Cell switching method and system based on cluster RSRP measurement and storage medium
By obtaining historical measurement values and establishing prediction models, and using spatial and temporal correlations to make predictions, the problem of large overhead of model deployment and management in the prior art is solved, and the prediction accuracy and effect of cell handover is improved.
Patent Information
- Application Number
- CN202411451975.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-17
- Publication Date
- 2025-02-14
AI Technical Summary
In the prior art, the cell handover method based on artificial intelligence has high overhead in model deployment and management, and it is difficult to adapt to the needs of terminals at different locations, resulting in low prediction accuracy and poor switching effect.
By obtaining historical measurements, establishing a prediction model, and using the spatial and temporal correlation of the data to predict, it reduces the training and deployment costs that require a large number of models to be established, and improves the prediction accuracy of the model.
It significantly improves the prediction accuracy and effect of cell handover, reduces the cost of model training and deployment, and enhances the generalization ability of the model.
Smart Images
Figure CN119450624A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of 5G communication technology, and in particular to a cell switching method, system, and storage medium based on cluster RSRP measurement. Background Art
[0002] UE L3 handover primarily involves communication between network nodes and mobility management of user equipment (UE). This process aims to ensure that communication continuity and service quality are maintained when the UE switches between different network cells. In related technologies, prediction methods for network-triggered L3 handovers typically employ AI-based cell handovers, where a model is trained for each cell to predict the cell's future RSRP. This approach incurs high deployment and management overhead for AI models, making it difficult to apply previously trained models to terminals in different locations. This presents significant obstacles in practical applications, resulting in low prediction accuracy and poor handover performance. Summary of the Invention
[0003] The purpose of the present invention is to solve one of the technical problems existing in the prior art to at least a certain extent.
[0004] Therefore, an object of the present invention is to provide a cell switching method, system and storage medium based on cluster RSRP measurement with high accuracy.
[0005] In order to achieve the above technical objectives, one aspect of an embodiment of the present invention provides a cell switching method based on cluster RSRP measurement, comprising the following steps: obtaining historical measurement values; the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes the first L1-RSRP of all beams; inputting the historical measurement values into a prediction model to obtain a prediction value; the prediction value is the second L1-RSRP of all cells within a second preset time period in the future; filtering the prediction value to obtain L3-RSRP; performing cell switching based on the L3-RSRP. The embodiment of the present application establishes a prediction model through the measurement information of all cells within the first preset time period, so that the model is established through the spatial correlation and temporal correlation of the data, alleviating the training and deployment costs of establishing a large number of models, and is conducive to improving the prediction accuracy of the model, thereby improving the cell switching effect.
[0006] In some embodiments, the cell handover method based on cluster RSRP measurement in an embodiment of the present invention obtains historical measurement values, including:
[0007] Get RSRP measurement value;
[0008] The RSRP measurement value is filtered by a clipping average filtering method to obtain a first L1-RSRP.
[0009] In some embodiments, in one embodiment of the present invention, filtering the RSRP measurement value by a clipping average filtering method to obtain a first L1-RSRP includes:
[0010] If the difference between the RSRP measurement values collected twice in succession is greater than a preset deviation, discard the RSRP measurement value collected second time in the two consecutive collections;
[0011] Put all the collected RSRP measurement values into an array, and replace the oldest RSRP measurement value with the latest collected RSRP measurement value;
[0012] Performing average filtering on the values in the array to obtain a first L1-RSRP.
[0013] In some embodiments, in one embodiment of the present invention, the prediction model is trained by the following steps:
[0014] Acquire measurement samples; the measurement samples include historical measurement samples and actual results of measurement predictions;
[0015] Inputting the measurement samples into a prediction model to obtain a prediction result;
[0016] According to the actual results and the predicted results, a loss function is used to calculate the loss value, and the parameters of the prediction model are updated to obtain a trained prediction model.
[0017] In some embodiments, in one embodiment of the present invention, inputting the measurement samples into a prediction model to obtain a prediction result includes:
[0018] Inputting the measurement sample into the LSTM layer to obtain a first feature;
[0019] Input the first feature into a fully connected layer to obtain a second feature; the fully connected layer includes two layers of FC, and the second layer of FC outputs the feature through a ReLU activation function;
[0020] The second feature is input into the output layer to obtain a prediction result; the output layer outputs the feature through the Elu activation function.
