Base station cell load intelligent prediction method, device, equipment, medium and product
By conducting prediction model analysis on the information and network operation data of the base station cell, the high-load cell is automatically identified and optimization solutions are generated, and the problem of low efficiency of identifying high-load cells in the prior art is solved, and the capacity expansion efficiency and resource utilization are improved.
Patent Information
- Application Number
- CN202410803486.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-20
- Publication Date
- 2024-11-05
AI Technical Summary
In the prior art, it is impossible to conveniently determine whether the base station cell is a high-load cell, resulting in high investment costs and low efficiency in capacity expansion.
By obtaining cell information and inputting it into the trained prediction model, the cell load level is automatically determined, and the cause of the load level is determined based on the network operation data to generate a targeted optimization plan.
Accurate identification and cause analysis of high-load cells are achieved, capacity expansion efficiency is improved, resource waste and investment costs are reduced.
Smart Images

Figure CN118900416A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, equipment, medium and product for intelligent prediction of base station cell load. Background Art
[0002] Network capacity directly affects user perception, and the cost of capacity expansion investment is limited. Therefore, how to accurately identify high-load cells that really need capacity expansion is the key to reducing capacity expansion investment costs and improving capacity expansion efficiency. If high-load cells can be accurately identified, the corresponding optimized capacity expansion plan can be output in a timely manner to improve resource utilization.
[0003] However, the actual conditions of different cells are different. In the prior art, different cells need to be manually determined one by one based on the actual conditions of the cells to determine whether they are high-load cells. There is no convenient method to determine whether a cell is a high-load cell, which is inefficient. Summary of the invention
[0004] The present invention provides a base station cell load intelligent prediction method, device, equipment, medium and product to solve the defect in the prior art that there is no convenient method for determining whether a cell is a high-load cell, thereby improving the efficiency of determining whether a cell is a high-load cell.
[0005] The present invention provides a base station cell load intelligent prediction method, comprising: Acquire cell information of a cell to be predicted, where the cell information reflects network attributes of the cell to be predicted and load conditions of the cell to be predicted in a historical period; Inputting the cell information into a trained first prediction model, and obtaining a load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of the first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0006] According to a base station cell load intelligent prediction method provided by the present invention, after obtaining the load level of the cell to be predicted output by the first prediction model, the method further includes: When the load level of the cell to be predicted reaches a preset threshold, network operation data of the cell to be predicted is obtained, the network operation data is input into the trained second prediction model, and the cause of the load level of the cell to be predicted outputted by the second prediction model is obtained, where the cause of the load level reflects the reason why the load level of the cell to be predicted in the future period reaches the preset threshold; The second prediction model is trained based on multiple groups of second training data, and each group of the second training data includes sample network operation data and load level cause labels corresponding to the sample network operation data.
[0007] According to a base station cell load intelligent prediction method provided by the present invention, after obtaining the cause of the load level of the cell to be predicted output by the second prediction model, the method includes: When the load level of the cell to be predicted is caused by a non-fault reason, the network configuration parameters of the cell to be predicted are obtained; and a load optimization plan for the cell to be predicted is generated based on the network configuration parameters of the cell to be predicted.
[0008] According to a method for intelligent prediction of base station cell load provided by the present invention, the network configuration parameters of the cell to be predicted include cell merging status, service access data, channel quality indicator value and surrounding cell load level; the load optimization scheme for the cell to be predicted is generated based on the network configuration parameters of the cell to be predicted, including: When there is a cell merger in the to-be-predicted cell, the load optimization scheme includes canceling the cell merger; When there is no cell merger in the cell to be predicted, determining whether the service access data of the cell to be predicted corresponds to unreasonable base station coverage of the cell to be predicted, or whether the channel quality indicator value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted, the load optimization scheme includes adjusting the base station coverage of the cell to be predicted or adjusting the interference data of the cell to be predicted; When the service access data of the cell to be predicted corresponds to the reasonable base station coverage of the cell to be predicted, and the channel quality indicator value of the cell to be predicted corresponds to the reasonable interference data of the cell to be predicted, if the load level of the cells surrounding the cell to be predicted is lower than the preset level, the load optimization plan includes adjusting the base station coverage of the cell to be predicted and the surrounding cells; if the load level of the cells surrounding the cell to be predicted is not lower than the preset level, the load optimization plan includes adjusting the carrier equalization parameters of the cell to be predicted or expanding the network capacity of the cell to be predicted.
