, satellite communication systems, vehicle-to-everything (V2X) networks, etc., and can effectively improve resource utilization and reduce communication latency.

[0028] Reference Figure 3 As shown, a network optimization method according to an embodiment of this application includes:

[0029] Step S100: Obtain the digital twin model of the target network device;

[0030] Step S200: Obtain terminal data reported by user terminals located within the signal range to be optimized of the target network device;

[0031] Step S300: Based on the terminal data, perform beam prediction on the target network device using a digital twin model to obtain target beam prediction data;

[0032] Step S400: Optimize the beam of the target network device based on the target beam prediction data.

[0033] Therefore, by using the digital twin model corresponding to the target network device to perform beam prediction on the terminal data reported by user terminals within the signal range to be optimized, target beam prediction data is obtained. Based on the target beam prediction data, the beam of the target network device is optimized. In this way, only a small amount of terminal data reported by user terminals is needed to obtain the target beam prediction data within the signal range to be optimized, reducing the amount of terminal data measurement and improving the utilization rate of measurement resources. Therefore, the embodiments of this application can optimize network devices while taking into account both beam utilization and optimization efficiency.

[0034] It should be noted that the target network device is a network device that uses a digital twin model for network optimization. When based on Figure 1 When the system deploys a digital twin model as shown, step S100 indicates that the target network device loads its own deployed digital twin model. When based on... Figure 2 When the system deploys a digital twin model, step S100 indicates that the third-party device runs the digital twin model corresponding to the target network device.

[0035] It should be noted that the digital twin model in step S100 is a pre-trained digital model. It is a model obtained by constructing a scenario based on digital twin technology and then training it using the current configuration data and measurement data of the network devices. The digital twin model can be trained online or offline. It should also be noted that the digital twin model is obtained by training the measurement data using a ray tracing propagation model to obtain the wireless signal propagation path and attenuation parameters. Therefore, training the digital twin model can be completed even with a small amount of measurement data.

[0036] It should be noted that the signal range to be optimized is the area within the coverage of the target network device that requires signal optimization.

[0037] It should be noted that the terminal data represents the performance of the user terminal. This application does not limit the number of user terminals or the duration of terminal data reporting.

[0038] It should be noted that the target beam prediction data reflects the beam performance of each beam of the target network device within its signal coverage area. This beam performance can be the performance over the prediction time period or the performance at the current moment. For example, in some embodiments, if terminal data is acquired at time t, the output target beam prediction data can be the direction of each beam, the corresponding signal strength, and the user terminal signal quality indicators at time t. In other embodiments, the output beamforming data can be multiple sets of beamforming data (such as beam direction, signal strength, etc.) that change over time within a duration of t+T.

[0039] It should be noted that, referring to Figure 1 When the digital twin model is deployed on the target network terminal, the relevant antenna parameters are directly configured in the target network device based on the target beam prediction data to process the beam direction, number, etc., such as element amplitude and phase. When referring to... Figure 2 As shown, when the digital twin model is deployed on a third-party device, the antenna parameters of the target network device are determined based on the target beam prediction data output by the third-party device, and the antenna parameters are configured on the target network device to achieve optimization.

[0040] It should be noted that the embodiments of this application use a small amount of terminal data as measurement data to replace the original beam measurement (such as beam measurement information used for spatial division decision) for beam management and beamforming, which can reduce the test resource overhead in the actual operation of the target network device and effectively improve resource utilization. The small amount of terminal data can also be achieved by extending the measurement cycle or reducing the number of measurements, thus reducing measurement resource overhead.

[0041] The network optimization method of this application embodiment can be used to dynamically select the beam most pointing to the user terminal from the configured preset beams, realizing dynamic beamforming. For example, based on the location information that changes over time within the predicted duration output by the digital twin model (which can indicate the location of the user terminal), combined with the output beam information (such as beam signal strength and signal quality), the amplitude and phase of the antenna elements are dynamically adjusted to generate a beam in real time, so that the generated beam accurately points to the user terminal. At this time, for the target network device, steps S200 to S400 can be repeated periodically to realize dynamic beamforming of the user terminal.

[0042] The network optimization method in this application embodiment can also be used for beam optimization in spatial multiplexing scenarios. For example, based on the beam information output by the digital twin model and the interference of related beams, it can determine whether different user terminals have spatial multiplexing conditions. If spatial multiplexing conditions are met, spatial multiplexing can be performed, thereby improving resource utilization.

