Cell service migration method, device and computer-readable storage medium

Abstract

The present invention discloses a method, device and computer-readable storage medium for the migration of services in a cell, wherein the method comprises the following steps: clustering the cells according to the measurement report data of each cell to obtain at least one co-coverage cluster; obtaining the capacity layer cells in the co-coverage cluster; determining the target neighboring area of ​​the capacity layer cell according to the predicted service volume of the capacity layer cell and the neighboring area in the current time period, wherein the predicted service volume is predicted by the historical measurement report data of the capacity layer cell and the neighboring area; and migrating the services of the capacity layer cell to the target neighboring area corresponding to the capacity layer cell. The present invention can avoid a large service load on other cells that undertake the services when migrating the services of the cell to other cells.

Description

Technical Field

[0001] The present invention relates to the field of mobile communication wireless technology, and in particular to a method and device for migrating services in a cell, and a computer-readable storage medium. Background Art

[0002] As network power consumption gradually increases, the demand for energy-saving processing of base stations is increasing. When saving energy for base stations, you can turn on energy saving in specific low-traffic periods according to experience. For example, from 0 to 6 in the morning, the business volume is often relatively low. At this time, you can save energy for specific cells. However, directly judging a specific period as a period with low business volume does not conform to the actual business volume. For example, during holidays, the business volume is also relatively high during this specific time period. At this time, if the cell is saved and its business is migrated to other cells, the load on other cells is large. The large load will further cause various quality problems. Therefore, the above method has at least the following problems: If the business of a certain cell is migrated to other cells within a preset specific time period, there may be a problem of heavy business load in the cell that undertakes the business.

[0003] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of the present invention is to provide a method, device and computer-readable storage medium for migrating cell services, aiming to solve the technical problem that when migrating the services of a cell to other cells within a preset specific time period, the service load of the cell taking over the services may be heavy.

[0005] To achieve the above object, the present invention provides a method for migrating services in a cell, the method comprising the following steps:

[0006] Clustering the cells according to measurement report data of each cell to obtain at least one co-coverage cluster;

[0007] Acquire a capacity layer cell in the co-coverage cluster;

[0008] Determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in the current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell;

[0009] Migrate the services of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell.

[0010] Optionally, the step of acquiring the capacity layer cell in the co-coverage cluster includes:

[0011] Acquire the frequency priority of each of the cells in the co-coverage cluster;

[0012] Mark the primary cell and the neighboring cells in the co-coverage cluster according to the frequency priority, wherein the cell with the highest frequency priority is the primary cell, and the cells other than the primary cell are the neighboring cells;

[0013] The primary cell whose average value of the reference signal received power of all the neighboring cells is greater than a preset value is used as the capacity layer cell;

[0014] The cells marked as the primary cell and the capacity layer cells in the co-coverage cluster are removed to update the co-coverage cluster, and the step of obtaining the frequency priority of each of the cells in the co-coverage cluster is returned to be executed.

[0015] Optionally, the step of acquiring the frequency priority of each of the cells in the co-coverage cluster includes:

[0016] Acquire a first primary service area in the co-coverage cluster, and an inter-frequency point corresponding to the first primary service area;

[0017] Determining a target switching strategy type for the first primary service area based on the frequency of the first primary service area and the inter-frequency point;

[0018] Determine the coverage attribute weight corresponding to the target switching strategy type;

[0019] Determine a ratio of a sum of the coverage attribute weights of the first primary service area with the same frequency to a sum of the corresponding numbers of the different-frequency frequency points;

[0020] The frequency priority of the first primary service area is determined according to the ratio, wherein the smaller the ratio is, the higher the frequency priority is.

[0021] Optionally, after the step of determining the frequency priority of the first primary service area according to the ratio, the method further includes:

[0022] Acquire a second primary service area in which the frequency priority is not determined in the first primary service area;

[0023] Determine the average value of the coverage attribute weights of the second primary service area at the same primary frequency point;

[0024] Determine the frequency priority of the second primary service area according to the mean value, wherein the smaller the mean value is, the higher the frequency priority is;

[0025] Acquire a third primary service area in which the frequency priority is not determined in the second service area;

[0026] Determine the frequency priority of the third primary service area according to the number of frequency points of the third primary service area, wherein the smaller the number of frequency points, the higher the frequency priority;

[0027] Acquire a fourth primary service area in which the frequency priority is not determined in the third primary service;

[0028] The frequency priority of the fourth primary service is determined according to a default priority order.

[0029] Optionally, the step of determining a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in the current time period includes:

[0030] Determine, according to the historical measurement report data, the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell;

[0031] The neighboring cell whose predicted traffic volume is less than or equal to the preset load threshold is defined as the target neighboring cell.

[0032] Optionally, after the step of migrating the service of the capacity layer cell to a target neighboring cell corresponding to the capacity layer cell, the method further includes:

[0033] Inputting the historical traffic volume data of the capacity layer cell into the long short-term memory network model to predict the traffic volume within the target time period;

[0034] Determine time series data for shutting down the capacity layer cell according to the traffic volume in the target time period;

[0035] The capacity layer cell is shut down according to the time series data, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods:

[0036] The symbol is off, turning off the power amplifier;

[0037] Channel shut down, shutting down the preset transmission channel of the capacity layer cell;

[0038] Carrier shutdown, shutting down at least one carrier of the capacity layer cell;

[0039] Cell shutdown, shutting down the capacity layer cell;

[0040] Deep sleep, shutting down the target devices of the capacity layer cell, the target devices being all devices except the power supply link;

[0041] Hard shutdown, shutting down the entire device of the capacity layer cell.

[0042] Optionally, the step of clustering the cells according to the measurement report data of each cell to obtain at least one co-coverage cluster includes:

[0043] Determine, according to the primary cell information and the neighboring cell information in the measurement report data, a neighboring cell that meets a preset condition and a primary cell corresponding to the neighboring cell, wherein the preset condition is that a ratio of the number of the neighboring cells having a received power greater than or equal to a preset reference signal to the number of inter-frequency sampling point groups is greater than a preset ratio, and a total number of the neighboring cells is less than or equal to a preset number;

[0044] The neighboring cells that meet the preset condition and the primary cells corresponding to the neighboring cells are divided into the same co-coverage cluster to obtain at least one co-coverage cluster.

[0045] In addition, to achieve the above object, the present invention further provides a service migration device for a cell, the service migration device for a cell comprising:

[0046] A clustering module, configured to cluster the cells according to the measurement report data of each cell to obtain at least one common coverage cluster;

[0047] An acquisition module, used for acquiring a capacity layer cell in the co-coverage cluster;

[0048] A determination module, configured to determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in a current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell;

[0049] A migration module is used to migrate the service of the capacity layer cell to a target neighboring cell corresponding to the capacity layer cell.

[0050] In addition, to achieve the above-mentioned purpose, the present invention also provides a service migration device for a cell, wherein the service migration device for the cell includes a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program implements the steps of any of the methods described above when executed by the processor.

[0051] In addition, to achieve the above objectives, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0052] The embodiments of the present invention provide a method, device and computer-readable storage medium for migrating services of a cell. The cells are clustered into at least one co-coverage cluster according to the measurement report data of each cell, and the capacity layer cells in the co-coverage cluster are obtained. The target neighboring cell of the capacity layer cell is determined according to the predicted service volume of the capacity layer cell and the neighboring cell in the current time period. The predicted service volume is obtained by predicting the historical measurement report data of the cell and the neighboring cell under the capacity layer. The service of the capacity layer cell is migrated to the target neighboring cell corresponding to the capacity layer cell. Before migrating the service of the cell, the impact of the service volume on the neighboring cell after the migration is first predicted. Based on this, the target neighboring cell for migration is obtained. The specific cell will not be directly shut down in a specific time period, so as to avoid that the service of the cell can only be migrated to other cells, which causes a large service load on other cells. This embodiment determines the target neighboring cell based on the predicted service volume, so that the target neighboring cell can be determined in combination with the predicted service volume during migration, and the service is migrated to the target neighboring cell. The service migration can be reasonably controlled to avoid migrating the service of a certain cell to other cells within a preset specific time period, and the service load of the cell that undertakes the service is large. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the structure of the hardware operating environment involved in the embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a flow chart of a first embodiment of a method for migrating a cell service according to the present invention;

[0055] Figure 3 A schematic diagram of a flow chart of a second embodiment of a method for migrating a cell service according to the present invention;

[0056] Figure 4 A schematic diagram of a flow chart of a third embodiment of a method for migrating a cell service according to the present invention;

[0057] Figure 5 A schematic diagram of a flow chart of a fourth embodiment of a method for migrating a service in a cell according to the present invention;

[0058] Figure 6 A schematic diagram of a flow chart of a fifth embodiment of a method for migrating a cell service according to the present invention;

[0059] Figure 7 A schematic diagram of a flow chart of a sixth embodiment of a method for migrating a service in a cell according to the present invention;

[0060] Figure 8 This is a schematic diagram of the functional modules of a service migration device for a cell.

