Resource scheduling method and device, medium and program product

Abstract

The invention relates to the technical field of wireless communication, and particularly provides a resource scheduling method and device, a medium and a program product, and the resource scheduling method comprises the steps: obtaining a plurality of historical load index values, respectively associated with a plurality of load indexes, of a plurality of cells in a historical time period; inputting a plurality of historical load index values associated with each load index of each cell in a historical time period into a load index prediction model to obtain a target load index value associated with each load index of each cell in a target time period; based on the target load index value and the load index threshold value, determining a matching score value of two cells in each cell group in the plurality of cell groups in the target time period, based on a matching score value set formed by the matching score values and a cell pairing strategy prediction model, determining a cell pairing result, and based on the cell pairing result, determining a matching result of the two cells in each cell group in the target time period; and performing resource scheduling in the target time period. The resource utilization rate and the resource allocation rationality can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of wireless communication technology, and in particular to a resource scheduling method, device, medium and program product. Background Art

[0002] With the rapid development of 5G commercial networks and the rapid increase in the penetration rate of 5G terminals, 5G is gradually developing into the main service bearer layer; due to differences in cell coverage scenarios, some cells will experience high network load in one period and low network load in another period. In order to effectively reuse network physical resources, cells with high and low load fluctuations can be identified and expanded or reduced in capacity.

[0003] In the related art, the network load index of each cell and / or other information that can reflect the cell load status is usually obtained. When it is determined that the cell load is too high, the cell is expanded; or, when it is determined that the cell load is too low, the cell is reduced in capacity to perform resource scheduling.

[0004] However, the resource scheduling scheme provided in the related technology mainly performs resource scheduling from the load dimension of a single cell. Usually, there will be a situation where there are no resources available in high-load cells and idle resources in low-load cells, resulting in unreasonable resource allocation and even affecting normal business services. Summary of the invention

[0005] The present disclosure is proposed in view of the above problems. The present disclosure provides a resource scheduling method, device, medium and program product, which can improve resource utilization and resource allocation rationality.

[0006] According to one aspect of the present disclosure, a resource scheduling method is provided, comprising:

[0007] Acquire multiple historical load index values ​​respectively associated with multiple load indexes in multiple cells within a historical period;

[0008] Inputting a plurality of historical load index values ​​associated with each load index in each cell in a historical period into a load index prediction model respectively, to obtain a target load index value associated with each load index in each cell in a target period;

[0009] Based on the target load index value and the load index threshold value associated with each load index of each cell in the target time period, determining the matching score value of two cells in each cell group in the target time period, wherein the multiple cell groups are obtained after matching two cells in the multiple cells, and the matching score value is used to characterize the degree of load difference between the two cells;

[0010] Based on a matching score value set consisting of matching score values ​​of the two cells in each cell group in the target time period and a cell pairing strategy prediction model, a cell pairing result is determined, and based on the cell pairing result, resource scheduling is performed in the target time period, wherein the cell pairing result includes a paired cell determined for each cell for resource scheduling.

[0011] According to another aspect of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the above-mentioned resource scheduling method.

[0012] According to another aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned resource scheduling method is implemented.

[0013] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, and the computer program implements the above-mentioned resource scheduling method when executed by a processor.

[0014] The resource scheduling method, device, medium and program product provided by the present invention obtain a target load index value associated with each load index of each cell in a target time period through multiple historical load index values ​​respectively associated with multiple load indexes in multiple cells in a historical time period and a load index prediction model; then, based on each target load index value and load index threshold associated with each load index of each cell in the target time period, determine the load difference degree between two cells in a cell group determined after pairing the multiple cells, and match a paired cell for resource scheduling for each cell based on the load difference degree between the two cells in the cell group and a cell pairing strategy prediction model, so as to perform resource scheduling based on the paired cell of each cell at a target time, and determine a paired cell for resource scheduling for each cell in the target time period through the load difference value between the two cells in the multiple cells in the target time period, so as to achieve reasonable allocation and scheduling of resources of multiple cells and improve resource utilization of multiple cells.

[0015] It is to be understood that both the foregoing general description and the following detailed description are exemplary, and are intended to provide further explanation of the technology as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0017] Figure 1 It is a schematic diagram of an implementation scenario of a resource scheduling solution according to an embodiment of the present disclosure.

[0018] Figure 2 is a flowchart illustrating resource scheduling according to an embodiment of the present disclosure.

[0019] Figure 3 It is a flowchart further illustrating a cell pairing strategy prediction model for training in resource scheduling in an embodiment of the present disclosure.

[0020] Figure 4 is a block diagram illustrating a resource scheduling apparatus according to an embodiment of the present disclosure.