[0043] The network optimization method in this application embodiment can also be used for beam optimization in scenarios with a large number of user equipment, such as selecting the optimal beam combination to cover user terminals based on the location information output by the digital twin model (such as the coverage range of each beam or the location of the terminal under the beam direction).

[0044] Therefore, this application does not limit the specific application scenarios. Those skilled in the art can choose to train the digital twin model according to actual needs, so that the target beam prediction data includes the data required for optimization processing.

[0045] It is understood that acquiring terminal data reported by user terminals located within the target network device's signal range to be optimized includes at least one of the following:

[0046] Within a preset time period, the terminal data reported by the user terminal is obtained multiple times from a predetermined location within the signal range to be optimized of the target network device.

[0047] Terminal data reported by user terminals is obtained from multiple predetermined locations within the signal range to be optimized of the target network device.

[0048] It should be noted that acquiring terminal data from the same predetermined location multiple times, as well as acquiring terminal data from multiple different predetermined locations, can enrich the input data of the digital twin model, thereby further ensuring the accuracy of the target beam prediction data.

[0049] It should be noted that in some embodiments, terminal data can be acquired multiple times for multiple different predetermined locations within a preset time period. In other embodiments, terminal data can be acquired multiple times only for the same predetermined location within a preset market. In still other embodiments, terminal data can be acquired only once for multiple different predetermined locations. Those skilled in the art can selectively configure these settings according to actual needs.

[0050] It should be noted that in practical applications, when terminal data from the same predetermined location is acquired multiple times, beam prediction can be performed on each acquired data set, resulting in multiple target beam prediction data sets. Alternatively, beam prediction can be performed simultaneously on terminal data from the same predetermined location acquired multiple times, resulting in a single target beam prediction data set. In practical applications, when terminal data is acquired from different predetermined locations, prediction can be performed on the data acquired from each location individually, or prediction can be performed simultaneously. Those skilled in the art can selectively configure this according to actual needs.

[0051] Understandably, in some embodiments, the target beam prediction data is a collection of beamforming data corresponding to multiple user terminals at different locations that change over time within the prediction duration. Taking dynamic beamforming in a mobile scenario as an example, refer to... Figure 4 As shown, it is possible to dynamically adjust the amplitude and phase of the array element based on the beam output from the digital twin model and the beam information within the predicted duration, thereby generating a precise beam pointing towards the user equipment in real time. Beamforming data is used to achieve beamforming.

[0052] as follows Figure 4 As shown, the user terminal is terminal device 1. The system acquires terminal data reported by terminal device 1 at its first position under beam 2. Based on the terminal data reported by terminal device 1 at the first position, the digital twin model outputs target beam prediction data to predict the beamforming data of terminal device 1's movement trajectory within the prediction time. The target beam prediction data includes the position information of the beam at different times during the prediction time and the corresponding beamforming data. At this point, the target network device does not need to rely on excessive user terminal measurement reports; a small amount of terminal data is sufficient to obtain the target beam prediction data. When terminal device 1 moves from the first position to the second position within the prediction time, the network device retrieves the beamforming data for the second position from the target beam prediction data generated at the first position. Figure 4The beamforming data corresponding to mid-beam 1 allows for network optimization based on the beamforming data of beam 2. Specifically, this involves dynamically adjusting the amplitude and phase information of the antenna array to ensure the beam is precisely pointed at terminal device 1. Repeating these steps enables the target network device to track the terminal device in real time, achieving dynamic beam adjustment for the user terminal.

[0053] It should be noted that the prediction duration can be output in units of time slots, subframes or frames, or fixed time periods. For example, the target beam prediction data can be a set of user terminal locations and beamforming data output in units of time slots within the prediction duration, or it can be a set of user terminal locations and beamforming data output in units of subframes or frames within the prediction duration, or it can be a set of user terminal locations and beamforming data output in units of a fixed time period within the prediction duration.

[0054] Understandably, target beam prediction data includes beamforming data; within the signal range to be optimized, there are multiple user terminals in different spatial directions that support spatial multiplexing. Based on the terminal data, beam prediction is performed on the target network equipment using a digital twin model to obtain target beam prediction data, including:

[0055] Based on terminal data from multiple user terminals in different spatial directions, beamforming prediction of the target network device is performed using a digital twin model to obtain beamforming data of candidate beams for each user terminal.