[0061] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0062] like Figure 1 As shown, Figure 1 It is a structural diagram of the hardware operating environment involved in the embodiment of the present invention.

[0063] The service migration device of the cell involved in the embodiment of the present invention may specifically be a server.

[0064] like Figure 1 As shown, the service migration device of the cell may include: a processor 1001, such as a CPU, a communication interface 1002, a memory 1003, and a communication bus 1004. Among them, the communication bus 1004 is used to realize the connection and communication between these components. The communication interface 1002 may optionally include a standard wired interface and a wireless interface. The memory 1003 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1003 may also be a storage device independent of the aforementioned processor 1001.

[0065] Those skilled in the art will understand that Figure 1 The structure of the service migration device for a cell shown in the figure does not constitute a limitation on the service migration device for a cell, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0066] like Figure 1 As shown, the memory 1003 as a computer storage medium may include a service migration program for a cell.

[0067] exist Figure 1 In the service migration device of the cell shown, the communication interface 1002 is mainly used to connect to the user terminal and perform data communication with the user terminal; and the processor 1001 can be used to call the service migration program of the cell stored in the memory 1003 and perform the following operations:

[0068] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0069] Clustering the cells according to measurement report data of each cell to obtain at least one co-coverage cluster;

[0070] Acquire a capacity layer cell in the co-coverage cluster;

[0071] Determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in the current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell;

[0072] Migrate the services of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell.

[0073] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0074] Acquire the frequency priority of each of the cells in the co-coverage cluster;

[0075] Mark the primary cell and the neighboring cells in the co-coverage cluster according to the frequency priority, wherein the cell with the highest frequency priority is the primary cell, and the cells other than the primary cell are the neighboring cells;

[0076] The primary cell whose average value of the reference signal received power of all the neighboring cells is greater than a preset value is used as the capacity layer cell;

[0077] The cells marked as the primary cell and the capacity layer cells in the co-coverage cluster are removed to update the co-coverage cluster, and the step of obtaining the frequency priority of each of the cells in the co-coverage cluster is returned to be executed.

[0078] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0079] Acquire a first primary service area in the co-coverage cluster, and an inter-frequency point corresponding to the first primary service area;

[0080] Determining a target switching strategy type for the first primary service area based on the frequency of the first primary service area and the inter-frequency point;

[0081] Determine the coverage attribute weight corresponding to the target switching strategy type;

[0082] Determine a ratio of a sum of the coverage attribute weights of the first primary service area with the same frequency to a sum of the corresponding numbers of the different-frequency frequency points;

[0083] The frequency priority of the first primary service area is determined according to the ratio, wherein the smaller the ratio is, the higher the frequency priority is.

[0084] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0085] Acquire a second primary service area in which the frequency priority is not determined in the first primary service area;

[0086] Determine the average value of the coverage attribute weights of the second primary service area at the same primary frequency point;

[0087] Determine the frequency priority of the second primary service area according to the mean value, wherein the smaller the mean value is, the higher the frequency priority is;

[0088] Acquire a third primary service area in which the frequency priority is not determined in the second service area;

[0089] Determine the frequency priority of the third primary service area according to the number of frequency points of the third primary service area, wherein the smaller the number of frequency points, the higher the frequency priority;

[0090] Acquire a fourth primary service area in which the frequency priority is not determined in the third primary service;

[0091] The frequency priority of the fourth primary service is determined according to a default priority order.

[0092] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0093] Determine, according to the historical measurement report data, the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell;

[0094] The neighboring cell whose predicted traffic volume is less than or equal to the preset load threshold is defined as the target neighboring cell.

[0095] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0096] Inputting the historical traffic volume data of the capacity layer cell into the long short-term memory network model to predict the traffic volume within the target time period;

[0097] Determine time series data for shutting down the capacity layer cell according to the traffic volume in the target time period;

[0098] The capacity layer cell is shut down according to the time series data, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods:

[0099] The symbol is off, turning off the power amplifier;

[0100] Channel shut down, shutting down the preset transmission channel of the capacity layer cell;

[0101] Carrier shutdown, shutting down at least one carrier of the capacity layer cell;

[0102] Cell shutdown, shutting down the capacity layer cell;

[0103] Deep sleep, shutting down the target devices of the capacity layer cell, the target devices being all devices except the power supply link;

[0104] Hard shutdown, shutting down the entire device of the capacity layer cell.

[0105] Further, the processor 1001 may call a service migration program of a cell stored in the memory 1003, and further perform the following operations:

[0106] Determine, according to the primary cell information and the neighboring cell information in the measurement report data, a neighboring cell that meets a preset condition and a primary cell corresponding to the neighboring cell, wherein the preset condition is that a ratio of the number of the neighboring cells having a received power greater than or equal to a preset reference signal to the number of inter-frequency sampling point groups is greater than a preset ratio, and a total number of the neighboring cells is less than or equal to a preset number;

[0107] The neighboring cells that meet the preset condition and the primary cells corresponding to the neighboring cells are divided into the same co-coverage cluster to obtain at least one co-coverage cluster.

[0108] Reference Figure 2 A first embodiment of the present invention provides a service migration method for a cell, the service migration method for a cell comprising:

[0109] Step S10, clustering the cells according to the measurement report data of each cell to obtain at least one common coverage cluster;

[0110] In this embodiment, the execution entity is the service migration device of the cell, which can specifically be a server; the cell, also known as a cellular cell, is an area covered by a base station or a part of a base station; the measurement report (Measurement Report, MR) data is uploaded by the user terminal; the co-coverage cluster is also called the co-coverage sector.

[0111] As the scale of communication network construction increases year by year, energy demand increases day by day, and network power consumption increases year by year. In order to reduce energy consumption, base stations will be energy-efficient. When saving energy, energy saving is mainly carried out for cells in the common coverage sector. The services of the cells that are energy-saving can be migrated to other cells in the common coverage sector, so that the services are not affected. In order to determine the cells that need energy saving, it is necessary to determine the cells with lower services and save energy for the cells with lower traffic volume. The exemplary technology will identify some specific cells as low-traffic cells and migrate their services to other cells. However, the energy-saving cells determined in this way are inaccurate, and the evaluation of traffic volume is inaccurate. It is not necessarily that the cell has low traffic volume in the specific time period, and it is not necessarily that the cell has high traffic volume and does not need energy saving in other time periods. Based on the inaccurate evaluation of traffic volume, It brings great risks to the network quality. For example, energy saving is performed on a certain cell from 0 to 6 o'clock, and the business originally belonging to the cell is migrated to other cells. During holidays, the business volume of the cell may still be high from 0 to 6 o'clock. At this time, after shutting down the cell to save energy, the original business of the cell is migrated to another cell, and the business volume of the other cell may be very high at this time. On this basis, the business of a certain cell is taken over again, resulting in overload of the business volume of another cell, which further leads to reduced communication quality. Based on the above analysis, it can be seen that if the business of a specific cell is migrated to other cells only according to a specific time period, although the effect of energy saving can be achieved in some cells, it will cause a large business load on other cells, thereby reducing the communication quality. Therefore, in this embodiment, at least the following problem is solved: avoid the business load of the cell that undertakes the business after the business migration is large. It can be understood that on the basis of solving this problem, the cell where the business is migrated can be further used as an energy-saving cell, such as using the capacity layer cell in this embodiment as an energy-saving cell, and shutting down the capacity layer cell to save energy, so as to achieve energy saving while avoiding the technical effect of large cell business load caused by business migration, and improve communication quality.

[0112] In this embodiment, MR data of each cell in the target area can be obtained and clustered to obtain at least one common coverage cluster. When determining the common coverage cluster, it can be determined based on the overlap of engineering parameters or original measurement report data. Alternatively, in order to eliminate the impact of coverage from the source, the common coverage cluster can be determined based on the further setting of the coverage level threshold.