[0021] Figure 5 is a schematic diagram illustrating a computer program product according to an embodiment of the present disclosure.

[0022] Figure 6 is a hardware block diagram illustrating an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present disclosure more obvious, the exemplary embodiments according to the present disclosure will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure, and it should be understood that the present disclosure is not limited to the exemplary embodiments described here.

[0024] In the related technology, during the resource scheduling process, after obtaining the network load indicators of each cell and / or other information that can reflect the cell load status, if it is determined that the cell load is too high, the cell is expanded; or, if it is determined that the cell load is too low, the cell is reduced in capacity to perform resource scheduling.

[0025] However, this resource scheduling scheme generally schedules resources for each cell in turn from the load dimension of a single cell. For high-load cells, resources usually need to be expanded. However, the premise of resource expansion is that resources are available. When there are no available resources in the high-load cell, resource scheduling identification will occur; and for low-load cells, after reducing the resource occupancy of the cell, the excess resources are idle, and idle resources occur, resulting in unreasonable resource scheduling.

[0026] In order to solve the above problems, the present disclosure provides a resource scheduling solution, such as Figure 1 As shown, Figure 1 FIG. 1 shows a schematic diagram of an implementation scenario of a resource scheduling solution provided by an exemplary embodiment of the present disclosure. Figure 1 As shown, the implementation scenario 100 includes a network device 101 and a resource scheduling device 102, wherein the network device 101 is a device that can monitor indicator data of a cell, and the resource scheduling device 102 can be a computer, a notebook, a tablet computer or a server.

[0027] A communication link may be established between the network device 101 and the resource scheduling device 102 to implement the resource scheduling solution provided in the embodiment of the present disclosure.

[0028] Figure 2 A flowchart of a resource scheduling method of an exemplary embodiment of the present disclosure is shown. The method can be applied to a resource scheduling device, such as Figure 2 As shown, the method of the embodiment of the present disclosure may include:

[0029] Step S201, obtaining a plurality of historical load index values ​​respectively associated with a plurality of load indexes in a historical period for a plurality of cells;

[0030] Step S202, inputting multiple historical load index values ​​associated with each load index in each cell in the historical period into the load index prediction model respectively, to obtain the target load index value associated with each load index in each cell in the target period;

[0031] Step S203, determining a matching score value of two cells in each cell group in the target time period based on the target load index value and the load index threshold value associated with each load index of each cell in the target time period;

[0032] Among them, the multiple cell groups are obtained by matching two cells in the multiple cells, and the matching score value is used to represent the load difference between the two cells;

[0033] Step S204, determining a cell pairing result based on a matching score set composed of matching score values ​​of two cells in each cell group in the target time period and a cell pairing strategy prediction model, and performing resource scheduling in the target time period based on the cell pairing result;

[0034] The cell pairing result includes a paired cell determined for each cell for resource scheduling.

[0035] In summary, the resource scheduling method provided by the embodiment of the present disclosure obtains the target load index value associated with each load index of each cell within the target time period through multiple historical load index values ​​respectively associated with multiple load indexes in multiple cells within the historical time period, and a load index prediction model; then, based on each target load index value and load index threshold associated with each load index of each cell within the target time period, the load difference degree between two cells in the cell group determined after pairing the multiple cells is determined, and based on the load difference degree between the two cells in the cell group and the cell pairing strategy prediction model, a paired cell for resource scheduling is matched for each cell, so as to perform resource scheduling based on the paired cell of each cell at the target time, and a paired cell for resource scheduling within the target time period is determined for each cell through the load difference value between the two cells in the multiple cells within the target time period, so as to realize the reasonable allocation and scheduling of resources of multiple cells and improve the resource utilization of multiple cells.

[0036] The following Figure 2 The specific implementation methods of each step in the embodiment shown are described in detail:

[0037] In step S201, the resource scheduling device may obtain a plurality of historical load index values ​​respectively associated with a plurality of load indexes for a plurality of cells within a historical period.

[0038] In an embodiment of the present disclosure, a resource scheduling device may periodically schedule resources of multiple cells. Specifically, this may be determined based on actual needs, and the embodiment of the present disclosure does not limit this. For example, the resource scheduling device may perform resource scheduling once every one or two days.

[0039] It should also be noted that the historical period is a preset duration before the current resource scheduling moment, which can be determined based on actual needs, and the embodiment of the present disclosure does not limit this; wherein, the historical period may include multiple unit durations, wherein each unit duration is determined based on actual needs, and the embodiment of the present disclosure does not limit this; for example, the historical period may be 7 days, and the unit duration may be 1 day; or, the historical duration may be 3 days, and the unit duration may be 12 hours; wherein, the duration of the resource scheduling cycle is greater than or equal to the unit duration in the historical period, which can prevent the process of repeatedly determining the cell pairing result and prevent waste of resources.