[0021] In some embodiments, in one embodiment of the present invention, the step of calculating a loss value using a loss function based on the actual result and the predicted result, updating the parameters of the prediction model, and obtaining a trained prediction model includes:
[0022] According to the actual results and the predicted results, the loss value is calculated using MSE; the parameters of the prediction model are updated using the Adam optimization algorithm to obtain a trained prediction model.
[0023] In some embodiments, in one embodiment of the present invention, filtering the predicted value to obtain L3-RSRP; and performing cell switching according to the L3-RSRP includes:
[0024] Perform average filtering on the second L1-RSRP of each cell to obtain L3-RSRP;
[0025] If the L3-RSRP of the current cell is less than the switching threshold, a cell switching is performed.
[0026] On the other hand, an embodiment of the present invention provides a cell handover system based on cluster RSRP measurement, including:
[0027] The first module is configured to obtain historical measurement values, wherein the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes the first L1-RSRP of all beams;
[0028] The second module is configured to input the historical measurement value into a prediction model to obtain a predicted value; the predicted value is a second L1-RSRP of all cells within a second preset time period in the future;
[0029] The third module is configured to filter the predicted value to obtain L3-RSRP; and perform cell switching according to the L3-RSRP.
[0030] On the other hand, an embodiment of the present invention provides a cell switching device based on cluster RSRP measurement, including:
[0031] at least one processor;
[0032] at least one memory for storing at least one program;
[0033] When the at least one program is executed by the at least one processor, the at least one processor is enabled to implement the above-mentioned cell switching method based on cluster RSRP measurement.
[0034] On the other hand, an embodiment of the present invention provides a storage medium storing a program executable by a processor. When executed by the processor, the program executable by the processor is used to implement the above-mentioned cell switching method based on cluster RSRP measurement.
[0035] The embodiments of the present application include at least the following beneficial effects: the embodiments of the present invention obtain historical measurement values; the historical measurement values are measurement information of all cells within the past first preset time period, and the measurement information includes the first L1-RSRP of all beams; the historical measurement values are input into a prediction model to obtain a prediction value; the prediction value is the second L1-RSRP of all cells within the future second preset time period; the prediction value is filtered to obtain L3-RSRP; cell switching is performed according to the L3-RSRP. The embodiments of the present application establish a prediction model through the measurement information of all cells within the first preset time period, so that the model is established through the spatial correlation and temporal correlation of the data, which alleviates the training and deployment costs of establishing a large number of models, is conducive to improving the prediction accuracy of the model, and thus improves the cell switching effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present invention or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.
[0037] Figure 1 A schematic flow chart of an embodiment of a cell switching method based on cluster RSRP measurement provided by the present invention;
[0038] Figure 2 A schematic diagram of an application scenario of an embodiment of a cell switching method based on cluster RSRP measurement provided by the present invention;
[0039] Figure 3 A schematic flow chart of an embodiment of a cell switching method based on cluster RSRP measurement provided by the present invention;
[0040] Figure 4 A flowchart of an embodiment of the prediction model implementation process provided by the present invention;
[0041] Figure 5 A schematic structural diagram of an embodiment of a cell switching system based on cluster RSRP measurement provided by the present invention;
[0042] Figure 6 This is a structural diagram of an embodiment of a cell switching device based on cluster RSRP measurement provided by the present invention. DETAILED DESCRIPTION
[0043] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention. The step numbers in the following embodiments are provided for ease of explanation only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.
[0044] Radio Resource Management (RRM) aims to provide high-quality service quality for terminals (UEs) under limited bandwidth conditions. Based on UE measurements (and reports), RRM allows the network to flexibly allocate and dynamically adjust radio (transmission) resources for UEs, maximizing radio spectrum utilization, preventing network congestion, and minimizing signaling overhead.
[0045] UE L3 handover primarily involves communication between network nodes and user equipment (UE) mobility management. This process ensures communication continuity and quality of service when the UE switches between different network cells. In 5G (NR) systems, RSRP measurement is performed by the UE (user equipment) and the results are used to perform cell selection, cell reselection, power control, mobility procedures, and beam management procedures. In 5G, RSRP measurements for UEs are generated and reported at the L1 and L3 layers.