[0113] Step S20, obtaining the capacity layer cells in the co-coverage cluster;

[0114] In order to obtain the capacity layer cells in the co-coverage cluster, the co-coverage cluster needs to be divided into capacity layer cells and coverage layer cells. The capacity layer cells mainly refer to cells with strong service absorption properties, and the coverage layer cells mainly refer to cells with weak service absorption properties, which are mainly used to ensure signal continuity and deep coverage to improve the basic needs of user communication connections. When obtaining the capacity layer cells, the cells with strong service absorption properties can be distinguished as capacity layer cells and coverage layer cells according to the MRO data in the MR data, and the capacity layer cells can be obtained.

[0115] Step S30, determining a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cells in the current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cells;

[0116] After obtaining the capacity layer cells in the common coverage cluster, only the capacity layer cells can be used as energy-saving cells. When the capacity layer cells are shut down, all the services of the capacity layer cells are taken over by other cells in the common coverage cluster. The other cells in the common coverage cluster can be capacity layer cells or coverage layer cells. It should be noted that if a capacity layer cell has been determined as a cell that needs energy saving, the cell will no longer be used as a cell to take over the services. When other cells in the common coverage cluster take over the services, the order of taking over is in accordance with the existing network's frequency interoperability strategy, and the neighboring cells with higher priorities are prioritized as migration objects, as shown in the following table. The following table lists the correspondence between frequency bands, priorities and migration thresholds. It is for example only and is not limited to this.

[0117] Combined with the following table, it is further explained how to obtain the neighboring cells that can be used as the preparatory migration objects from the co-coverage cluster. It should be noted that before obtaining the target neighboring cells, it is necessary to first determine the preparatory migration objects. The method of determining the preparatory migration objects is as follows: obtain all sampling points of a period of the preparatory shutdown cell, and select the neighboring cell with the highest priority and the largest RSRP level strength value in the MR sampling data as the service migration object cell. Among them, if the highest priority cell cannot meet the minimum migration threshold, then select the neighboring cell with the second highest priority that appears in the sampling data and meets the minimum migration threshold of the frequency as the preparatory migration object, and compare them in turn until all sampling points of the capacity layer cell in a sampling period are attributed to other co-coverage neighboring cells, and all preparatory migration objects are determined; by summarizing and counting all the MR sampling points of the capacity layer cell in a period, the proportion of services that can be migrated in each co-coverage neighboring cell can be predicted.

[0118]

[0119] Taking the above table as an example, the first step is to obtain the level strengths of other neighboring cells in MR, which are: D_1(-110), D_2(-95), F_1(-100), E_1(-90), E_2(-95), and F_2(-101). Then, the second step is to screen out the level threshold load requirements, which are D_2(-95), F_1(-100), E_1(-90), E_2(-95), and F_2(-101). The third step is to compare the priorities, and it can be obtained that the E band has the highest priority, including E_1(-90) and E_2(-95). The fourth step is to compare and obtain the maximum level value, which is E_1(-90), so that the E_1 cell is determined to be the object of the migration. Then, in order to avoid the problem of excessive traffic after migration, resulting in heavy load and reduced network quality, the target neighboring cell is further determined. The target neighboring cell is a neighboring cell that will not cause excessive traffic after the migration, resulting in excessive load and reduced network quality.

[0120] After determining the prepared migration object, the service volume after migrating the service from the capacity layer cell to the prepared migration object is further predicted to obtain the predicted service volume. When the predicted service volume is less than or equal to the preset service volume, the prepared migration object is used as the target neighboring cell.

[0121] When determining the target cell of the capacity layer according to the predicted traffic volume of the capacity layer cell and the neighboring cell, the predicted traffic volume of the neighboring cell can be predicted after the traffic volume of the capacity layer cell is migrated to the neighboring cell. When the predicted traffic volume of the neighboring cell is lower than the preset threshold, the neighboring cell can be used as the target neighboring cell. The historical measurement report data is the measurement report data within a period of history. According to the historical measurement report data, the traffic volume of the neighboring cell can be predicted in the spatial dimension after the traffic of the capacity layer cell is migrated to the neighboring cell, so as to obtain the impact on the traffic volume of the neighboring cell. In order to ensure the accuracy of the prediction, the latest period of historical measurement report data can be used to predict the predicted traffic volume.

[0122] The predicted business volume of the capacity layer cell and the neighboring cell can also be predicted. For example, if the predicted business volume of the capacity layer cell is low, lower than the first preset threshold, and the predicted business volume of the neighboring cell is also low, lower than the second preset threshold, then it can be determined that shutting down the capacity layer cell will not generate a large business load on the neighboring cell that undertakes the business, and thus the neighboring cell can also be used as the target neighboring cell.

[0123] Step S40: Migrate the service of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell.

[0124] After determining the target neighboring cell, the services of the capacity layer cell are migrated to the target neighboring cell corresponding to the capacity layer cell, and the capacity layer cell can be shut down to achieve energy saving, thereby achieving energy saving of the cell. At the same time, it can also avoid the problem of excessive service load of other cells caused by shutting down the capacity layer cell, resulting in a decrease in network quality.

[0125] In this embodiment, the cells are clustered into at least one co-coverage cluster according to the measurement report data of each cell, the capacity layer cell in the co-coverage cluster is obtained, and the target neighboring cell of the capacity layer cell is determined according to the predicted business volume of the capacity layer cell and the neighboring cell in the current time period. The predicted business volume is obtained by predicting the historical measurement report data of the cell and the neighboring cell under the capacity layer. The business of the capacity layer cell is migrated to the target neighboring cell corresponding to the capacity layer cell. Before the business of the migrated cell, the impact of the business volume on the neighboring cell after the migration is first predicted. On this basis, the target neighboring cell for migration is obtained. The specific cell will not be directly shut down in a specific time period, so that the business of the cell can only be migrated to other cells, which causes a large business load on other cells. This embodiment determines the target neighboring cell based on the predicted business volume, so that the target neighboring cell can be determined in combination with the predicted business volume during migration, and the business is migrated to the target neighboring cell. The business migration can be reasonably controlled to avoid migrating the business of a certain cell to other cells within a preset specific time period, and the problem of a large business load of the cell that undertakes the business.

[0126] Reference Figure 3 The second embodiment of the present invention provides a method for migrating services in a cell. Figure 2 In the first embodiment shown, step S20 includes:

[0127] Step S21, obtaining the frequency priority of each of the cells in the co-coverage cluster;

[0128] After obtaining the common coverage cluster, the cells in the common coverage cluster are sorted by frequency priority, which is also called energy-saving order priority. Then, coverage evaluation and service migration prediction are performed based on MRO data. Finally, the capacity layer cells can be used as energy-saving cells, and the energy-saving cell set is output. The energy-saving cells are shut down according to the energy-saving cell set, which can save energy while avoiding network quality degradation caused by excessive service load.

[0129] The frequency priority can be set in the default priority order, or can be set in sequence according to the switching order, reselection order, frequency point number order and default order.

[0130] Frequency priority is used in the network to distinguish different cell network levels, so as to achieve the purpose of allocating users between cells. Generally, the higher the cell level, that is, the higher the frequency priority, the stronger the service absorption attribute of the cell, and the lower the cell level, that is, the lower the frequency priority, the stronger the deep coverage attribute. In order to ensure continuous and deep coverage, as well as the basic requirements of mobile network user perception, in the order of implementing objects of energy-saving strategies in co-coverage clusters, priority is given to strong coverage attributes and weak coverage attributes to ensure that network coverage is not affected. Frequency priority can be used for coverage prediction and service migration prediction after energy saving.

[0131] Step S22, marking the primary cell and the neighboring cells in the co-coverage cluster according to the frequency priority, wherein the cell with the highest frequency priority is the primary cell, and the cells other than the primary cell are the neighboring cells;

[0132] All cells in the co-coverage cluster unit take turns as the main cell according to the frequency priority from high to low, and other cells in the co-coverage cluster serve as neighboring cells. The MR coverage prediction algorithm is used to calculate and distinguish the capacity layer and the coverage layer until all cells in the co-coverage cluster unit complete the calculation as the main cell. When taking turns as the main cell for calculation, if the cell with high frequency priority has been determined as the capacity layer after calculation, it will no longer be a neighboring cell and will no longer participate in the subsequent low frequency priority coverage prediction.

[0133] In this embodiment, the cell with the highest frequency priority is first taken as the main cell, and the cells other than the main cell are taken as neighboring cells to further determine the capacity layer cell. After that, the cell that has been marked as the main cell is proposed, and the main cell and the neighboring cells in the co-coverage cluster are returned according to the frequency priority mark. In this way, the main cell is taken in turn from high to low frequency priority, and other cells in the co-coverage cluster are taken as neighboring cells. The capacity layer and the coverage layer are distinguished through the MR coverage prediction algorithm.