[0040] The load indicator is a parameter indicator used to characterize the load status of a cell. Specifically, it can be determined based on actual needs. The embodiment of the present disclosure does not limit this. For example, the load indicator may include: average evolved radio access bearer (Evolved Radio Access Bearer, E-RAB) traffic, the number of radio resource controls (Radio Resource Control, RRC) with data transmission, physical uplink shared channel (Physical Uplink Shared CHannel, PUSCH) utilization, physical downlink shared channel (Physical Downlink Shared Channel, PDSCH) utilization, physical downlink control channel (Physical Downlink Control Channel, PDCCH) utilization, uplink traffic and / or downlink traffic, etc.; Optionally, the load indicators for different business services may also be different. In this embodiment of the present disclosure, resource scheduling can be performed for different business services, and multiple load indicators are load indicators associated with target services that require resource scheduling.

[0041] In an optional implementation, the process of the resource scheduling device obtaining multiple historical load index values ​​associated with multiple load indexes in a historical period for multiple cells may include: sending a load data acquisition request to a network device, and obtaining multiple historical load index values ​​associated with multiple load indexes in a historical period for the cells returned by the network device. The historical load index data can be obtained in real time through the network device, thereby improving the efficiency of obtaining the historical load index values.

[0042] In an optional implementation, the process by which a resource scheduling device obtains multiple historical load indicator values ​​respectively associated with multiple load indicators for multiple cells within a historical period may include: reading multiple historical load indicator values ​​respectively associated with multiple load indicators for multiple cells within a historical period in the resource scheduling device.

[0043] It can be understood that, in the embodiment of the present disclosure, the multiple historical load index values ​​respectively associated with the multiple load indexes include: the historical load index value of each load index associated with each unit duration of the historical period.

[0044] For example, the historical period is 7 days, the unit time can be 1 day, and the multiple load indicators include: E-RAB traffic, PUSCH utilization and PDCCH utilization. Then the multiple historical load indicator values ​​associated with the multiple load indicators include: historical E-RAB traffic, PUSCH utilization and PDCCH utilization associated with each day within 7 days.

[0045] In step S202, the resource scheduling device inputs multiple historical load index values ​​associated with each load index of each cell in the historical period into the load index prediction model to obtain the target load index value associated with each load index of each cell in the target period.

[0046] In the embodiment of the present disclosure, the load index prediction model is a time series model, which can be determined based on actual needs. The embodiment of the present disclosure is not limited to this. For example, the load index prediction model can be a Long Short Term Memory Networks (LSTM) model, or an Auto-Regression and Moving Average Model (ARMA), etc.; the target period is later than the historical period, and the target period can include at least one unit time length, wherein the time length of each unit time length contained in the target period is the same as the time length of each unit time length contained in the historical period.

[0047] In an optional embodiment, the process of training a load index prediction model of a resource scheduling device may include: obtaining a sample data set, then dividing the sample data set into a training set and a validation set, and using the training set to train a time series network, after the time series network converges, verifying the time series network through the validation set, and, if the verification passes, obtaining a load index prediction model; wherein the sample data set includes: multiple sample historical load index values ​​​​respectively associated with multiple load indicators for each cell within a sample historical period, and load index label values ​​​​respectively associated with multiple load indicators for each cell within a sample target period, wherein the sample target period is later than the sample historical period.

[0048] For example, taking the load index prediction model as an LSTM model, the process of training the load index prediction model of the resource scheduling device may include: collecting sample historical load index values ​​of n types of load indicators respectively associated with M unit durations within the target duration; translating the sample historical load index values ​​of each load index Fi associated with M unit durations within the target duration in a time window of T+S time length to obtain MT-S+1 samples associated with each load index Fi; and combining the MT-S+1 samples associated with each load index Fi to obtain a sample data set, wherein the sample data set is (MT-S+1, T+S, n), the number of samples associated with each load index in the sample data set is MT-S+1, and the time step of each sample is T+S; wherein the time step T is the duration of the sample historical period, indicating that the sample historical period contains T unit durations, the time step S is the duration of the sample target period, indicating that the sample target period contains S unit durations, the feature dimension of each sample is n, indicating that it contains n types of load indicators, and the value range of i is 1 to n.

[0049] Furthermore, the sample data set is divided into a training set tensor_x and a validation set tensor_y, and the mean square error is used as the loss function. The LSTM model is trained by the back propagation algorithm in the training set tensor_x; after the model converges, the validation set tensor_y is used to verify the prediction effect of the trained LSTM model, where the verification indicator can be the mean absolute percentage error.