[0009] According to a method for intelligent prediction of base station cell load provided by the present invention, the determining whether the service access data of the cell to be predicted corresponds to unreasonable base station coverage of the cell to be predicted, or whether the channel quality indicator value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted includes: When the access distance average in the service access data of the cell to be predicted is greater than the first preset average or the long-distance access ratio is greater than the first preset ratio, it is determined that the service access data of the cell to be predicted corresponds to the base station coverage of the cell to be predicted is unreasonable; When the mean of the channel quality indication values of the cell to be predicted is less than a second preset mean or the proportion of low channel quality indication values is greater than a second preset proportion, it is determined that the interference data of the cell to be predicted corresponding to the channel quality indication value of the cell to be predicted is unreasonable.
[0010] According to a base station cell load intelligent prediction method provided by the present invention, the cell information includes network attribute parameters of the cell to be predicted, load index values of the cell to be predicted, and perception index values of the cell to be predicted.
[0011] The present invention also provides a base station cell load intelligent prediction device, comprising: An information acquisition module, used to acquire cell information of a cell to be predicted, wherein the cell information reflects the network attributes of the cell to be predicted and the load of the cell to be predicted in a historical period; A model prediction module, used for inputting the cell information into a trained first prediction model, and obtaining the load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of the first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0012] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for intelligently predicting the load of a base station cell as described above is implemented.
[0013] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for intelligently predicting the load of a base station cell as described above is implemented.
[0014] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned methods for intelligent prediction of base station cell load.
[0015] The intelligent prediction method, device, equipment, medium and product for base station cell load provided by the present invention obtain cell information reflecting the network attributes of the cell to be predicted and the load conditions of the cell to be predicted in a historical period, and input the information into a trained first prediction model. The first prediction model is trained based on multiple groups of first training data including sample cell information and load level labels corresponding to the sample cell information. In this way, the first prediction model can learn the intrinsic relationship between the cell information and the cell load level, so that after the cell information of the cell to be predicted is input, the load level of the cell to be predicted can be output, thereby realizing automatic determination of the load level of the cell to be predicted and improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0017] Figure 1 It is a flow chart of a method for intelligent prediction of base station cell load provided by the present invention; Figure 2 It is a schematic diagram of a training set calling process of a prediction model in a base station cell load intelligent prediction method provided by the present invention; Figure 3 It is a schematic diagram of the data processing process of the second prediction model in the intelligent prediction method for base station cell load provided by the present invention; Figure 4 It is a logic block diagram of generating a load optimization scheme in a base station cell load intelligent prediction method provided by the present invention; Figure 5 It is a structural schematic diagram of the base station cell load intelligent prediction device provided by the present invention; Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0018] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] Combine the following Figure 1-4 The intelligent prediction method for base station cell load provided by the present invention is described as follows: Figure 1 As shown, the method comprises the steps of: S110, obtaining cell information of the cell to be predicted, where the cell information reflects the network attributes of the cell to be predicted and the load of the cell to be predicted in a historical period; S120, inputting the cell information into the trained first prediction model, and obtaining the load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0020] How to accurately identify the high-load cells that really need to be expanded is the key to reducing the investment cost of expansion and improving the efficiency of expansion. If the high-load cells can be accurately identified, the corresponding optimized expansion plan can be output in time to improve resource utilization. However, in practical applications, the actual conditions of different cells (including network attributes and load conditions) are different, and it is impossible to use a unified standard to determine whether the cell is a high-load cell that needs to be considered for expansion, which will result in the need for manual one-by-one judgment, which is inefficient. The method provided by the present invention obtains cell information reflecting the network attributes of the cell to be predicted and the load conditions of the cell to be predicted in the historical period, and inputs it into the trained first prediction model, which is based on a plurality of groups of first training data including sample cell information and load level labels corresponding to the sample cell information. In this way, the first prediction model can learn the intrinsic relationship between the cell information and the load level of the cell, so that after inputting the cell information of the cell to be predicted, the load level of the cell to be predicted can be output, and the automatic judgment of the load level of the cell to be predicted is realized, thereby improving efficiency.