[0134] When marking the primary cell and the neighboring cell according to the frequency priority, MRO data is required. The user terminal uploads an MRO data every 5 ms. Each MRO data includes but is not limited to the following information: terminal UE_id, serving cell reference signal received power (Reference Signal Received Power, RSRP), neighboring cell RSRP level strength, frequency, physical cell identifier (Physical Cell Identifier, PCI).

[0135] In this embodiment, when dividing the capacity layer cell and the coverage layer cell in the co-coverage cluster, the neighboring cell with the strongest RSRP is screened out according to each MRO data. If its RSRP is greater than or equal to the preset value, it is regarded as a valid sampling point. The preset value can be -110dBm. The valid sampling points of all main cells in a period are summarized. If the proportion of valid sampling points is greater than 95%, the main cell is determined to be a capacity layer cell, otherwise it is determined to be a coverage layer cell. Only the capacity layer cell can be used as an energy-saving cell, and only the capacity layer cell can be shut down to save energy.

[0136] Step S23, taking the primary cell whose average value of the reference signal received power of all the neighboring cells is greater than a preset value as the capacity layer cell;

[0137] When determining the mean value of the reference signal received power, coverage prediction may be performed to obtain a coverage prediction value, wherein the coverage prediction value is the mean value of the reference signal received power.

[0138] For example, assuming that a common coverage cluster [MR_Cell_1…MR_Cell_M] includes M cells, the number of MR sampling data reported in a period T is n, and the level of the strongest neighbor in each MR data is RSRP_neighbour. The calculation formula for the coverage prediction value F is as follows:

[0139]

[0140] If F_server≥95%, the primary cell is the capacity layer; if F_server<95%, the primary cell is the coverage layer.

[0141] Step S24, removing the cells marked as the primary cell and the capacity layer cells in the co-coverage cluster to update the co-coverage cluster, and returning to the step of obtaining the frequency priority of each cell in the co-coverage cluster.

[0142] Eliminate cells and capacity layer cells that have been marked as primary cells to update the co-coverage cluster, and continue to return to execute step S22 until each cell has been marked as a primary cell or as a capacity layer cell. The purpose of this cycle is to enable each cell to be marked as a capacity layer cell or a coverage layer cell, thereby realizing the division of capacity layer cells and coverage layer cells.

[0143] Taking two sampling points in the MRO data as an example, each sampling point has four data. The cells that also need to save energy during the same period are eliminated, and the cells with the strongest RSRP levels in the remaining two neighboring cells are selected to retain one sampling data. Among them, some capacity layer cells may also need to save energy, so they will not be the objects of undertaking the business.

[0144] Time ue_id Main area Primary RSRP Neighborhood Neighbor RSRP t1 id_1 cell_server_1 -80 Cell_neighbour_1 -85 Energy saving during the same period t1 id_1 cell_server_1 -80 Cell_neighbour_2 -89 t1 id_1 cell_server_1 -80 Cell_neighbour_3 -97 Energy saving during the same period t1 id_1 cell_server_1 -80 Cell_neighbour_4 -75 √ t2 id_1 cell_server_1 -115 Cell_neighbour_1 -118 t2 id_1 cell_server_1 -90 Cell_neighbour_2 -80 Energy saving during the same period t2 id_1 cell_server_1 -90 Cell_neighbour_3 -94 √ t2 id_1 cell_server_1 -112 Cell_neighbour_4 -100 Energy saving during the same period

[0145] Each user terminal samples and obtains a piece of data of the neighboring cell with the strongest RSRP level, that is, the above two samples obtain the strongest sampling data of the two neighboring cells:

[0146] Time ue_id Main area Primary cell RSRP Neighborhood Neighboring cell RSRP t1 id_1 cell_server_1 -80 Cell_neighbour_4 -75 √ t2 id_1 cell_server_1 -90 Cell_neighbour_3 -94 √

[0147] Assuming that the total number of RSRP sampling points in the main cell is 100, and the number of sampling points with RSRP ≥ -110dBm is 95, then X is 95%. If the number of sampling points with RSRP ≥ -110dBm in the neighboring cell is 80, then F is 80%, and this cell is a coverage layer cell and cannot save energy. If the number of sampling points with RSRP ≥ -110dBm in the neighboring cell is 95 or more, then F ≥ X, and this cell is judged to be a capacity layer cell and can participate in energy saving.

[0148] By calculating the MR coverage prediction, all cells in the energy-saving area are labeled as capacity layer cells / coverage layer cells, providing a basis for judging whether the cells can be energy-efficient. The actual calculation example is as follows:

[0149] time Service area Proportion Layer Identification 2020 / 7 / 16 460-00-314093-130 100％ Capacity layer 2020 / 7 / 16 460-00-598319-130 26.38％ Covering

[0150] The percentage refers to the percentage of neighboring cells with optimal RSRP ≥ -110dBm.

[0151] In this embodiment, by obtaining the frequency priority of each cell in the common coverage cluster, the main cell and the neighboring cells in the common coverage cluster are marked according to the frequency priority, wherein the cell with the highest frequency priority is the main cell, and the cells other than the main cell are the neighboring cells, and the main cell whose average reference signal receiving power of all neighboring cells is greater than a preset value is used as the capacity layer cell, and the cells and capacity layer cells marked as the main cell in the common coverage cluster are eliminated to update the common coverage cluster, and return to execute the step of determining the frequency priority of each cell in the common coverage cluster, thereby dividing the capacity layer cell and the coverage layer cell, and only using the capacity layer cell as the energy-saving cell can ensure the coverage effect of the coverage layer cell, and according to the frequency priority network structure layering, the order of implementing objects of the energy-saving strategy in the common coverage cluster is prioritized from strong coverage attributes to weak coverage attributes to ensure that network coverage is not affected.

[0152] Reference Figure 4 The third embodiment of the present invention proposes a service migration method for a cell. Figure 3 In the second embodiment shown in the figure, only steps S211 to S2112 are shown, and step S21 includes:

[0153] Step S211, obtaining a first primary service area in the co-coverage cluster and an inter-frequency point corresponding to the first primary service area;

[0154] In this embodiment, when determining the frequency priority, the frequency priority is determined in four ways, namely, switching order, reselection order, frequency point number order, and default energy-saving order. When the frequency priority of the cell is successfully determined based on any one of the methods, the original frequency priority can be maintained, and the frequency priority is not re-determined by other methods. When the frequency priority is not determined by a certain method, the next method can be used to determine the frequency priority. The reason for not determining the frequency priority may be parameter loss, or that the frequency priorities of multiple cells are the same. In order to accurately distinguish the frequency priority, the next method can be used to continue to determine the frequency priority.

[0155] In this embodiment, steps S211 to S215 are switching sorting, steps S216 to S218 are reselection sorting, steps S219 to S2110 are frequency point number sorting, and steps S2111 to S2112 are default priority sorting.

[0156] The first primary service area is all the primary service areas in the co-coverage cluster. The primary service area is also called the primary frequency (F server ), the main service usually configures several different frequency points F neibour .

[0157] Step S212, determining a target switching strategy type of the first primary service area based on the frequency of the first primary service area and the inter-frequency point;

[0158] The main principles of the switching strategy based on the switching sorting are as follows: A5 event (poor source side, good target side) is used for switching from high priority band to low priority band, and switching to low priority band is performed only when the high priority signal is poor and the low priority signal is good. A4 event is used for switching from low priority band to high priority band. As long as the quality of the high priority signal is good, it will switch to the high priority band. A3 event (best compared) is used between bands with the same priority level to ensure that users switch to cells with good signals in time. It can be seen that services occupy high priority frequencies first.

[0159] Step S213, determining the coverage attribute weight corresponding to the target switching strategy type;

[0160] By assigning a value to the switching strategy of the different frequency points, defined as the coverage attribute weight S, and then summing up all the coverage attribute weights, the main frequency F server The coverage attribute strength C is mathematically represented to quantify the coverage attribute strength of the main frequency in the co-coverage cluster; the A4 event is set when the low-priority cell switches to the high-priority cell, and the coverage attribute weight S is equal to 1; the A3 event is set when the same-priority cell switches, and the coverage attribute weight S is equal to 0; the A5 event is set when the high-priority cell switches to the low-priority cell, and the coverage attribute weight S is equal to -1.