[0056] Based on the target beam prediction data, the beam of the target network device is optimized, including:

[0057] Based on beamforming data, determine the configurable spatially multiplexed user terminals within the signal range to be optimized for the target network device;

[0058] Beamforming processing is performed on the target network device based on beamforming data.

[0059] It should be noted that beamforming data is used for beamforming and includes at least one of the following: beam direction, signal strength, and coverage area. Since each user terminal performs beam prediction within the coverage area of ​​the target network device, it is possible to determine whether there is interference between the beams corresponding to the user terminals based on the beamforming data of two user terminals. For example, interference can be determined based on coverage area and beam direction, or by filtering based on coverage area, beam direction, and signal strength, etc.

[0060] For example, such as Figure 5As shown, in a spatial division multiplexing (SDM) scenario, terminal device 1 and terminal device 2 are user terminals in different spatial directions. Therefore, based on the beamforming data of terminal device 1 and terminal device 2 output by the digital twin model, it can be determined whether terminal device 1 and terminal device 2 meet the SDM conditions, and SDM multiplexing is performed on the user devices that meet the conditions. Compared to existing technologies that require extensive measurements to determine whether SDM conditions are met, the digital twin model can continuously output beamforming data in different beam directions with fewer measurements. Figure 5 As shown, the predicted mutual interference between beam 2 and beam 1 used by terminal device 1 and terminal device 2 is relatively low. Therefore, terminal device 1 and terminal device 2 can be configured with spatial multiplexing to improve resource utilization.

[0061] It should be noted that beamforming prediction can be continuously performed on multiple user terminals in different spatial directions within a preset duration. In this case, multiple target beam prediction data within the preset duration can be obtained. Alternatively, prediction can be performed on target beam data within a prediction duration after the terminal data is reported. Both the preset duration and the prediction duration can be output in units of time slots, subframes, frames, or a fixed time period.

[0062] Understandably, the target beam prediction data includes interference data; within the signal range to be optimized, there are multiple user terminals in different spatial directions, all supporting spatial multiplexing. Based on the terminal data, beam prediction is performed on the target network equipment using a digital twin model to obtain the target beam prediction data, including:

[0063] Based on terminal data from multiple user terminals in different spatial directions, beam interference analysis is performed on the target network device using a digital twin model to obtain interference data.

[0064] Optimizing the beam of the target network device based on the target beam prediction data also includes:

[0065] Based on the interference data, at least two terminal devices in different spatial directions where there is no interference are identified as configurable spatial multiplexing devices.

[0066] It should be noted that, in some embodiments, the digital twin model will predict the beam where the user terminal is located and then perform interference analysis based on the predicted beamforming data to obtain interference data.

[0067] like Figure 5As shown, if the digital twin model predicts that the mutual interference between the beams used by terminal device 1 and terminal device 2 is relatively low, it directly outputs the beamforming data of beams 1 and 2 that do not interfere with each other. In this case, the network device can configure terminal device 1 and terminal device 2 for spatial multiplexing, thereby improving resource utilization.

[0068] It should be noted that, in some embodiments, the digital twin model can output beamforming data and interference data of the user terminal in different directions.

[0069] Understandably, based on terminal data, beam prediction is performed on the target network device using a digital twin model to obtain target beam prediction data, including:

[0070] Based on terminal data, a first user terminal and a second user terminal are identified; wherein, the user density in the area where the first user terminal and the second user terminal are distributed meets a preset density threshold, the signal quality of the first user terminal meets a preset quality requirement, and the signal quality of the second user terminal does not meet the preset quality requirement.

[0071] Using a digital twin model, beam prediction is performed on the terminal data of the first user terminal and the second user terminal within a preset time period to obtain candidate beam prediction data that makes both the first user terminal and the second user terminal meet the preset quality requirements.

[0072] The candidate beam prediction data obtained within a preset time period are combined to obtain the target beam prediction data.

[0073] It should be noted that the density threshold can be selectively set according to actual needs. The preset quality requirements can also be selectively set according to requirements.