[0050] Among them, using the mean square error as the loss function, the process of training the LSTM model through the back propagation algorithm in the training set tensor_x may include: inputting multiple sample historical load index values ​​of each load index in the sample historical period in the training set into the LSTM model to obtain the sample load index prediction value of the load index in the sample target period; then, inputting the sample load index prediction value and the load index label value of the load index in the sample target period into the loss function to obtain the loss function value of the current iteration; when the loss function value is greater than or equal to the loss function value threshold, repeating the above iterative process to obtain the LSTM model; the loss function value threshold can be determined based on actual needs, and the embodiments of the present disclosure do not limit this.

[0051] In an optional embodiment, when the target time period includes a unit time length, the resource scheduling device inputs multiple historical load index values ​​associated with each load index for each cell in the historical time period into the load index prediction model respectively, and the process of obtaining the target load index value associated with each load index for each cell in the target time period may include: inputting multiple historical load index values ​​associated with each load index for each cell in the historical time period into the load index prediction model respectively, and obtaining the target load index value associated with each load index for each cell in the target time period.

[0052] In an optional embodiment, when the target time period includes multiple unit time periods, the resource scheduling device inputs the multiple historical load index values ​​associated with each load index of each cell in the historical time period into the load index prediction model respectively, and the process of obtaining the target load index value associated with each load index of each cell in the target time period may include: inputting the multiple historical load index values ​​associated with each load index of each cell in the historical time period into the load index prediction model respectively, to obtain the target load index value associated with each load index of each cell in the earliest unit time period of the target time period; then, determining the earliest unit time period and the target number of unit time periods before the earliest unit time period as the updated historical time period, and repeating the above process of obtaining the target load index value associated with each load index of each cell in the earliest unit time period of the target time period until the target load index value associated with each load index of each cell in the last unit time period of the target time period is obtained; wherein the target number is the difference between the number of unit time periods in the historical time period and 1.

[0053] For example, if the historical period contains T unit time lengths, the target period contains S unit time lengths, and the number of load indicators is n, then in the process of the resource scheduling device determining the target load indicator value associated with each load indicator in each cell within the target period, for each cell, for the i-th load indicator among the n load indicators, the data input to the LSTM model is a tensor of (1, T, ni), and the output is a tensor of the form (1, S, ni), which represents the target load indicator value of the i-th load indicator among the n load indicators in the target period S, and the above process is repeated to obtain the target load indicator value associated with each load indicator in each cell within the target period S.

[0054] In step S203, the resource scheduling device determines the matching score value of two cells in each cell group in the target time period based on the target load index value and the load index threshold associated with each load index of each cell in the target time period.

[0055] In the embodiment of the present disclosure, multiple cell groups are obtained after matching two cells in multiple cells, and the matching score value is used to characterize the degree of load difference between the two cells; the load index threshold associated with each load index can be determined based on actual needs, and the embodiment of the present disclosure does not limit this. Among them, the load index threshold associated with each load index may include: the minimum load index threshold and the maximum load index threshold of the load index threshold interval associated with each load index; or, the load index threshold associated with each load index may include: the load index threshold reference value associated with each load index.

[0056] In an optional implementation, the resource scheduling device determines the matching score values ​​of two cells in each cell group in a target time period based on the target load indicator value and the load indicator threshold associated with each load indicator of each cell in the target time period, and the process may include: for each cell group, according to the load indicator threshold associated with each load indicator, and the comparison result of the target load indicator value associated with each load indicator of the first cell in the cell group in the target time period, obtaining a first modified target load indicator value associated with each load indicator of the first cell in the target time period; and according to the load indicator threshold associated with each load indicator, and the comparison result of the target load indicator value associated with each load indicator of the second cell in the cell group in the target time period. The target load index values ​​associated with the load indicators are compared to obtain the second modified target load index value associated with each load indicator for the second cell during the target period; then, the first modified target load index value and the second modified target load index value associated with each load indicator for the first cell and the second cell in each cell group are determined during the target period to obtain the load difference amount associated with each load indicator for the first cell and the second cell in each cell group during the target period; further, the load difference amounts respectively associated with multiple load indicators for the first cell and the second cell in each cell group during the target period are weightedly summed to obtain the matching score value of the two cells in each cell group during the target period. According to the load index threshold value associated with each load index, the target load index value associated with each load index in the target time period can be corrected for the two cells in each cell group, and the load difference amount associated with each load index in the target time period between the two cells in each cell group can be determined according to the corrected value. The actual load state of each cell in the cell group can be measured by the load index threshold value, and the target load index value associated with each load index of each cell can be corrected to ensure that the determined load difference amount matches the actual situation; and the load difference amount associated with each load index in the target time period between the two cells in each cell group is weighted and summed to obtain the matching score value of the two cells in each cell group in the target time period. The influence of different load indicators on the cell load state can be considered to improve the reliability of the matching score value of the two cells in each cell group in the target time period.