[0161] Assume that the main frequency of a cell is F server , configure k different frequency points F neibour , the main frequency of the cell is F server The coverage attribute weight for:

[0162]

[0163] Step S214, determining the ratio of the sum of the coverage attribute weights of the first primary service area with the same frequency to the sum of the corresponding numbers of the different-frequency frequency points;

[0164] The ratio is the strength of the coverage attribute. Assuming that there are n cells with the same main frequency as the above cell in the co-coverage cluster, the main frequency F in the cluster server The coverage attribute strength is equal to the sum of the coverage attribute weights of n cells with the same main frequency and then the average of the number of switching strategies:

[0165]

[0166] Similarly, the above formula is used to calculate all the main frequencies in the co-covering cluster to obtain the coverage attribute strength C of all the main frequencies.

[0167] Take the following table as an example:

[0168] Among them, the main cell frequency F1, the cell main frequency F1 coverage attribute weight

[0169] Primary cell frequency Inter-frequency neighboring frequency Switching strategy type Cover attribute weight S F1 D1 A4 1 F1 D3 A4 1 F1 E1 A4 1 F1 F2 A3 0 F1 FDD900 A5 -1

[0170] In addition, based on the corresponding relationship between the number of main frequency points, the number of different frequency points and the weight of the coverage attribute in the following table, the coverage attribute strength can be obtained:

[0171] Among them, coverage attribute strength

[0172] Step S215, determining the frequency priority of the first primary service area according to the ratio, wherein the smaller the ratio is, the higher the frequency priority is;

[0173] According to the strength of coverage attribute, that is, sorted by ratio, the larger the ratio, the more it is biased towards the coverage layer and the lower the frequency priority; the smaller the ratio, the more it is biased towards the capacity layer and the higher the frequency priority. The frequency priority can be set based on this principle.

[0174] Step S216, obtaining a second primary service area in which the frequency priority is not determined in the first primary service area;

[0175] When the frequency priority cannot be determined, a second main service area in which the frequency priority has not been determined in the first main service area is obtained. The second main service area refers to a service area in which the frequency priority is not marked in the first main service area. In addition, the second main service area may also include a main service area with the same frequency priority as the first main service area.

[0176] Step S217, determining the average value of the coverage attribute weights of the second primary service area at the same primary frequency point;

[0177] For the second primary service area, the frequency priority can be obtained by reselection sorting. The reselection priority of the existing 4G / 5G network cell ranges from 3 to 7. The configuration coverage attribute weights are as follows:

[0178]

[0179]

[0180] Take the co-covered cluster as the unit, according to the main frequency F server Classify, average the coverage attribute weights S of cells with the same main frequency, and obtain the F of each main frequency under the co-coverage cluster. server The coverage attribute strength based on reselection, the coverage attribute strength of reselection is the mean value in this embodiment:

[0181]

[0182] Step S218, determining the frequency priority of the second primary service area according to the mean value, wherein the smaller the mean value is, the higher the frequency priority is;

[0183] Sorted by mean value, the larger the mean value, the more it leans towards the coverage layer and the lower the frequency priority; the smaller the mean value, the more it leans towards the capacity layer and the higher the frequency priority.

[0184] Step S219, obtaining a third primary service area in the second service area for which the frequency priority has not been determined;

[0185] The third primary service area is a primary service area whose frequency priority has not yet been determined in the second primary service area. In order to further sort the frequency priorities, this embodiment determines the frequency priority based on the number of frequency points.

[0186] Step S2110, determining the frequency priority of the third primary service area according to the number of frequency points of the third primary service area, wherein the smaller the number of frequency points, the higher the frequency priority;

[0187] Among them, the D-band frequencies are more complex. D1, D2, D4, D5, and D6 are currently allocated to 5G, and D3, D7, and D8 are allocated to LTE.

[0188] Step S2111, obtaining a fourth primary service area for which the frequency priority is not determined in the third primary service;

[0189] The fourth main service area is the main service area whose frequency priority cannot be determined in the third main service area; after the switching sorting, reselection sorting and frequency point number sorting mentioned above, the main service area whose frequency priority has not been determined is sorted according to the default priority.

[0190] Step S2112: Determine the frequency priority of the fourth primary service according to a default priority order.

[0191] The default priority order correspondence table is as follows:

[0192]

[0193]

[0194] Through the above method, the frequency priority of each cell in the co-coverage cluster can be determined.

[0195] In this embodiment, the first main service area in the co-coverage cluster and the inter-frequency frequency point corresponding to the first main service area are obtained; based on the frequency and the inter-frequency frequency point of the first main service area, the target switching strategy type of the first main service area is determined; the coverage attribute weight corresponding to the target switching strategy type is determined; the ratio of the sum of the coverage attribute weights of the first main service area with the same frequency to the sum of the corresponding number of inter-frequency frequency points is determined; the frequency priority of the first main service area is determined according to the ratio, wherein the smaller the ratio, the higher the frequency priority, and the second main service area whose frequency priority is not determined in the first main service area is obtained; the average value of the coverage attribute weights of the second main service area with the same main frequency point is determined; Determine the frequency priority of the second main service area according to the mean, wherein the smaller the mean, the higher the frequency priority; obtain the third main service area whose frequency priority has not been determined in the second service area; determine the frequency priority of the third main service area according to the number of frequency points of the third main service area, wherein the smaller the number of frequency points, the higher the frequency priority; obtain the fourth main service area whose frequency priority has not been determined in the third main service; determine the frequency priority of the fourth main service according to the default priority order; thereby being able to completely determine the frequency priority of each cell based on the switching order, reselection order, frequency point number order and default priority order, thereby avoiding cells with undetermined frequency priority.

[0196] Reference Figure 5 The fourth embodiment of the present invention proposes a service migration method for a cell. Figure 2 In the first embodiment shown, step S30 includes:

[0197] Step S31, determining the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell according to the historical measurement report data;

[0198] According to the historical measurement report data, the neighboring cell for service migration is determined, that is, the neighboring cell that undertakes the service, or the preparatory migration object of the first embodiment. Thereafter, the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell is determined. The predicted service volume can be characterized by the utilization rate of the physical resource block (PRB). It is calculated based on 1 equivalent carrier for an ordinary 20MTDD cell, 0.5 equivalent carrier for a 10M F-band cell, 0.75 equivalent carrier for an FDD900 or A-band, 1.5 equivalent carrier for an FDD1800, and 2.5 equivalent carriers for a 3D-MIMIO cell.

[0199] Refer to the table below:

[0200]

[0201] Assume that the PRB utilization rate RPB of the content layer cell in a certain period is Server , the PRB utilization rate of a neighboring cell or a pre-migration object is RPB Neighbour , the service migration ratio is m%, then the PRB utilization rate of the migration object after energy saving of the capacity layer cell is RPB_ 迁移后 , the prediction formula is as follows:

[0202]

[0203] The above announcement is for migration under the same network standard, such as TDD to TDD or FDD to FDD.

[0204] The following is a list of migration scenarios under different network standards, such as migration from TDD to FDD or from FDD to TDD:

[0205]

[0206] Step S32: The neighboring cell whose predicted traffic volume is less than or equal to the preset load threshold is defined as the target neighboring cell.

[0207] The preset load threshold, also called the safety load threshold in this embodiment, predicts that the neighboring cell with a traffic volume less than or equal to the preset load threshold is the target neighboring cell. For example, if the traffic volume PRB utilization rate RPB_ 迁移后If the predicted value does not exceed the safety load threshold (PRB utilization rate 40%), the next highest priority capacity layer cell of the co-coverage cluster will be evaluated again with reference to the above service migration prediction algorithm. The evaluation objects include the current capacity layer cell and the previously evaluated capacity layer cell, that is, the service migration ratio needs to be accumulated with the previously migrated capacity layer cell. Until all capacity layer cells are accumulated and predicted and evaluated, or the neighboring cell traffic PRB utilization rate RPB_ 迁移后 If the predicted value exceeds the safety load threshold, the migration prediction is terminated and each target neighboring area is obtained.

[0208] In this embodiment, the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell is determined based on historical measurement report data; the neighboring cell whose predicted service volume is less than or equal to the preset load threshold is taken as the target neighboring cell, thereby improving the quality of migration, avoiding quality problems caused by directly shutting down the cell for energy saving, and avoiding the problem of triggering various network quality problems after energy saving through performance indicator monitoring, and only being able to urgently restore the energy-saving cell.