[0074] It should be noted that in some embodiments, once the network device configuration is determined, the coverage area of ​​the network device is basically determined. However, in practical applications, when there are a large number of users within this coverage area, some areas within the target network device's coverage area may still not achieve optimal coverage, such as the signal quality of user terminals at the edge not meeting preset quality requirements. In this case, based on the terminal data reported by the user terminals, candidate beam prediction data for multiple time units within a preset time period or multiple candidate beam prediction data at the current moment are output. The time unit can be a time slot, frame, subframe, or a fixed time period. Then, the optimal candidate beam prediction data is selected from the multiple candidate beam prediction data for network optimization. Each candidate beam prediction data reflects the signal strength of the beam in different beam directions output in the corresponding unit of time and the signal quality of the user terminal.

[0075] For example, refer to Figure 6As shown, the area covered by beam 1 of the target network device has a large number of users, and some areas of beam 1 may not achieve optimal coverage. Using a digital twin model, the model can provide target beam prediction data that satisfies the signal quality requirements of both the first and second user terminals. This target beam prediction data includes antenna parameters that satisfy the optimal beam for both user terminals. The target network terminal can then directly configure itself based on these antenna parameters to perform beamforming on beam 1.

[0076] Understandably, based on terminal data, beam prediction is performed on the target network device using a digital twin model to obtain target beam prediction data, including:

[0077] Based on terminal data, a first user terminal and a second user terminal are identified; wherein, the user density in the area where the first user terminal and the second user terminal are distributed meets a preset density threshold, the signal quality of the first user terminal meets a preset quality requirement, and the signal quality of the second user terminal does not meet the preset quality requirement.

[0078] Using a digital twin model, beam prediction is performed on the terminal data of the first user terminal and the second user terminal within a preset time period to obtain the first candidate beam prediction data of the first user terminal and the second candidate beam prediction data of the second user terminal.

[0079] The target beam prediction data is obtained based on the prediction data of the first candidate beam and the prediction data of the second candidate beam.

[0080] For example, such as Figure 6 As shown, the area covered by network device beam 1 has a large number of users, and some areas of this beam may not achieve optimal coverage. Using a digital twin model, the optimal beam configuration for all user terminals under the network device (i.e., the first candidate beam prediction data and the second candidate beam data), including the optimal beam configuration over a subsequent period, is output to obtain the target beam prediction data. Based on the target beam prediction data output by the digital twin model, the network device dynamically adjusts the overall coverage of this beam. In this case, with a large number of terminal devices, the network device dynamically adjusts the optimal beam coverage based on the optimized output of the digital twin model.

[0081] Understandably, digital twin models are trained through the following steps:

[0082] Acquire configuration data and sample measurement data of the target network device;

[0083] The digital twin model is trained by using the configuration data and sample measurement data to train the digital twin model determined by the twin physical model of the target network device.

[0084] It should be noted that the training of the digital twin model is based on the ray tracing propagation model to obtain the wireless signal propagation path and attenuation parameters. Therefore, after training based on measurement data, it is possible to predict information such as beam strength, signal strength, and signal quality at different locations within the overall coverage area of ​​the target network device. Thus, after the final trained digital twin model is deployed online, by inputting current terminal data, such as the current user's configuration and a small amount of measurement information, it can output target beam prediction data composed of terminal location and beam information.

[0085] It should be noted that the training and inference process of digital models can utilize artificial intelligence (AI) related technologies, including but not limited to deep learning methods such as Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Long Short-Term Memory (LSTM), or other non-AI methods such as multinomial fitting, Kalman filtering, and random forest algorithms.

[0086] It should be noted that the twin physical model is a digital twin scenario constructed based on the coverage area of ​​network equipment. Taking 5G NR (New Radio) system network equipment as an example, the digital twin scenario may include, but is not limited to, base station information, antenna information, and environmental information. Base station information may include, but is not limited to, the components of the base station and its location information; antenna information may include, but is not limited to, the number of antennas, antenna type, and antenna gain; environmental information may include, but is not limited to, buildings, roads, and trees surrounding the base station. Those skilled in the art can selectively set these parameters based on actual data.

[0087] It should be noted that the current configuration of the network equipment used for training may include, but is not limited to, the beam configuration, power configuration, cell configuration, etc. of the current environment.