[0021] The cell information includes the network attribute parameters of the cell to be predicted, the load type index value of the cell to be predicted, and the perception type index value of the cell to be predicted. Specifically, the network attribute parameters reflect the network attributes of the cell to be predicted, such as the standard, bandwidth, frequency band, etc. of the cell to be predicted. The network attribute parameters of the cell to be predicted can be matched from the network attribute data of each cell based on the cell name and the ECGI (Extended Cell Global Identity) of the cell. The load type index value of the cell to be predicted reflects the load data of the cell to be predicted. The load type index of the cell to be predicted may include the number of users, traffic, utilization rate, etc. of the cell to be predicted. The perception type index of the cell to be predicted reflects the status of the factors affecting the network perception of the cell to be predicted. The perception type index of the cell to be predicted may include call connection status, disconnection status, switching status, perception rate, etc.
[0022] It is worth noting that the load data of the cell to be predicted cannot directly reflect whether the load level of the cell to be predicted is high or low, because the network attribute parameters and perception index values that affect the load level of the cell to be predicted are different. The load type index value of the cell to be predicted is high, but the perception index value of the cell to be predicted is also high, and the load level of the cell to be predicted cannot be judged to be high.
[0023] The network attribute parameters, perception index values and load type indicators of the cell influence and restrict each other. Different cells have different network attribute parameters and perception index values. It is obviously not accurate to judge whether the cell is high-loaded only based on the state of the load factors that can affect the predicted cell without considering the network attribute parameters, perception index values, etc. Therefore, in the prior art, it is impossible to formulate a unified standard to determine whether the load level of the cell is high-loaded. It is necessary to judge one by one according to the actual situation of the cell, which is very inefficient.
[0024] In the method provided by the present invention, the intrinsic relationship between the cell information and the load level of the cell is learned through a deep learning model. Specifically, a first prediction model is constructed, and the first prediction model is a deep learning model. After being trained with multiple sets of first training data, the first prediction model learns the relationship between the cell information and the load level of the cell, so that after the cell information of the cell to be predicted is input into the trained first prediction model, the first prediction model can output the load level of the cell to be predicted.
[0025] In one possible implementation, the load level of the cell to be predicted can be set to three levels: high level, medium level, and low level. More or fewer levels can also be set, such as two levels, four levels, etc. When the first prediction model outputs that the load level of the cell to be predicted is a high level, it is determined that the load level of the pre-stored cell reaches a preset threshold.
[0026] The load level label in the first training data can be realized by manual labeling, that is, reading the cell information of the existing cell whose load level is manually determined as the sample cell information, and taking the load level of the cell determined manually as the load level label corresponding to the sample cell information, thereby obtaining a set of first training data. Figure 2 As shown, after the first prediction model is trained based on multiple groups of first training data, the first prediction model can be deployed in a terminal device that executes the method provided by the present invention or in a terminal device that is communicatively connected to the terminal device that executes the method provided by the present invention, so that the terminal device that executes the method provided by the present invention can call the first prediction model through a model calling interface.
[0027] The method provided by the present invention can integrate various types of data and combine artificial intelligence algorithms through the above content to more accurately realize the judgment of high-load cells, avoid the disadvantages of excessive influence of factors considered in traditional capacity optimization processes, low efficiency, and high investment in optimization personnel and resources, realize accurate investment of resources, and reduce the cost of cell capacity optimization.