[0046] Currently, the method for predicting network-triggered L3 handovers is an AI-based cell handover method. This method trains a model for each cell to predict its future RSRP, then deploys multiple trained models on the network side. While this AI-based cell handover prediction method is conceptually simple, it suffers from high deployment and management overhead for AI models. It also makes it difficult to reuse previously trained models for terminals in different locations, creating significant obstacles in practical application.
[0047] To address this issue, the present invention proposes an AI-assisted L3 handover method based on RSRP clustering time prediction. Using cluster-level RSRP measurement information, an AI-based approach is used to predict the future RSRP of multiple adjacent cells. This prediction model demonstrates excellent generalization capabilities and fully considers data characteristics in RSRP filtering. Combining the advantages of traditional filtering methods, it is robust against occasional pulsed interference, eliminating sampling deviations caused by such interference and providing reasonable guidance for L3 handover.
[0048] The cell switching method and system based on cluster RSRP measurement according to the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. First, the cell switching method based on cluster RSRP measurement according to the embodiments of the present invention will be described with reference to the accompanying drawings.
[0049] Reference Figure 1 In an embodiment of the present invention, a cell switching method based on cluster RSRP measurement is provided. The cell switching method based on cluster RSRP measurement in the embodiment of the present invention can be applied to a terminal, a server, or software running in a terminal or a server. The terminal can be a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The cell switching method based on cluster RSRP measurement in the embodiment of the present invention mainly includes the following steps:
[0050] S100: Obtain historical measurement values; the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes the first L1-RSRP of all beams;
[0051] S200: Inputting the historical measurement value into a prediction model to obtain a predicted value; the predicted value is the second L1-RSRP of all cells within a second preset time period in the future;
[0052] S300: Filter the predicted value to obtain L3-RSRP; perform cell switching according to the L3-RSRP.
[0053] In some possible implementations, the prediction model in the present application performs prediction based on spatial correlation and temporal correlation between L1-RSRP measurements. The first L1-RSRP is historical data, and the second L1-RSRP is predicted future data.
[0054] RSRP is obtained by measuring the received power of the reference signal. RSRP is then filtered to obtain L1-RSRP (various filtering methods are available, and the present invention does not impose specific limitations). L1-RSRP is then filtered again to obtain L3-RSRP. L1-RSRP is primarily filtered at the physical layer (Layer 1, L1 for short). It is useful for processes that require minimal latency, such as beam management, where the UE needs to quickly switch between beams. L3-RSRP is filtered at the radio resource control layer (RRC, Layer 3, L3 for short). It is useful for radio resource management decisions that require long-term monitoring of channel conditions, such as triggering handover procedures to reduce the risk of ping-pong handovers between serving cells. L3-RSRP results can be reported at either the beam or cell level and to the gNB in a measurement report message.
[0055] In some embodiments of the present invention, referring to Figure 2 The wireless communication scenario shown in Figure 1 has a user equipment (UE) and 7 base stations (BS). The orange dotted line represents the movement trajectory of the UE. The prediction mechanism of cell switching based on artificial intelligence is as follows: Figure 3 As shown. Cell-level measurement information from the past h measurements is used to predict the upcoming g L1-RSRP values. These measurements will be used to perform cell switching. Only one AI model needs to be trained to complete the L1-RSRP prediction for all cells. This method utilizes the spatial correlation between L1-RSRP measurements of different cells and the temporal correlation between L1-RSRP measurements of the same cell at different times, which significantly reduces the cost of model training and deployment and ensures prediction accuracy. Spatial correlation and temporal correlation mean that when a UE is in motion, its position changes over time, and the received signal strength also changes. Spatial correlation is the relationship between RSRP at different locations. Simply put, the signal is weaker when it is far away from the base station and stronger when it is close to the base station. Temporal correlation is the change in signal strength over time when the UE is not moving.
[0056] It is understood that the prediction model in this application uses the first L1-RSRP of all beams of all cells in the past h measurements to predict the second L1-RSRP of all beams of all cells in the upcoming g measurements. The AI model output (second L1-RSRP) is then filtered to derive the L3-RSRP for performing cell handover. The L1-RSRP output by the AI model can be applied to both beam and mobility management.