[0088] It should be noted that the training process for the digital twin model can be conducted online or offline. The trained digital twin model can be deployed to network devices, servers, or other third-party deployable devices. Based on the current user terminal data (including status and a small amount of relevant measurement data), the deployed digital twin model can generate beamforming data or the user terminal's location information online and provide it to the network device. The network device selects the final beam based on the twin model's output and communicates with the user device using beamforming. In spatial division multiplexing scenarios, the network device can determine whether to perform spatial division multiplexing based on the twin model's output. In scenarios with a large number of users, the network device can also select the optimal beam combination to cover the user devices based on the twin model's output.

[0089] Understandably, the sample measurement data includes terminal sample data, which includes first detection reference signal quality index data and channel quality index data.

[0090] It should be noted that the channel quality index data is an index obtained by the user terminal based on channel measurements, such as the channel quality indicator (CQI) measured by the user equipment. The first sounding reference signal quality index data is an index obtained by the base station from the user terminal, which may include the precoding matrix indicator (PMI).

[0091] Understandably, the terminal sample data also includes user location data, downlink performance data, and beam parameters.

[0092] It should be noted that beam parameters may include the RSRP of the serving beam and surrounding beams; downlink performance data may include at least one of downlink RSRP, SINR, CQI, etc. It should also be noted that user location data may be in the form of GPS-measured latitude and longitude, angular information, or other location-representing information.

[0093] Understandably, the sample measurement data also includes network device measurement data of the target network device; the network device measurement data includes at least one of the second probe reference signal quality index data and physical uplink shared channel quality index data.

[0094] It should be noted that the first sounding reference signal quality index data is based on the sounding reference signal (SRS) measurement data, and may include at least one or more of the following: reference signal receiving power (RSRP), signal-to-interference plus noise ratio (SINR), and angle of arrival (DOA). The channel quality index data is based on the physical uplink shared channel (PUSCH) measurement data, and may include one or more of the following: RSRP, SINR, etc.

[0095] Understandably, the configuration data includes at least one of beam configuration parameters, power configuration parameters, and cell configuration parameters.

[0096] It should be noted that beam configurations include phase, dipole amplitude, etc., cell configurations include the cells covered by the beam, etc., and power configurations include the power level of the beam, etc. This application does not impose limitations on these aspects in its embodiments.

[0097] Understandably, the twin physical model includes at least one of the target network device's base station data, antenna data, and environmental data.

[0098] It is understood that base station data includes at least one of the following: base station location data and base station equipment composition data;

[0099] Antenna data includes at least one of the following: number of antennas, antenna gain, and antenna type;

[0100] Environmental data includes at least one of the following: natural environment data, building data, and road data.

[0101] Understandably, the sample measurement data is at least one of historical measurement data or real-time measurement data, and the real-time measurement data is the measurement data obtained by requesting the target network device to collect the data before the digital twin model starts training.

[0102] For example, the following example uses real-time measurement data for online training, but the specific details are not as follows:

[0103] Step 1: Collect network device operating parameters based on the digital twin model. The digital twin model requests or the target network device actively reports the operating parameters. The target network device reports the operating parameters to the digital twin model, thereby obtaining a digital model based on the physical twin model. The operating parameters include, but are not limited to, the latitude and longitude, altitude, azimuth, downtilt angle, antenna type, number of antennas, etc. of the network device.

[0104] Step 2: User terminals within the coverage area of ​​the target network device report sample measurement data to the digital twin model. Terminal data includes, but is not limited to, RSRP, SINR, DOA, CQI, etc.

[0105] Step 3: After receiving sufficient sample measurement data, the digital twin model starts training. The training methods include, but are not limited to, some AI algorithms such as LTSM and CNN, as well as some non-AI algorithms such as polynomial fitting and Kalman filtering.

[0106] Step 4: Once the digital twin model is trained, deploy it directly online.

[0107] Step 5: The digital twin model continuously receives terminal data reported by user terminals and infers target beam prediction data, and then performs further optimization based on the target beam prediction data.

[0108] Understandably, the method also includes:

[0109] Obtain optimized terminal performance data;

[0110] The optimized terminal performance data is compared with the terminal performance data before optimization to obtain the comparison results.

[0111] When the comparison results indicate that the performance of the target network device has deteriorated, optimization is performed using a digital twin model.