[0028] After identifying that the cell to be predicted is a high-load cell with a load level higher than a preset threshold, the method provided by the present invention further determines the reason why the cell to be predicted is a high-load cell. By determining the cause of the cell to be predicted being a high-load cell, effective reference information can be provided for improving the load of the high-load cell. After obtaining the load level of the cell to be predicted in the future period output by the first prediction model, the method provided by the present invention further includes the steps of: When the load level of the cell to be predicted in the future period reaches a preset threshold, network operation data of the cell to be predicted is obtained, the network operation data is input into the trained second prediction model, and the load level cause of the cell to be predicted output by the second prediction model is obtained, where the load level cause reflects the reason why the load level of the cell to be predicted reaches the preset threshold; The second prediction model is obtained by training based on multiple groups of second training data, and each group of second training data includes sample network operation data and load level cause labels corresponding to the sample network operation data.
[0029] The network operation data of the cell to be predicted reflects the network data that records the network operation status generated during the network operation of the cell to be predicted. The network operation data may include log data, performance data, alarm data, signaling data, network element basic data, etc. The network operation data contains multi-dimensional data, which can provide multi-faceted information for the prediction of the load level cause of the cell to be predicted, thereby facilitating the second prediction model to locate the load level cause of the cell to be predicted.
[0030] The method provided by the present invention obtains the network operation data of the cell to be predicted and inputs it into a trained second prediction model. The second prediction model is trained based on multiple groups of second training data including sample network operation data and load level cause labels corresponding to the sample network operation data. In this way, the second prediction model can learn the intrinsic relationship between the network operation data and the load level cause of the cell. After the network operation data of the cell to be predicted is input, the load level cause of the cell to be predicted can be output, thereby realizing automatic positioning of the load level cause of the cell to be predicted and further improving efficiency.
[0031] like Figure 3As shown, the second prediction model can be a deep learning model for classification. After inputting network operation data including log data, performance data, alarm data, signaling data, network element basic data, etc., features are extracted through the network elements of the hidden layer, and finally the features are mapped to different categories through the softmax function of the output layer. In one possible implementation, three load level cause categories can be pre-set: failure, interference, and high traffic volume. Of course, it is understandable that more or fewer load level cause categories can also be set.
[0032] The load level cause label in the second training data can be realized by manual labeling, which can be to read the network operation data of the cell whose load level cause is manually determined as the sample network operation data, and use the load level cause of the cell determined manually as the load level cause label corresponding to the sample network operation data, so as to obtain a set of second training data, and the load level cause label corresponds to one of the set load level cause categories. Figure 2 As shown, after the second prediction model is trained based on multiple sets of second training data, the second prediction model can be deployed in a terminal device that executes the method provided by the present invention or in a terminal device that is communicatively connected to the terminal device that executes the method provided by the present invention, so that the terminal device that executes the method provided by the present invention can call the second prediction model through a model calling interface.
[0033] By integrating various types of data and combining them with artificial intelligence algorithms, we can more accurately and conveniently determine the causes of high loads, further improve the efficiency of cell capacity optimization, achieve accurate resource investment, and reduce optimization costs.
[0034] Furthermore, in a possible implementation, the method provided by the present invention, after obtaining the cause of the load level of the cell to be predicted output by the second prediction model, includes: When the load level of the cell to be predicted is caused by a non-fault reason, obtaining network configuration parameters of the cell to be predicted; A load optimization plan for the cell to be predicted is generated based on the network configuration parameters of the cell to be predicted.
[0035] After determining the cause of the load level of the cell to be predicted, a corresponding load optimization plan is generated based on the cause of the load level of the cell to be predicted, which can further improve the efficiency of cell capacity optimization.