[0057] Optionally, in one embodiment of the present invention, obtaining historical measurement values includes:
[0058] Get RSRP measurement value;
[0059] The RSRP measurement value is filtered by a clipping average filtering method to obtain a first L1-RSRP.
[0060] Optionally, in one embodiment of the present invention, filtering the RSRP measurement value by a clipping average filtering method to obtain a first L1-RSRP includes:
[0061] If the difference between the RSRP measurement values collected twice in succession is greater than a preset deviation, discard the RSRP measurement value collected second time in the two consecutive collections;
[0062] Put all the collected RSRP measurement values into an array, and replace the oldest RSRP measurement value with the latest collected RSRP measurement value;
[0063] Performing average filtering on the values in the array to obtain a first L1-RSRP.
[0064] RSRP (Reference Signal Receiving Power) is obtained by measuring the received power of a reference signal. To eliminate the effects of rapid fading and shadow fading, the RSRP measurements are filtered to obtain L1-RSRP and L3-RSRP for model training.
[0065] ID RSRP 0001 -125 0002 -129 0003 -134 0004 -124
[0066] Table 1
[0067] The RSRP measurement is filtered using the clipping average filtering method to obtain L1-RSRP. The clipping average filtering method is equivalent to the "clipping filter method" and "recursive average filtering method". First, each new data sampled is clipped and then sent to the queue for recursive average filtering. The details are as follows:
[0068] Limiting: Refer to the measurement values shown in Table 1. For two consecutive data collections, if the difference between the first and second data collections is greater than the maximum deviation value, the second data collection is considered incorrect and not recorded. The first data collection is assumed to be the second data collection. If the difference is less than the maximum deviation value, the second data collection is considered valid and saved.
[0069] Recursive averaging: All collected data are stored in an array. Each time a new data is sampled, the latest data replaces the oldest data. Then, the arithmetic average operation is performed on the data in the array to obtain a new filtering result. The final processed data is shown in Table 2 below:
[0070]
[0071]
[0072] Table 2
[0073] Optionally, in one embodiment of the present invention, the prediction model is trained by the following steps:
[0074] Acquire measurement samples; the measurement samples include historical measurement samples and actual results of measurement predictions;
[0075] Inputting the measurement samples into a prediction model to obtain a prediction result;
[0076] According to the actual results and the predicted results, a loss function is used to calculate the loss value, and the parameters of the prediction model are updated to obtain a trained prediction model.
[0077] Optionally, in one embodiment of the present invention, inputting the measurement samples into a prediction model to obtain a prediction result includes:
[0078] Inputting the measurement sample into the LSTM layer to obtain a first feature;
[0079] Input the first feature into a fully connected layer to obtain a second feature; the fully connected layer includes two layers of FC, and the second layer of FC outputs the feature through a ReLU activation function;
[0080] The second feature is input into the output layer to obtain a prediction result; the output layer outputs the feature through the Elu activation function.
[0081] Optionally, in one embodiment of the present invention, the calculating a loss value using a loss function based on the actual result and the predicted result, updating the parameters of the prediction model, and obtaining a trained prediction model includes:
[0082] According to the actual results and the predicted results, the loss value is calculated using MSE; the parameters of the prediction model are updated using the Adam optimization algorithm to obtain a trained prediction model.
[0083] The prediction model in the embodiments of the present application can be any artificial intelligence model. For example, it can be an artificial intelligence model based on an LSTM network combined with a fully connected layer, which can predict future L1-RSRP based on historical L1-RSRP. Figure 4 The overall framework structure of the LSTM-FC model is given, which consists of three parts:
[0084] The first part consists of two LSTM layers and Dropout. Its input needs to be converted into an h*M*N matrix, where h is the number of data items, M is the number of features, and N is the number of time steps, that is, how long the previous data is used as input for prediction. The two LSTM layers contain 512 and 64 nodes, respectively.
[0085] The second part includes two fully connected layers and Dropout. The number of nodes in the two fully connected layers is 128 and 64 respectively. A layer of Dropout is set after the two layers to prevent overfitting. The Dropout rate is set to 0.25, and the ReLU activation function is used in the second fully connected layer.
[0086] The third part is the output layer. Because the final output of the regression task is a single value, the final layer is a fully connected layer with only one node, and this fully connected layer uses Elu as the activation function. The model uses MSE as the loss function and the Adam optimization algorithm as the optimizer for weight updates.