[0057] In an optional implementation, when the load index threshold associated with each load index includes: a minimum load index threshold and a maximum load index threshold of a load index threshold interval associated with each load index, the resource scheduling device may obtain a first modified target load index value associated with each load index in the first cell within the target period according to the load index threshold associated with each load index and a comparison result of the target load index value associated with each load index in the first cell in the cell group within the target period, and the process may include:

[0058] For each load indicator, if the target load indicator value associated with the load indicator for the first cell in the cell group within the target time period is greater than or equal to the minimum load indicator threshold associated with the load indicator, and the target load indicator value associated with the load indicator for the first cell in the cell group within the target time period is less than or equal to the maximum load indicator threshold associated with the load indicator, then it is determined that the first modified target load indicator value associated with the load indicator for the first cell within the target time period is 0.

[0059] If the target load index value associated with the load index of the first cell in the cell group during the target time period is less than the minimum load index threshold associated with the load index, the difference between the target load index value associated with the load index and the minimum load index threshold is determined as the first revised target load index associated with the load index of the first cell during the target time period.

[0060] If the target load index value associated with the load index of the first cell in the cell group during the target time period is greater than the maximum load index threshold associated with the load index, then the difference between the target load index value associated with the load index and the maximum load index threshold is determined as the first modified target load index associated with the load index of the first cell during the target time period.

[0061] In an optional implementation, when the load index threshold associated with each load index includes: a load index threshold reference value associated with each load index, the resource scheduling device obtains a first modified target load index value associated with each load index in the first cell within the target period according to the load index threshold associated with each load index and a comparison result of the target load index value associated with each load index in the first cell in the cell group within the target period, and the process may include:

[0062] For each load indicator, if the target load indicator value associated with the load indicator of the first cell in the cell group within the target time period is equal to the load indicator threshold reference value associated with the load indicator, then it is determined that the first modified target load indicator value associated with the load indicator of the first cell within the target time period is 0.

[0063] If the target load index value associated with the load index of the first cell in the cell group during the target time period is less than the load index threshold reference value associated with the load index, then the difference between the target load index value associated with the load index and the load index threshold reference value is determined as the first corrected target load index associated with the load index of the first cell during the target time period.

[0064] If the target load index value associated with the load index of the first cell in the cell group during the target time period is greater than the load index threshold reference value associated with the load index, then the difference between the target load index value associated with the load index and the load index threshold reference value is determined as the first corrected target load index associated with the load index of the first cell during the target time period.

[0065] It should be noted that in the embodiment of the present disclosure, the process by which the resource scheduling device determines the first modified target load indicator value associated with each load indicator for the second cell within the target time period is similar to the process by which the resource scheduling device determines the first modified target load indicator value associated with each load indicator for the first cell within the target time period, and the embodiment of the present disclosure will not go into details about this.

[0066] For example, for a load index Fi among n types of load indexes, the maximum load index threshold associated with the load index Fi is Fi_high, the minimum load index threshold associated with the load index Fi is Fi_low, the j-th cell included in a cell group is cellj, the k-th cell is cellk, j is any one of a plurality of cells, k is also any one of a plurality of cells, and j is different from k; wherein, the target load index value associated with the load index Fi for the j-th cell cellj in the target time period is Fi_cellj, then the resource scheduling device may obtain the first modified target load index value associated with each load index in the first cell in the cell group in the target time period according to the load index threshold associated with each load index and the comparison result of the target load index value associated with each load index in the first cell in the cell group in the target time period, and the process may include:

[0067] If the target load index value Fi_cellj is greater than or equal to the minimum load index threshold Fi_low, and the target load index value Fi_cellj is less than or equal to the maximum load index threshold Fi_high, then it is determined that the first modified target load index value Feature_Tensor_j associated with the load index Fi for the jth cell cellj within the target period is 0.

[0068] If the target load index value Fi_cellj is less than the minimum load index threshold value Fi_low, then the first modified target load index value Feature_Tensor_j associated with the load index Fi for the j-th cell cellj during the target period is determined to be Fi_cellj-Fi_low; or, if the target load index value Fi_cellj is greater than the maximum load index threshold value Fi_high, then the first modified target load index value Feature_Tensor_j associated with the load index Fi for the j-th cell cellj during the target period is determined to be Fi_cellj-Fi_high.