[0112] It should be noted that network devices can monitor key system indicators in real time by comparing optimized terminal data with unoptimized terminal data. This process monitors key communication indicators of the target network device before and after adjustment. Key indicators may include, but are not limited to, user terminal traffic, user terminal signal quality, user terminal block error rate, etc. This indicator information is fed back to the digital twin model, and the digital twin model can consider further optimization processing if the indicators deteriorate.

[0113] It should be noted that terminal performance data may or may not be included in the terminal data. Optimization using a digital twin model may include retraining the digital twin model based on existing data and making predictions based on the retrained model, or using the existing digital twin model to make predictions again.

[0114] Therefore, this application embodiment predicts beam-related data using a digital twin model, enabling network devices to dynamically adjust beam direction based on the output of the digital twin model. It also allows for real-time monitoring of key indicators of the communication system before and after adjustment, feeding these indicators back to the twin model for further optimization. This method effectively improves resource utilization and communication quality.

[0115] It is understood that a network device provided according to an embodiment of this application includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the device management method described above.

[0116] Understandably, referring to Figure 7 As shown, one embodiment of this application also provides a network device, including:

[0117] At least one processor 701;

[0118] At least one memory 702 is used to store at least one program, which implements the above-described device management method when the at least one program is executed by at least one processor 701.

[0119] The memory 702, as a non-transitory network system, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, the memory 702 may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 702 may optionally include remotely located memories 702 relative to the processor 701, which can be connected to the processor 701 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0120] The memory 702 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 702 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 702 and is called and executed by the processor 701.

[0121] The processor 701 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0122] In some embodiments, the network device further includes:

[0123] Input / output interfaces are used to implement information input and output;

[0124] The communication interface is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0125] The bus transmits information between various components of the device (such as processor 701, memory 702, input / output interface and communication interface);

[0126] The processor 701, memory 702, input / output interface, and communication interface can communicate with each other within the device via a bus.

[0127] Fourthly, one embodiment of this application also provides a computer-readable storage medium storing computer-executable instructions for performing the above-described method.

[0128] Fifthly, an embodiment of this application also provides a computer program product, including a computer program or computer instructions stored in a computer-readable storage medium, wherein a processor of a computer device reads the computer program or computer instructions from the computer-readable storage medium, and the processor executes the computer program or computer instructions to cause the computer device to perform the above-described method.

[0129] The system architecture and application scenarios described in this application are intended to more clearly illustrate the technical solutions of this application and do not constitute a limitation on the technical solutions provided in this application. Those skilled in the art will understand that as system architectures evolve and new application scenarios emerge, the technical solutions provided in this application are also applicable to similar technical problems.

[0130] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0131] The above description, with reference to the accompanying drawings, illustrates some embodiments of this application, but does not limit the scope of the invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of this invention should be considered within the scope of this application.

Claims

1. A network optimization method, the method comprising: obtaining a digital twin model of a target network device; obtaining terminal data reported by a user terminal located within a signal range to be optimized of the target network device; performing beam prediction on the target network device through the digital twin model according to the terminal data to obtain target beam prediction data; performing optimization processing on a beam of the target network device according to the target beam prediction data.

2. The method of network optimization of claim 1, wherein, The obtaining of the terminal data reported by the user terminal located within the signal range to be optimized of the target network device comprises at least one of the following: obtaining terminal data reported by a user terminal from a predetermined position within the signal range to be optimized of the target network device multiple times within a preset time period; obtaining terminal data reported by a user terminal from multiple different predetermined positions within the signal range to be optimized of the target network device.

3. The method of network optimization of claim 1, wherein, The target beam prediction data comprises beamforming data; the user terminals in multiple different spatial directions within the signal range to be optimized all support spatial division multiplexing, and the performing of the beam prediction on the target network device through the digital twin model according to the terminal data to obtain target beam prediction data comprises: performing beamforming prediction on the target network device through the digital twin model based on terminal data of the user terminals in multiple different spatial directions to obtain beamforming data of candidate beams corresponding to the user terminals; The performing of the optimization processing on the beam of the target network device according to the target beam prediction data comprises: determining user terminals that can be configured for spatial division multiplexing within the signal range to be optimized of the target network device according to the beamforming data; performing beamforming processing on the target network device according to the beamforming data.