[0087] The model outputs the L1-RSRP of all beams in all cells from the g measurements (i.e., the second L1-RSRP). The average L1-RSRP of each cell is then selected as the L3-RSRP. Network operators set an appropriate handover threshold based on the specific network environment and service requirements. Handover occurs when the L3-RSRP meets the threshold.
[0088] Optionally, in one embodiment of the present invention, filtering the predicted value to obtain L3-RSRP; and performing cell switching according to the L3-RSRP includes:
[0089] Perform average filtering on the second L1-RSRP of each cell to obtain L3-RSRP;
[0090] If the L3-RSRP of the current cell is less than the switching threshold, a cell switching is performed.
[0091] The embodiment of the present application uses training and prediction of existing RSRP values to predict RSRP changes in advance and make handover deployments in advance, which is more scientific and effective than switching based on real-time measurement data. The embodiment of the present application uses a limited average filtering method to filter RSRP measurement values to obtain L1-RSRP. It fully considers data characteristics and combines the advantages of traditional filtering methods to resist occasional pulse interference and eliminate sampling value deviations caused by pulse interference. The embodiment of the present application uses an LSTM plus FC model to train this type of time series data, which can well capture the long-term dependencies in time series data and improve prediction accuracy. In the embodiment of the present application, the input and output of the prediction model are both L1-RSRP values. L3-RSRP is obtained by filtering the prediction results again, rather than directly filtering twice before inputting the model to obtain L3-RSRP and then inputting the model to directly obtain L3-RSRP. Continuous filtering twice may cause the data to be overly smoothed and the prediction error to be large. The embodiment of the present application measures the moving UE, and the trained model has good generalization ability and can adapt to situations where the UE is at different distances from the base station.
[0092] This paper proposes an AI-based measurement prediction method for L3 mobility based on cell measurement information. The method includes AI model input, output, data collection and processing, and training methods. This method enables accurate prediction of L3-RSRP, significantly reducing model training and deployment costs and improving the model's generalization capabilities.
[0093] In summary, the above method obtains historical measurement values; the historical measurement values are measurement information of all cells within the past first preset time period, and the measurement information includes the first L1-RSRP of all beams; the historical measurement values are input into the prediction model to obtain prediction values; the prediction values are the second L1-RSRP of all cells within the future second preset time period; the prediction values are filtered to obtain L3-RSRP; and cell switching is performed according to the L3-RSRP. The embodiment of the present application establishes a prediction model through the measurement information of all cells within the first preset time period, so that the model is established through the spatial correlation and temporal correlation of the data, which alleviates the training and deployment costs of establishing a large number of models, is conducive to improving the prediction accuracy of the model, and thus improves the cell switching effect.
[0094] Secondly, refer to the attached Figure 5 A cell handover system based on cluster RSRP measurement proposed according to an embodiment of the present invention is described.
[0095] Figure 5 1 is a schematic diagram of a cell switching system structure based on cluster RSRP measurement according to an embodiment of the present invention. The system specifically includes:
[0096] The first module 510 is configured to obtain historical measurement values, wherein the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes first L1-RSRPs of all beams;
[0097] The second module 520 is configured to input the historical measurement value into a prediction model to obtain a prediction value; the prediction value is a second L1-RSRP of all cells within a second preset time period in the future;
[0098] The third module 530 is configured to filter the predicted value to obtain L3-RSRP; and perform cell switching according to the L3-RSRP.
[0099] It can be seen that the contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.
[0100] Reference Figure 6 An embodiment of the present invention provides a cell switching device based on cluster RSRP measurement, comprising:
[0101] at least one processor 610;
[0102] at least one memory 620, for storing at least one program;
[0103] When the at least one program is executed by the at least one processor 610, the at least one processor 610 implements the cell switching method based on cluster RSRP measurement.
[0104] Similarly, the contents of the above method embodiments are applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.
[0105] An embodiment of the present invention further provides a computer-readable storage medium storing a program executable by a processor. When the program is executed by the processor, it is used to perform the above-mentioned cell switching method based on cluster RSRP measurement.
[0106] Similarly, the contents of the above method embodiments are applicable to the present storage medium embodiment. The functions specifically implemented by the present storage medium embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.
[0107] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operation and logic flow presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.