[0036] The network configuration parameters of the cell to be predicted include cell merging status, service access data, channel quality indicator value and load level of surrounding cells; based on the network configuration parameters of the cell to be predicted, a load optimization scheme for the cell to be predicted is generated, including: When there is a cell merger in the cell to be predicted, the load optimization solution includes canceling the cell merger; When there is no cell merger in the cell to be predicted, it is determined whether the service access data of the cell to be predicted corresponds to unreasonable coverage of the base station of the cell to be predicted, or whether the channel quality indicator value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted, and the load optimization plan includes adjusting the base station coverage of the cell to be predicted or adjusting the interference data of the cell to be predicted; When the service access data of the cell to be predicted corresponds to the base station coverage of the cell to be predicted reasonably, and the channel quality indicator value of the cell to be predicted corresponds to the interference data of the cell to be predicted reasonably, if the load level of the cells surrounding the cell to be predicted is lower than the preset level, the load optimization plan includes adjusting the base station coverage of the cell to be predicted and the weekly standard cells; if the load level of the cells surrounding the cell to be predicted is not lower than the preset level, the load optimization plan includes adjusting the carrier equalization parameters of the cell to be predicted or expanding the network capacity of the cell to be predicted.
[0037] In the method provided by the present invention, when the cell to be predicted is a high-load cell and the cause of the load level of the cell to be predicted is not due to a fault, it is not directly determined that the cell to be predicted needs to be expanded, because in addition to the fault, there may be many factors that cause the cell to be predicted to be a high-load cell. Figure 4 As shown, in the method provided by the present invention, when the cell to be predicted is a high-load cell and the cause of the load level of the cell to be predicted is not due to a fault, the cell to be predicted is included in the scope of capacity expansion consideration, and the load optimization plan for the cell to be predicted is determined in combination with the cell merging situation, service access data, channel quality indicator value and load level of the surrounding cells of the cell to be predicted. In this way, a more accurate load optimization plan can be provided, the reasonable allocation of capacity resources is achieved, and the problem of waste of capacity expansion resources that may be caused by directly expanding the capacity of the cell to be predicted is reduced.
[0038] When there is a cell merger in the cell to be predicted, that is, when the cell to be predicted is actually a merger of two cells that can be independent, the cell merger can be released for the cell to be predicted, and the cells contained therein can be re-evaluated as independent cells to determine whether the independent cells are high-load cells.
[0039] The service access data of the cell to be predicted includes the geographical location of the base station accessed by the network connection generated in the cell to be predicted; determining whether the service access data of the cell to be predicted corresponds to unreasonable coverage of the base station of the cell to be predicted includes: When the access distance average in the service access data of the cell to be predicted is greater than the first preset average or the long-distance access ratio is greater than the first preset ratio, it is determined that the base station coverage of the cell to be predicted corresponding to the service access data of the cell to be predicted is unreasonable.
[0040] The mean access distance in the service access data of the cell to be predicted refers to the mean distance between the accessed base station and the cell to be predicted in the service access data of the cell to be predicted. Long-distance access refers to the number of times that the distance between the accessed base station and the cell to be predicted in the service access data of the cell to be predicted is greater than a certain value. If the mean access distance is greater than the first preset mean or the proportion of long-distance access is greater than the first preset proportion, it means that most of the network connections of the cell to be predicted are connected to base stations at a longer distance, which means that the base station coverage setting of the cell to be predicted is unreasonable. At this time, the output load optimization plan should include adjusting the base station coverage setting of the cell to be predicted.
[0041] Determining whether the channel quality indication value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted, including: when the mean of the channel quality indication values of the cell to be predicted is less than a second preset mean or the proportion of low channel quality indication values is greater than a second preset proportion, determining that the channel quality indication value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted.
[0042] The channel quality indicator value (CQI) of the cell to be predicted reflects the channel quality of the cell to be predicted. When the mean value of the channel quality indicator value of the cell to be predicted is less than a second preset mean value or the channel quality indicator value is less than a certain setting and the proportion of low channel quality indicator values thereof is greater than a second preset proportion, it indicates that the high load of the cell to be predicted may be caused by serious interference in the cell to be predicted. At this time, the output load optimization plan should include adjusting the interference data of the cell to be predicted.
[0043] When the service access data of the cell to be predicted corresponds to the base station coverage of the cell to be predicted reasonably, and the channel quality indicator value of the cell to be predicted corresponds to the interference data of the cell to be predicted reasonably, a load optimization plan for the cell to be predicted is generated in combination with the load levels of the surrounding cells of the cell to be predicted. Figure 4 As shown, if the load level of the surrounding cells of the cell to be predicted is low, the load of the cell to be predicted can be adjusted, which can be achieved by balancing the base station coverage of the cell to be predicted and the surrounding cells with lower load levels.