[0209] Reference Figure 6 The fifth embodiment of the present invention proposes a service migration method for a cell. Figure 2 In the first embodiment shown, after step S40, the following steps are further included:

[0210] Step S50, inputting the historical traffic volume data of the capacity layer cell into the long short-term memory network model to predict the traffic volume within the target time period;

[0211] The long short-term memory network model is used to predict the business volume within a target time period in the time dimension. The long short-term memory network model is pre-trained based on historical data.

[0212] To determine whether a capacity layer cell can be shut down in a certain period of time to save energy, it is necessary to predict its network load. To predict future network load, it is necessary to first analyze historical data and find patterns from it. The present invention predicts future data through the Long Short-Term Memory (LSTM) algorithm. LSTM is a special recurrent neural network (RNN); during the training of the original RNN, as the training time increases and the number of network layers increases, it is easy to have gradient explosion or gradient disappearance problems, resulting in the inability to process longer sequence data, and thus unable to obtain information on long-distance data, and LSTM is suitable for processing such problems.

[0213] LSTM neural network is mainly implemented through three gating logics (forgetting, input, and output) to solve the gradient vanishing and gradient exploding problems in the long sequence training process. As a typical nonlinear network model prediction mechanism, the autonomous learning function of the neural network is used to train the network model sequence, which can achieve high-precision network capacity prediction with a long prediction cycle. At the same time, the machine learning logistic regression algorithm is used to predict the burst type network capacity. The following is the process of constructing the long short-term memory network model in this embodiment:

[0214] This patent uses the Keras Sequential model to build the model, and verifies whether the algorithm is effective by predicting the indicators for the next day through the downlink PRB utilization rate in the Lin'an area over the past seven days.

[0215] Data preprocessing, first normalize the data. The purpose of normalization is to limit the data to a certain range after processing, reduce the impact of singular sample data on the prediction results, and improve the convergence speed of the algorithm. The Keras model generally uses the MinMaxScaler function to shrink the data to between 0 and 1.

[0216] Model building, the model adopted is: model = keras.models.Sequential()

[0217] model.add(LSTM(input_shape=(168,1),units=16,dropout=0.2,activation=None,stateful=False,)

[0218] input_shape: The first layer of the Sequential model needs to specify the input_shape parameter, but the subsequent layers do not need it. input_shape contains two elements. The first element represents the length of each input sample sequence, and the second element represents how many input data an element in each sequence has. This embodiment uses an indicator for 7 days and 24 hours for prediction, so this parameter is set to (168,1). units: is the dimension of the output of this layer. If the value is too large, it will overfit and need to be debugged. In this embodiment, it is set to units=16, and the output dimension is (168,16); dropout: how much proportion of neuron output (unit output) activation is invalidated, the default value is 0, the purpose is to prevent overfitting, and this embodiment is set to 0.2; activation: the type of activation function, this activation function determines the output range of each unit each time it is calculated. This embodiment does not need to change the range of the output value, so the activation function is not set. stateful: whether to expand. Expanding the LSTM calculation will speed up the regression calculation process, but it will take up more memory. It can be expanded on a small sequence. For model training, 7 days of hourly data were used as the training sample number for training, and 1 day of data was used for validation. The model indicators, such as loss function and accuracy, were tested after each epoch. The model was trained 100 times in a cycle. After about 20 trainings, the Training and Validation loss values ​​were greatly reduced and stabilized below 0.005, with no overfitting or underfitting, and the training network was normal. One week of historical data was imported into the trained model for calculation, and finally the prediction results for the next day were output.

[0219] Step S60, determining time series data for shutting down the capacity layer cell according to the traffic volume in the target time period;

[0220] According to the business volume in the target time period and the preset shutdown processing method, the time series data of the shutdown processing of the capacity layer cell is obtained, and the time series data is the sequence data of the shutdown processing method based on time changes.

[0221] Step S70, shutting down the capacity layer cell according to the time series data, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods: symbol shutdown, shutting down the power amplifier; channel shutdown, shutting down the preset transmission channel of the capacity layer cell; carrier shutdown, shutting down at least one carrier of the capacity layer cell; cell shutdown, shutting down the capacity layer cell; deep sleep, shutting down the target devices of the capacity layer cell, the target devices are all devices except the power supply link; hard shutdown, shutting down the entire devices of the capacity layer cell.

[0222] In this embodiment, soft and hard shutdown technologies are adopted. Based on the five soft shutdown energy-saving technologies of "symbol shutdown, channel shutdown, carrier shutdown, cell shutdown, and deep sleep" and one "hard shutdown" energy-saving technology characteristic, energy-saving means are matched and applied to the energy-saving cell output in step 4 according to the characteristics and technical requirements of each energy-saving technology. The order of power-saving method decision is determined according to the power-saving efficiency as follows: hard shutdown → deep sleep → cell shutdown → carrier shutdown → channel shutdown → symbol shutdown, and the optimal power-saving solution is output.

[0223] The soft and hard shutdown characteristics and implementation conditions are as follows:

[0224]

[0225] In this embodiment, the historical traffic volume data of the capacity layer cell is input into the long short-term memory network model to predict the traffic volume in the target time period; according to the traffic volume in the target time period, the time series data for shutting down the capacity layer cell is determined; according to the time series data, the capacity layer cell is shut down, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods: symbol shutdown, shutting down the power amplifier; channel shutdown, shutting down the preset transmission channel of the capacity layer cell; carrier Shutdown, shut down at least one carrier of the capacity layer cell; cell shutdown, shut down the capacity layer cell; deep sleep, shut down the target devices of the capacity layer cell, the target devices are all devices except the power supply link; hard shutdown, shut down the entire devices of the capacity layer cell; thereby achieving the following technical effects: it can avoid the problem of business compliance after power saving: it avoids the power-saving cell estimation method in the exemplary technology that mainly uses daily or hourly business volume, and does not take into account the impact of business volume changes in long periods (such as festivals, holidays, fixed activities, etc.), and business load problems are likely to occur after the node. A better power-saving effect can be achieved: avoid direct hard shutdown to save power in the exemplary technology. The existing 4 / 5 network structure is complex in level, and a single hard shutdown power-saving strategy is no longer suitable for complex network structures. This embodiment can adapt different power-saving strategies to different coverage scenarios. At the same time, various power-saving means also formulate strategies and apply them layer by layer, which can adapt to the power-saving needs of various scenarios; avoid the problem that only manual shutdown can be selected in the early morning in the exemplary technology, and avoid the technical problem that the exemplary technology is affected by the mixed construction of different base stations and different manufacturers in the same coverage area, and cannot carry out energy saving around the clock. The differences in energy-saving functions of base stations of different manufacturers also make the power-saving operation cumbersome, and the overall difficulty of power saving increases;

[0226] In addition, this embodiment adopts a new MRO data analysis method to identify co-coverage clusters, applies intelligent algorithms to successfully establish coverage and business prediction models, accurately outputs energy-saving cells, and introduces soft and hard shutdown energy-saving technologies. According to different scenarios and the characteristics of each shutdown technology, accurate energy-saving plans are intelligently allocated to achieve all-weather, low-risk, and highly accurate intelligent energy saving, and achieve system-level energy saving for the entire network; use long short-term memory network algorithms to predict network capacity, solve the problem of capacity prediction distortion caused by periodic business changes, and truly achieve accurate business prediction.

[0227] Reference Figure 7 The sixth embodiment of the present invention proposes a service migration method for a cell. Figure 2 In the first embodiment shown, step S10 includes:

[0228] Step S11, determining, according to the primary cell information and the neighboring cell information in the measurement report data, a neighboring cell that meets a preset condition and a primary cell corresponding to the neighboring cell, wherein the preset condition is that a ratio of the number of the neighboring cells having a received power greater than or equal to a preset reference signal to the number of inter-frequency sampling point groups is greater than a preset ratio, and the total number of the neighboring cells is less than or equal to a preset number;

[0229] In order to eliminate the impact of weak coverage caused by power saving at the source, this embodiment eliminates weak coverage cells when obtaining the common coverage cluster; in the exemplary technology, when determining the common coverage cluster, only the antenna angle is combined to determine the overlapping part to determine the common coverage cluster, but this method is not accurate.