[0069] It can be understood that in the embodiment of the present disclosure, the load difference Feature_Tensor_substract_jk associated with the load index Fi between the j-th cell cellj and the k-th cell cellk during the target period is:

[0070] Feature_Tensor_substract_jk=Feature_Tensor_j-Feature_Tensor_k; (Formula 1)

[0071] In Formula 1, Feature_Tensor_k is the first modified target load index value associated with the load index Fi for the k-th cell cellk in the target period.

[0072] Among them, the resource scheduling device performs weighted summation on the load difference amounts associated with multiple load indicators respectively for the first cell and the second cell in each cell group within the target time period to obtain the matching score value of the two cells in each cell group during the target time period, which may include: for each cell group, based on the load difference amount associated with each load indicator between the first cell and the second cell in the cell group during the target time period, constructing a target matrix, wherein the columns of the target matrix are n types of load indicator dimensions, and the rows of the target matrix are S unit durations within the target time period; then, based on the preset weight configuration, multiplying the load difference amount in each column of the target matrix by the weight value associated with each column of the load indicator to obtain an updated target matrix, and further, solving the sum of the updated target matrices for each cell group to obtain the matching score value of the two cells in each cell group during the target time period.

[0073] It is understandable that, in the embodiment of the present disclosure, the weight value configured for each load index can be determined based on actual needs, and the embodiment of the present disclosure does not limit this. Among them, the sum of the weight values ​​configured for multiple load indicators is 1.

[0074] In step S204, the resource scheduling device may determine the cell pairing result based on the matching score value set composed of the matching score values ​​of the two cells in each cell group in the target time period and the cell pairing strategy prediction model, and perform resource scheduling in the target time period based on the cell pairing result.

[0075] In the embodiment of the present disclosure, the cell pairing result includes a paired cell determined for each cell for resource scheduling.

[0076] It should be noted that in the disclosed embodiment, the greater the load difference between cells, the higher the demand for capacity expansion resources of the high-load cell among the two cells, and the higher the remaining resources of the low-load cell. Providing the resources of the low-load cell to the high-load cell can utilize resources more efficiently. In the disclosed embodiment, the goal of cell pairing is to maximize the utilization of multiple cells, so that the strategy prediction model determines a paired cell for resource scheduling for each cell based on a set of matching score values ​​between two cells in multiple cells, with the goal of maximizing resource utilization.

[0077] In an optional embodiment, the process of training a strategy prediction model of a resource scheduling device may include: obtaining a sample data set, and using the sample data set to train a time series network, and obtaining a load index prediction model after the time series network converges; then, inputting multiple sample historical load index values ​​associated with each load index of each cell in the sample historical period into the load index prediction model, respectively, to obtain a sample load index prediction value associated with each load index of each cell in the sample target period; based on the sample load index prediction value and load index threshold associated with each load index of each cell in the sample target period, determining the sample matching score value of the two cells in each cell group in the sample target period, and training the cell pairing strategy prediction model based on the sample matching score value of the two cells in each cell group in the sample target period. In the process of training the cell pairing strategy prediction model, the sample matching score value used is obtained by processing multiple sample historical load index values ​​associated with each load index for each cell within the sample historical period through the load index prediction model, and obtaining the sample load index prediction value associated with each load index for each cell within the sample target period, as well as the load index threshold. The sample matching score value that can be used to train the cell pairing strategy prediction model is more in line with the matching score determination strategy in the actual application process, so as to improve the reliability of the trained cell pairing strategy prediction model in actual applications.

[0078] It should be noted that, in the embodiments of the present disclosure, the process of training the load index prediction model of the resource scheduling device can be the process of training the load index prediction model of the resource scheduling device in the above-mentioned embodiments, which is not described in detail in the embodiments of the present disclosure; the process of determining the sample matching score value of two cells in each cell group in the sample target time period based on the sample load index prediction value and the load index threshold associated with each load index of each cell in the sample target time period can refer to the process of determining the matching score value of two cells in each cell group in the target time period based on the target load index value and the load index threshold associated with each load index of each cell in the target time period in the above-mentioned embodiments, which is not described in detail in the embodiments of the present disclosure.

[0079] In an optional implementation, the process of training the strategy prediction model of the resource scheduling device may include: determining the sample matching score value of the two cells in each cell group in the sample target period based on the sample target load index value (i.e., load index label value) and load index threshold value associated with each load index in each cell in the sample target period; further, training the cell pairing strategy prediction model based on the sample matching score value of the two cells in each cell group in the sample target period. In the process of training the cell pairing strategy prediction model, the sample matching score value may be determined directly based on the sample target load index value associated with each load index in each cell in the sample target period, and used to train the cell pairing strategy prediction model, thereby reducing the amount of data processing and improving the training efficiency of the cell pairing strategy prediction model.