[0044] If the surrounding cells of the predicted cell are also high-load cells, then determine whether the predicted cell is a single carrier. If it is a dual carrier, first determine whether both carriers meet the preset expansion threshold. If only one carrier meets the threshold and the other carrier load does not meet the threshold, then adjust the carrier balancing parameters. If both carriers meet the preset expansion threshold or the predicted cell is a single carrier, then generate a load optimization plan for expanding the predicted cell. Expanding the predicted cell may include expanding carriers, adding micro stations, and building additional base stations.
[0045] The following is a description of the base station cell load intelligent prediction device provided by the present invention. The base station cell load intelligent prediction device described below and the base station cell load intelligent prediction method described above can be referred to each other. Figure 5 As shown, the base station cell load intelligent prediction device provided by the present invention includes: The information acquisition module 510 is used to acquire the cell information of the cell to be predicted, where the cell information reflects the network attributes of the cell to be predicted and the load of the cell to be predicted in the historical period; The model prediction module 520 is used to input the cell information into the trained first prediction model to obtain the load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0046] Furthermore, the model prediction module 520 can also be used to obtain network operation data of the cell to be predicted when the load level of the cell to be predicted reaches a preset threshold, input the network operation data into the trained second prediction model, and obtain the cause of the load level of the cell to be predicted output by the second prediction model, where the load level cause reflects the reason why the load level of the cell to be predicted in the future time period reaches the preset threshold; wherein the second prediction model is trained based on multiple groups of second training data, and each group of second training data includes sample network operation data and load level cause labels corresponding to the sample network operation data.
[0047] The base station cell load intelligent prediction device provided by the present invention may also include a load optimization module, which is used to obtain network configuration parameters of the cell to be predicted when the load level of the cell to be predicted is caused by non-fault reasons; and generate a load optimization plan for the cell to be predicted based on the network configuration parameters of the cell to be predicted.
[0048] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor (processor) 610, a communication interface (Communications Interface) 620, a memory (memory) 630 and a communication bus 640, wherein the processor 610, the communication interface 620, and the memory 630 complete mutual communication through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute the intelligent prediction method for the load of the base station cell, the method comprising: obtaining the cell information of the cell to be predicted, the cell information reflects the network attributes of the cell to be predicted and the load of the cell to be predicted in the historical period; inputting the cell information into the trained first prediction model, and obtaining the load level of the cell to be predicted output by the first prediction model; wherein the first prediction model is obtained by training based on multiple groups of first training data, and each group of first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0049] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, 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, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a 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: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0050] On the other hand, the present invention also provides a computer program product, which includes a computer program, and the computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the base station cell load intelligent prediction method provided by the above methods, and the method includes: obtaining cell information of the cell to be predicted, the cell information reflects the network attributes of the cell to be predicted and the load conditions of the cell to be predicted in a historical period; inputting the cell information into a trained first prediction model to obtain the load level of the cell to be predicted output by the first prediction model; wherein the first prediction model is trained based on multiple groups of first training data, and each group of first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0051] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the intelligent prediction method for base station cell load provided by the above-mentioned methods, the method comprising: obtaining cell information of the cell to be predicted, the cell information reflecting the network attributes of the cell to be predicted and the load condition of the cell to be predicted in a historical period; inputting the cell information into a trained first prediction model to obtain the load level of the cell to be predicted output by the first prediction model; wherein the first prediction model is trained based on multiple groups of first training data, and each group of first training data includes sample cell information and a load level label corresponding to the sample cell information.
[0052] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0053] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for intelligent prediction of base station cell load, characterized in that: include: Acquire cell information of a cell to be predicted, where the cell information reflects network attributes of the cell to be predicted and load conditions of the cell to be predicted in a historical period; Inputting the cell information into a trained first prediction model, and obtaining a load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of the first training data includes sample cell information and a load level label corresponding to the sample cell information.