[0230] According to the measurement report data, obtain the required original measurement report data and extract the required MRO information. An example of MRO information is shown in the following table:

[0231] name What the data means data eNB id Base station ID 595632 object id Service cell CELL ID 152481921 MmeUeS1apId UE terminal ID 361261 TimeStamp Timestamp 2020-06-17T10:00:00.122 MR.LteScRSRP Serving cell RSRP 56 MR.LteScEarfcn Service cell frequency 38400 MR.LteScPci Serving cell PCI logical number 308 MR.LteNcRSRP Neighboring cell RSRP 53 MR.LteNcEarfcn Neighboring cell frequency 36275 MR.LteNcPci Neighboring cell PCI 308

[0232] There are multiple sampling points in each cell in the MRO data. Each sampling point contains different RSRP level values ​​of the main neighboring cells and different UE_ids. The UE reports to the base station according to the MR sampling period. The MRO data of each cell is calculated according to the number of sampling points within the RSRP threshold statistical period, and the proportion of neighboring cells to the main cell is calculated. (Neighboring cell sampling point RSRP ≥ -110dBm sampling point) / hetero-frequency sampling point group > 10% and the number of neighboring cell sampling points TOP10 is used as the judgment condition. The main neighboring cells that meet the threshold are included in a co-coverage cluster. The whole network is divided into different co-coverage clusters according to the MRO data, and the scene cluster unit library is constructed. When the MRO data changes, its scene cluster unit library changes accordingly. The judgment method for the effective hetero-frequency MRO measurement cell is: MRO hetero-frequency sample number / MRO total sample number * 100> 10%.

[0233] Step S12: group the neighboring cells satisfying a preset condition and the primary cells corresponding to the neighboring cells into the same co-coverage cluster to obtain at least one co-coverage cluster.

[0234] Through MRO data calculation, the 4G cell co-coverage sector group under the physical site has been delineated, and the 5G cell is added to the 4G co-coverage sector group by latitude, longitude and azimuth, and the coverage layer and capacity layer are distinguished according to the MRO coverage prediction module. Each time the hetero-frequency MRO data is turned on, the valid hetero-frequency MRO measurement cells that meet the mobile network are added to the MRO business model, and the latest MRO business model is called each time. If the proportion of hetero-frequency MRO sampling points is too small, the MRO data collection of the cell is considered invalid, and the co-coverage sector group adopts the industrial parameter-level co-coverage sector algorithm.

[0235] In this embodiment, the neighboring cells that meet the preset conditions and the main cells corresponding to the neighboring cells are determined according to the main cell information and the neighboring cell information in the measurement report data, and the preset conditions are that the ratio of the number of neighboring cells with a received power greater than or equal to the preset reference signal to the number of heterofrequency sampling point groups is greater than the preset ratio, and the total number of neighboring cells is less than or equal to the preset number; the neighboring cells that meet the preset conditions and the main cells corresponding to the neighboring cells are divided into the same co-coverage cluster to obtain at least one co-coverage cluster, which can eliminate the influence of weak coverage when dividing the co-coverage clusters, thereby improving the accuracy of the determination of the co-coverage clusters.

[0236] In addition, the seventh embodiment of the present invention provides a process of overall implementation of the present invention in combination with a specific application scenario:

[0237] Taking the Lin'an urban area of ​​Hangzhou as a pilot, there are 366 physical sites in the pilot area, including 1,406 4 / 5G logical cells.

[0238] 1) First, enable MR inter-frequency measurement in all cells within the planned energy-saving area, collect three days of MRO data in the cells, and extract the required MRO data information.

[0239] 2) Based on the MRO data information of the extracted 1406 cells, the sampling point ratios of the main cell and the TOP10 neighboring cells are calculated. When the sampling point ratio of the neighboring cell RSRP ≥ -110dBm is greater than 10%, the main neighboring cell is classified as the same co-coverage cluster, and the co-coverage clusters of all cells are calculated and divided in turn. A total of 368 co-coverage clusters are calculated in the pilot area.

[0240] 3) After obtaining the co-coverage cluster, the energy-saving priority of the frequencies involved in each co-coverage cluster is sorted to obtain the frequency priority. The frequency priority is determined according to the switching sorting, reselection and other sorting methods to finally determine the priority of each frequency in the co-coverage cluster.

[0241] 4) After obtaining the frequency priority of each cell, the cells are evaluated for coverage in order. The neighboring cell with the strongest RSRP is selected through each MRO data. If its RSRP ≥ -110dBm is considered as a valid sampling point, the valid sampling points of a main cell in a cycle are summarized. If the proportion of valid sampling points is greater than 95%, the main cell is determined to be a capacity layer cell, otherwise it is determined to be a coverage layer cell. The calculation formula of MR coverage prediction value F is as follows:

[0242]

[0243] F server ≥95%, the service cell is the capacity layer; F server <95%, the service cell is the coverage layer. In this pilot area, 368 co-coverage clusters were gathered from 1406 cells. The MR prediction algorithm was used to calculate that there were 1134 capacity layer cells and 372 coverage layer cells in the pilot area.

[0244] 5) After distinguishing the coverage layer cells and the capacity layer cells, it is necessary to make a capacity prediction for the cells in order to make the next business migration prediction. The capacity prediction in this case is based on the LSTM long short-term memory network algorithm. The hourly KPI indicators of the cells 7 days before the prediction date are extracted for training, and a prediction model for the main indicators at the cell level is established. The capacity is predicted for all cells, and the hourly business model for November 27 is predicted.

[0245] 6) Output power-saving cells based on capacity prediction, service migration prediction and soft and hard shutdown technical characteristics; first, according to the service migration prediction algorithm (T_L=T_N+T_CM), calculate that after the capacity layer cell is shut down, the predicted value of the traffic volume PRB utilization rate (T_L) of each neighboring cell under the co-coverage cluster does not exceed the safe load threshold (40%), that is, the capacity cell is considered to be power-saving, and the power-saving method judgment rules after sliding calculation of all capacity layer cells are finally output. Among the 372 capacity cells, 236 cells can actually participate in power saving, and the overall power saving ratio is 16.79%. So far, the cells and energy-saving methods that can be energy-saving in each period are judged through MRO data and long short-term memory network prediction methods, and are executed in order from high to low priority according to hard shutdown → deep sleep → cell shutdown → carrier shutdown → channel shutdown → symbol shutdown.

[0246] 7) After the implementation of the above power saving plan, the number of kWh of electricity saved in the pilot area was counted, with a total daily saving of 4,235.6 kWh and a saving of approximately RMB 2,404.4 per day. If the full-day forecast is subsequently adopted and the results of the pilot area are used to predict the entire Zhejiang province, it is estimated that 226.17 million kWh of electricity can be saved throughout the year, saving approximately RMB 113 million in electricity bills per year.

[0247] 8) For the power-saving operations in this pilot area, a business evaluation was conducted after the power-saving operation. Mainly from the comparison of daily KPI indicators, all major indicators remained basically stable, and the number of high-load cells and high-traffic warning cells did not increase significantly, achieving the expected effect.

[0248] like Figure 8 As shown, Figure 8 This is a schematic diagram of an embodiment of a service migration device for a cell according to the present invention. The service migration device for a cell includes:

[0249] A clustering module 10, configured to cluster the cells according to the measurement report data of each cell to obtain at least one common coverage cluster;

[0250] An acquisition module 20, configured to acquire a capacity layer cell in the co-coverage cluster;

[0251] A determination module 30, configured to determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in a current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell;

[0252] The migration module 40 is used to migrate the service of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell.

[0253] In one embodiment, the acquisition module 20 is also used to perform the following steps: obtain the frequency priority of each of the cells in the common coverage cluster; mark the main cell and the neighboring cells in the common coverage cluster according to the frequency priority, wherein the cell with the highest frequency priority is the main cell, and the cells other than the main cell are the neighboring cells; the main cell whose average reference signal receiving power of all the neighboring cells is greater than a preset value is used as the capacity layer cell; eliminate the cells in the common coverage cluster that have been marked as the main cell and the capacity layer cells to update the common coverage cluster, and return to execute the step of obtaining the frequency priority of each of the cells in the common coverage cluster.

[0254] In one embodiment, the acquisition module 20 is also used to perform the following steps: obtain the first main service area in the co-coverage cluster, and the hetero-frequency point corresponding to the first main service area; determine the target switching strategy type of the first main service area based on the frequency of the first main service area and the hetero-frequency point, and determine the coverage attribute weight corresponding to the target switching strategy type; determine the ratio of the sum of the coverage attribute weights of the first main service area with the same frequency to the sum of the corresponding numbers of hetero-frequency points; determine the frequency priority of the first main service area according to the ratio, wherein the smaller the ratio, the higher the frequency priority.