[0080] In an optional embodiment, if Figure 3 As shown, the resource scheduling device may include: training the cell pairing strategy prediction model based on the sample matching score values ​​of the two cells in each cell group in the sample target period;

[0081] Step S301 : constructing a sample matching score matrix through the sample matching score values ​​of two cells in each cell group during the sample target period.

[0082] In the embodiment of the present disclosure, the sample matching score value matrix is ​​a NumCell*NumCell dimensional matrix, where NumCell is the number of multiple cells. For example, when the number of multiple cells is 100, the sample matching score value matrix is ​​a 100*100 dimensional matrix.

[0083] It can be understood that, since a cell cannot match itself, the values ​​of the elements on the diagonal of the sample matching score matrix are 0.

[0084] Step S302, determining multiple cell groups as action spaces, and constructing a grid environment based on a sample matching score value matrix;

[0085] In the embodiment of the present disclosure, the grid environment includes a plurality of cell groups and a target matching score value corresponding to each cell group, and the target matching score value is determined based on the sample matching score value.

[0086] In an optional implementation, the process of constructing a grid environment based on a sample matching score matrix by a resource scheduling device may include: constructing a target grid with a side length equal to the number of cells, wherein a subgrid in the target grid represents a cell group; then, determining the target matching score values ​​of multiple subgrids on the diagonal of the target grid to be 0, further, obtaining the mean and variance of multiple sample matching score values ​​in the sample matching score matrix, and constructing a normal distribution of multiple sample matching score values ​​in the sample matching score matrix based on the mean and variance of multiple sample matching score values; and, for each non-diagonal target subgrid in the target grid, extracting a sample matching score value to be processed for the target subgrid in the normal distribution, and normalizing the sample matching score value to be processed to obtain a target matching score value, and finally, filling the target matching score value of each target word grid into the target word grid to obtain a grid environment. In the process of training a cell pairing strategy prediction model based on a reinforcement learning algorithm, a new grid environment can be constructed based on a sample matching score matrix to prevent the grid environment from being relatively single, improve the diversity of the grid environment, and improve the generalization ability of the trained sample matching score matrix.

[0087] Step S303, repeating the process of determining sample pairing cells for each cell in the action space from the current state of the grid environment, obtaining the sample cell pairing result of the current iteration process, the updated state of the grid environment, and the reward value corresponding to the updated state of the grid environment, until the iteration stop condition is met, and obtaining the first target sample data set and the second target sample data set.

[0088] In the disclosed embodiment, the first target sample data set includes the current state of the grid environment in each iteration process, the sample cell pairing result, the reward value obtained after determining the sample pairing cell for each cell, the reward value is the sample matching score value of the cell group composed of the cell and the sample pairing cell, and the updated state of the grid environment. The second target sample data set includes the updated state of the grid environment in each iteration process, the updated state of the grid environment is the state of the grid environment after determining the sample pairing cell for each cell in each iteration process, and the reward value corresponding to the updated state of the grid environment, the reward value is the sum of the reward values ​​obtained after determining the sample pairing cell for each cell.

[0089] Among them, the iteration stopping condition can be determined based on actual needs, and the embodiments of the present disclosure do not limit this. For example, the iteration stopping condition can be that the reward value obtained during multiple consecutive iterations is greater than the reward value threshold. The reward value threshold can be determined based on actual needs, and the embodiments of the present disclosure do not limit this.

[0090] In an optional implementation, the resource scheduling device determines a sample pairing cell for each cell in the action space from the current state of the grid environment, obtains the sample cell pairing result of the current iteration process, the updated state of the grid environment, and the reward value corresponding to the updated state of the grid environment, and the process may include: initializing the root node based on the current state of the grid environment; then, repeatedly searching for child nodes starting from the root node, and when it is determined that the current child node has not been expanded, expanding all child nodes of the current child node, and, based on the confidence interval upper bound algorithm, selecting the next child node from all child nodes of the current child node for search until the condition is satisfied. In the process of searching for termination conditions, multiple optional cell pairing results are obtained; further, the optional cell pairing result associated with the search path with the most visits in the search paths corresponding to the multiple optional cell pairing results is determined as the sample cell pairing result of the current iteration process; and the updated state of the grid environment associated with the matching result of the sample optional cell is determined as the updated state of the grid environment of the current iteration, and the total value of the sample matching score values ​​corresponding to each cell and the sample matching cell in the sample cell pairing result is determined in the sample matching score values ​​corresponding to the multiple cell groups, and the reward value corresponding to the updated state of the grid environment is obtained. The target sample data set can be collected based on the Monte Carlo tree search algorithm for training the cell pairing strategy prediction model, which can improve the randomness of the sample data in the collected target sample data set, and further improve the generalization ability of the trained cell pairing strategy prediction model.