[0108] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the ordinary skill of an engineer. Therefore, a person skilled in the art will be able to implement the present invention set forth in the claims using ordinary skill without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.
[0109] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several programs for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
[0110] The logic and / or steps represented in a flowchart or otherwise described herein, for example, may be considered as an ordered list of executable programs for implementing the logical functions, and may be embodied in any computer-readable medium for use by, or in conjunction with, a program execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can retrieve and execute a program from a program execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, a program execution system, apparatus, or device.
[0111] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.
[0112] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable program execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0113] In the above description of this specification, reference to the terms "one embodiment / example," "another embodiment / example," or "certain embodiments / examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
[0114] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.
[0115] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.
Claims
1. A cell switching method based on cluster RSRP measurement, characterized in that: The following steps are involved: Obtaining historical measurement values; the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes the first L1-RSRP of all beams; Inputting the historical measurement value into a prediction model to obtain a prediction value; the prediction value is the second L1-RSRP of all cells within a second preset time period in the future; The predicted value is filtered to obtain L3-RSRP; and cell switching is performed according to the L3-RSRP.
2. The cell switching method based on cluster RSRP measurement according to claim 1, characterized in that: The obtaining of historical measurement values comprises: Get RSRP measurement value; The RSRP measurement value is filtered by a limiting average filtering method to obtain a first L1-RSRP.
3. The cell switching method based on cluster RSRP measurement according to claim 2, characterized in that: The filtering the RSRP measurement value by a limiting average filtering method to obtain a first L1-RSRP includes: If the difference between the RSRP measurement values collected twice in succession is greater than a preset deviation, discard the RSRP measurement value collected second time in the two consecutive collections; Put all the collected RSRP measurement values into an array, and replace the oldest RSRP measurement value with the latest collected RSRP measurement value; An average filtering process is performed on the values in the array to obtain a first L1-RSRP.
4. The cell switching method based on cluster RSRP measurement according to claim 1, characterized in that: The prediction model is trained by the following steps: Acquire measurement samples; the measurement samples include historical measurement samples and actual results of measurement predictions; Inputting the measurement samples into a prediction model to obtain a prediction result; According to the actual result and the predicted result, a loss function is used to calculate the loss value, and the parameters of the prediction model are updated to obtain a trained prediction model.
5. The cell switching method based on cluster RSRP measurement according to claim 4, characterized in that: The step of inputting the measurement samples into a prediction model to obtain a prediction result includes: Inputting the measurement sample into the LSTM layer to obtain a first feature; Input the first feature into a fully connected layer to obtain a second feature; the fully connected layer includes two layers of FC, and the second layer of FC outputs the feature through a ReLU activation function; The second feature is input into the output layer to obtain a prediction result; the output layer outputs the feature through the Elu activation function.
6. The cell switching method based on cluster RSRP measurement according to claim 4, characterized in that: The method of calculating the loss value by using a loss function according to the actual result and the predicted result, updating the parameters of the prediction model, and obtaining a trained prediction model comprises: According to the actual results and the predicted results, the loss value is calculated using MSE; the parameters of the prediction model are updated using the Adam optimization algorithm to obtain a trained prediction model.
7. The cell switching method based on cluster RSRP measurement according to claim 1, characterized in that: The filtering the predicted value to obtain L3-RSRP; and performing cell switching according to the L3-RSRP, including: Perform average filtering on the second L1-RSRP of each cell to obtain L3-RSRP; If the L3-RSRP of the current cell is less than the switching threshold, a cell switching is performed.
8. A cell switching system based on cluster RSRP measurement, characterized in that: include: The first module is used to obtain historical measurement values; the historical measurement values are measurement information of all cells within a first preset time period in the past, and the measurement information includes the first L1-RSRP of all beams; The second module is used to input the historical measurement value into the prediction model to obtain the prediction value; The predicted value is the second L1-RSRP of all cells within a second preset time period in the future; The third module is used to filter the predicted value to obtain L3-RSRP; and perform cell switching according to the L3-RSRP.
9. A cell switching device based on cluster RSRP measurement, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the cell switching method based on cluster RSRP measurement as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The program executable by the processor is used to implement the cell switching method based on cluster RSRP measurement as claimed in any one of claims 1 to 7 when executed by the processor.