[0255] In one embodiment, the acquisition module 20 is also used to perform the following steps: acquire a second main service area in the first main service area for which the frequency priority has not been determined; determine the mean of the coverage attribute weights of the second main service area with the same main frequency point; determine the frequency priority of the second main service area according to the mean, wherein the smaller the mean, the higher the frequency priority; acquire a third main service area in the second service area for which the frequency priority has not been determined; determine the frequency priority of the third main service area according to the number of frequency points of the third main service area, wherein the smaller the number of frequency points, the higher the frequency priority; acquire a fourth main service area in the third main service area for which the frequency priority has not been determined; and determine the frequency priority of the fourth main service area according to a default priority order.

[0256] In one embodiment, the determination module 30 is further used to perform the following steps: determining the predicted traffic volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell according to the historical measurement report data; and taking the neighboring cell whose predicted traffic volume is less than or equal to the preset load threshold as the target neighboring cell. In one embodiment, the service migration device of the cell further includes a prediction module and a shutdown module, wherein:

[0257] The prediction module is used to input the historical traffic volume data of the capacity layer cell into the long short-term memory network model to predict the traffic volume within a target time period;

[0258] The determination module 30 is further configured to determine the time series data for shutting down the capacity layer cell according to the traffic volume in the target time period;

[0259] The shutdown module is used to shut down the capacity layer cell according to the time series data, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods: symbol shutdown, shutting down the power amplifier; channel shutdown, shutting down the preset transmission channel of the capacity layer cell; carrier shutdown, shutting down at least one carrier of the capacity layer cell; cell shutdown, shutting down the capacity layer cell; deep sleep, shutting down the target devices of the capacity layer cell, the target devices are all devices except the power supply link; hard shutdown, shutting down the entire devices of the capacity layer cell.

[0260] The clustering module 10 is also used to perform the following steps: determine the neighboring cells that meet the preset conditions and the main cells corresponding to the neighboring cells according to the main cell information and the neighboring cell information in the measurement report data, wherein the preset conditions are that the ratio of the number of the neighboring cells that are greater than or equal to the preset reference signal receiving power to the number of heterofrequency sampling point groups is greater than the preset ratio, and the total number of the neighboring cells is less than or equal to the preset number; and divide the neighboring cells that meet the preset conditions and the main cells corresponding to the neighboring cells into the same co-coverage cluster to obtain at least one co-coverage cluster.

[0261] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0262] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0263] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a service migration device (server or other computer equipment) of a cell to execute the methods described in each embodiment of the present invention.

[0264] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for migrating services in a cell, It is characterized in that The service migration method of the cell comprises the following steps: Clustering the cells according to measurement report data of each cell to obtain at least one co-coverage cluster; Obtain the frequency priority of each of the cells in the co-coverage cluster, wherein a first main service area in the co-coverage cluster and an inter-frequency frequency point corresponding to the first main service area are obtained, and a target switching strategy type of the first main service area is determined based on the frequency of the first main service area and the inter-frequency frequency point, and a coverage attribute weight corresponding to the target switching strategy type is determined, and a ratio of the sum of the coverage attribute weights of the first main service area with the same frequency to the sum of the corresponding number of inter-frequency frequency points is determined, and the frequency priority of the first main service area is determined according to the ratio, wherein the smaller the ratio, the higher the frequency priority; Mark the primary cell and the neighboring cells in the co-coverage cluster according to the frequency priority, wherein the cell with the highest frequency priority is the primary cell, and the cells other than the primary cell are the neighboring cells; The primary cell whose average value of the reference signal received power of all the neighboring cells is greater than a preset value is used as a capacity layer cell in the co-coverage cluster; Determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in the current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell; Migrate the services of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell.

2. The service migration method of a cell according to claim 1, It is characterized in that After the step of using the primary cell whose average value of the reference signal received power of all the neighboring cells is greater than a preset value as a capacity layer cell in the co-coverage cluster, the method further includes: The cells marked as the primary cell and the capacity layer cells in the co-coverage cluster are removed to update the co-coverage cluster, and the step of obtaining the frequency priority of each of the cells in the co-coverage cluster is returned to be executed.

3. The service migration method of a cell as claimed in claim 1, It is characterized in that After the step of determining the frequency priority of the first primary service area according to the ratio, the method further includes: Acquire a second primary service area in which the frequency priority is not determined in the first primary service area; Determine the average value of the coverage attribute weights of the second primary service area at the same primary frequency point; Determine the frequency priority of the second primary service area according to the mean value, wherein the smaller the mean value is, the higher the frequency priority is; Acquire a third primary service area in which the frequency priority is not determined in the second primary service area; Determine the frequency priority of the third primary service area according to the number of frequency points of the third primary service area, wherein the smaller the number of frequency points, the higher the frequency priority; Acquire a fourth primary service area in which the frequency priority is not determined in the third primary service; The frequency priority of the fourth primary service is determined according to a default priority order.

4. The service migration method of a cell according to claim 1, It is characterized in that The step of determining the target neighboring cell of the capacity layer cell according to the predicted traffic volume of the capacity layer cell and the neighboring cell in the current time period comprises: Determine, according to the historical measurement report data, the predicted service volume of the neighboring cell after the service of the capacity layer cell is migrated to the neighboring cell; The neighboring cell whose predicted traffic volume is less than or equal to the preset load threshold is defined as the target neighboring cell.

5. The service migration method of a cell as claimed in claim 1, It is characterized in that After the step of migrating the service of the capacity layer cell to the target neighboring cell corresponding to the capacity layer cell, the method further includes: Inputting the historical traffic volume data of the capacity layer cell into the long short-term memory network model to predict the traffic volume within the target time period; Determine time series data for shutting down the capacity layer cell according to the traffic volume in the target time period; The capacity layer cell is shut down according to the time series data, wherein the time series data is sequence data of a shutdown processing method based on time changes, and the shutdown processing method includes at least two of the following processing methods: The symbol is off, turning off the power amplifier; Channel shut down, shutting down the preset transmission channel of the capacity layer cell; Carrier shutdown, shutting down at least one carrier of the capacity layer cell; Cell shutdown, shutting down the capacity layer cell; Deep sleep, shutting down the target devices of the capacity layer cell, the target devices being all devices except the power supply link; Hard shutdown, shutting down the entire device of the capacity layer cell.

6. The service migration method of a cell as claimed in claim 1, It is characterized in that The step of clustering the cells according to the measurement report data of each cell to obtain at least one co-coverage cluster comprises: Determine, according to the primary cell information and the neighboring cell information in the measurement report data, a neighboring cell that meets a preset condition and a primary cell corresponding to the neighboring cell, wherein the preset condition is that a ratio of the number of the neighboring cells having a received power greater than or equal to a preset reference signal to the number of inter-frequency sampling point groups is greater than a preset ratio, and a total number of the neighboring cells is less than or equal to a preset number; The neighboring cells that meet the preset condition and the primary cells corresponding to the neighboring cells are divided into the same co-coverage cluster to obtain at least one co-coverage cluster.

7. A service migration device for a cell, It is characterized in that The service migration device of the cell includes: A clustering module, configured to cluster the cells according to the measurement report data of each cell to obtain at least one common coverage cluster; An acquisition module is used to obtain the frequency priority of each cell in the co-coverage cluster, wherein the first main service area in the co-coverage cluster and the hetero-frequency points corresponding to the first main service area are obtained, and based on the frequency of the first main service area and the hetero-frequency points, the target switching strategy type of the first main service area is determined, and the coverage attribute weight corresponding to the target switching strategy type is determined, and the ratio of the sum of the coverage attribute weights of the first main service area with the same frequency to the sum of the corresponding hetero-frequency points is determined, and the frequency priority of the first main service area is determined according to the ratio, wherein the smaller the ratio, the higher the frequency priority; the main cell and the neighboring cell in the co-coverage cluster are marked according to the frequency priority, wherein the cell with the highest frequency priority is the main cell, and the cells other than the main cell are the neighboring cells; the main cell whose average reference signal received power of all the neighboring cells is greater than a preset value is used as the capacity layer cell in the co-coverage cluster; A determination module, configured to determine a target neighboring cell of the capacity layer cell according to predicted traffic volumes of the capacity layer cell and the neighboring cell in a current time period, wherein the predicted traffic volumes are predicted by historical measurement report data of the capacity layer cell and the neighboring cell; A migration module is used to migrate the service of the capacity layer cell to a target neighboring cell corresponding to the capacity layer cell.

8. A service migration device for a cell, It is characterized in that The service migration device of the cell includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program implements the steps of the method according to any one of claims 1 to 6 when executed by the processor.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images