4. The method of network optimization of claim 1, wherein, The target beam prediction data comprises interference data; the user terminals in multiple different spatial directions within the signal range to be optimized all support spatial division multiplexing, and the performing of the beam prediction on the target network device through the digital twin model according to the terminal data to obtain target beam prediction data comprises: performing beam interference analysis on the target network device through the digital twin model based on terminal data of the user terminals in multiple different spatial directions to obtain interference data; The performing of the optimization processing on the beam of the target network device according to the target beam prediction data further comprises: determining terminal devices in at least two different spatial directions in which no interference exists as user terminals that can be configured for spatial division multiplexing according to the interference data.

5. The method of network optimization of claim 1, wherein, The performing of the beam prediction on the target network device through the digital twin model according to the terminal data to obtain target beam prediction data comprises: determining a first user terminal and a second user terminal according to the terminal data; wherein a user density of a region in which the first user terminal and the second user terminal are jointly distributed satisfies a preset density threshold, a signal quality of the first user terminal satisfies a preset quality requirement, and a signal quality of the second user terminal does not satisfy the preset quality requirement. The terminal data of the first user terminal and the second user terminal within a preset time period is combined to obtain the target beam prediction data. The target network device is predicted by the digital twin model according to the terminal data to obtain target beam prediction data, including:

6. The method of network optimization of claim 1, wherein, The first user terminal and the second user terminal are determined according to the terminal data, wherein the user density of the area where the first user terminal and the second user terminal are distributed together meets a preset density threshold, the signal quality of the first user terminal meets a preset quality requirement, and the signal quality of the second user terminal does not meet the preset quality requirement; The terminal data of the first user terminal and the second user terminal within a preset time period is combined to obtain the target beam prediction data. The first candidate beam prediction data of the first user terminal and the second candidate beam prediction data of the second user terminal are obtained by the digital twin model. The target beam prediction data is obtained according to the first candidate beam prediction data and the second candidate beam prediction data.

7. The method of network optimization of claim 1, wherein, The digital twin model is obtained by the following steps: Configuration data and sample measurement data of the target network device are obtained; The configuration data and the sample measurement data are trained by a digital model determined based on a twin physical model of the target network device to obtain a trained digital twin model.

8. The method of network optimization according to claim 7, characterized in that, The sample measurement data includes terminal sample data, and the terminal sample data includes first sounding reference signal quality index data and channel quality index data.

9. The method of network optimization of claim 8, wherein, The terminal sample data also includes user location data, downlink performance data, and beam parameters.

10. The method of network optimization of claim 8, wherein, The sample measurement data also includes network device measurement data of the target network device; the network device measurement data includes at least one of second sounding reference signal quality index data and physical uplink shared channel quality index data.

11. The method of network optimization of claim 8, wherein, The configuration data includes at least one of beam configuration parameters, power configuration parameters, and cell configuration parameters.

12. The method of network optimization of claim 7, wherein, The twin physical model includes at least one of base station data, antenna data, and environment data of the target network device.

13. The method of network optimization of claim 12, wherein, The base station data includes at least one of base station location data and base station device composition data; The antenna data includes at least one of antenna quantity, antenna gain, and antenna type; The environment data includes at least one of natural environment data, building data, and road data.

14. The method of network optimization of claim 7, wherein, The sample measurement data is at least one of historical measurement data or real-time measurement data, and the real-time measurement data is measurement data collected by the target network device before the digital twin model starts training.

15. The method of network optimization of claim 1, wherein, The method further includes: Obtaining optimized terminal performance data; Comparing the optimized terminal performance data with the terminal performance data before optimization to obtain a comparison result; In a case where the comparison result indicates that the performance of the target network device is deteriorated, optimization processing is performed by the digital twin model.

16. A network device comprising: at least one processor; at least one memory for storing at least one program; when at least one of the programs is executed by the at least one processor, the method of network optimization according to any one of claims 1 to 15 is implemented.

17. A computer readable storage medium storing computer executable instructions for performing the method of network optimization according to any one of claims 1 to 15.

18. A computer program product comprising computer programs or computer instructions, characterized in that, The computer program or the computer instructions are stored in a computer readable storage medium, and the processor of the network device reads the computer program or the computer instructions from the computer readable storage medium, and executes the computer program or the computer instructions, so that the network device performs the method of network optimization according to any one of claims 1 to 15.

Images