2. The intelligent prediction method for base station cell load according to claim 1, characterized in that: After obtaining the load level of the cell to be predicted output by the first prediction model, the method further includes: When the load level of the cell to be predicted reaches a preset threshold, network operation data of the cell to be predicted is obtained, the network operation data is input into the trained second prediction model, and the cause of the load level of the cell to be predicted outputted by the second prediction model is obtained, where the cause of the load level reflects the reason why the load level of the cell to be predicted in the future period reaches the preset threshold; The second prediction model is trained based on multiple groups of second training data, and each group of the second training data includes sample network operation data and load level cause labels corresponding to the sample network operation data.
3. The intelligent prediction method for base station cell load according to claim 2, characterized in that: After obtaining the cause of the load level of the cell to be predicted output by the second prediction model, the method includes: When the load level of the cell to be predicted is caused by a non-fault reason, the network configuration parameters of the cell to be predicted are obtained; and a load optimization plan for the cell to be predicted is generated based on the network configuration parameters of the cell to be predicted.
4. The intelligent prediction method for base station cell load according to claim 3, characterized in that: The network configuration parameters of the cell to be predicted include cell merging status, service access data, channel quality indicator value, and load level of surrounding cells; the load optimization scheme for the cell to be predicted is generated based on the network configuration parameters of the cell to be predicted, including: When there is a cell merger in the to-be-predicted cell, the load optimization scheme includes canceling the cell merger; When there is no cell merger in the cell to be predicted, determining whether the service access data of the cell to be predicted corresponds to unreasonable base station coverage of the cell to be predicted, or whether the channel quality indicator value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted, the load optimization scheme includes adjusting the base station coverage of the cell to be predicted or adjusting the interference data of the cell to be predicted; When the service access data of the cell to be predicted corresponds to the reasonable base station coverage of the cell to be predicted, and the channel quality indicator value of the cell to be predicted corresponds to the reasonable interference data of the cell to be predicted, if the load level of the cells surrounding the cell to be predicted is lower than the preset level, the load optimization plan includes adjusting the base station coverage of the cell to be predicted and the surrounding cells; if the load level of the cells surrounding the cell to be predicted is not lower than the preset level, the load optimization plan includes adjusting the carrier equalization parameters of the cell to be predicted or expanding the network capacity of the cell to be predicted.
5. The intelligent prediction method for base station cell load according to claim 4, characterized in that: The determining whether the service access data of the cell to be predicted corresponds to unreasonable base station coverage of the cell to be predicted, or whether the channel quality indicator value of the cell to be predicted corresponds to unreasonable interference data of the cell to be predicted, includes: When the access distance average in the service access data of the cell to be predicted is greater than the first preset average or the long-distance access ratio is greater than the first preset ratio, it is determined that the service access data of the cell to be predicted corresponds to the base station coverage of the cell to be predicted is unreasonable; When the mean of the channel quality indication values of the cell to be predicted is less than a second preset mean or the proportion of low channel quality indication values is greater than a second preset proportion, it is determined that the interference data of the cell to be predicted corresponding to the channel quality indication value of the cell to be predicted is unreasonable.
6. The intelligent prediction method for base station cell load according to claim 1, characterized in that: The cell information includes network attribute parameters of the cell to be predicted, load index values of the cell to be predicted, and perception index values of the cell to be predicted.
7. A base station cell load intelligent prediction device, characterized in that: include: An information acquisition module, used to acquire cell information of a cell to be predicted, wherein the cell information reflects the network attributes of the cell to be predicted and the load of the cell to be predicted in a historical period; A model prediction module, used for inputting the cell information into a trained first prediction model, and obtaining the load level of the cell to be predicted output by the first prediction model; The first prediction model is obtained through training based on multiple groups of first training data, and each group of the first training data includes sample cell information and a load level label corresponding to the sample cell information.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the base station cell load intelligent prediction method as described in any one of claims 1 to 6 is implemented.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for intelligent prediction of base station cell load according to any one of claims 1 to 6 is implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for intelligent prediction of base station cell load according to any one of claims 1 to 6 is implemented.