[0091] It can be understood that in the embodiment of the present disclosure, the process of the resource scheduling device searching for child nodes is the process of determining a sample pairing cell for each cell in the action space, and the search termination condition may be that the action space is empty; determining a sample pairing cell for a cell can be understood as performing an action, wherein the reward for performing an action is the sample matching score value of the cell group consisting of the cell and the paired cell; the reward value of the current iteration is the total sample matching score value of the cell group consisting of the sample pairing cells determined for each cell after the current iteration is completed.

[0092] In which, in the action space, each cell can form a cell group with multiple other cells. In the process of determining the sample paired cell for each cell in the action space, the resource scheduling device searches for one candidate cell group in multiple candidate cell groups that include the target cell in the action space as a sample target cell group for each target cell for which a paired cell has not been determined. Then, another cell other than the target cell in the sample target cell group can be determined as the sample paired cell of the target cell; and other candidate cell groups that include the target cell are deleted from the action space to prevent multiple sample paired cells from being determined for the same cell.

[0093] Among them, the upper confidence bound (UCB) algorithm is:

[0094]

[0095] In Formula 2, ni is the number of times node Si is explored, Vi is the total reward obtained by exploring node Si, N is the total number of explorations, and c is a constant, which can generally be 2.

[0096] Optionally, after repeatedly obtaining the target sample data set (including the first target sample data set and the second target sample data set) for a specific number of iterations, the resource scheduling device may update the grid environment based on the method for constructing the grid environment provided in the above embodiment, and continue to obtain the target sample data set using the updated grid environment, which may further improve the generalization ability of the trained cell pairing strategy prediction model. The specific number of iterations may be set based on actual needs, and for example, the specific number of iterations may be 25 or 30 times.

[0097] Step S304, jointly iteratively training the cell pairing strategy prediction model based on the first target sample data set and the value network based on the second target sample data set, and obtaining the cell pairing strategy prediction model after the cell pairing strategy prediction model and the value network converge.

[0098] In the disclosed embodiment, the cell pairing strategy prediction model can be trained by a reinforcement learning algorithm. In the process of training the cell pairing strategy prediction model, the reward value is used as the model optimization target. Since the reward value is the sum of the reward values ​​obtained after determining the sample pairing cell for each cell, and the reward value is the sample matching score value of the cell group composed of the cell and the sample pairing cell, the matching score value can also characterize the load difference between the cells, thereby achieving the optimization target of maximizing resource utilization, so as to ensure that the trained cell pairing strategy can match each cell in multiple cells and obtain a cell pairing result that can maximize the utilization of resources of multiple cells.

[0099] It should be noted that in the embodiments of the present disclosure, the cell pairing strategy prediction model and the value network model can be neural network models respectively, and the specific model structure can be set based on actual needs, which is not limited in the embodiments of the present disclosure.

[0100] In an optional implementation, the resource scheduling device may perform joint iterative training on the cell pairing strategy prediction model based on the first target sample data set and the value network based on the second target sample data set, which may include: determining the current state of the grid environment obtained in each iteration process in the first target sample data set, the sample cell pairing result, the reward value obtained after determining the sample pairing cell for each cell, and the updated state of the grid environment as a first sample, and determining the updated state of the grid environment obtained in each iteration process in the second target sample data set, and the reward value corresponding to the updated state of the grid environment as a second sample; then, repeatedly inputting the current state of the grid environment in the first sample into the cell pairing strategy prediction model; A test model is used to obtain a predicted cell matching result, and the updated state of the grid environment in the second sample is input into the value model to obtain a predicted reward value corresponding to the updated state of the grid environment; then, based on the predicted cell matching result, the predicted reward value corresponding to the updated state of the grid environment, the sample cell pairing result in the first sample, a reward value obtained after the sample pairing cell is determined for each cell, the updated state of the grid environment, the reward value corresponding to the updated state of the grid environment in the second sample, and the process of determining the loss function value by the loss function, when the loss function value is less than the loss function threshold, it is determined that the cell pairing strategy prediction model and the value network converge, wherein the loss function threshold can be determined based on actual needs, and the embodiments of the present disclosure are not limited to this.

[0101] Among them, the loss function is:

[0102]

[0103] In Formula 3, k is the number of times the cell pairing strategy prediction model and the value network are repeatedly trained, P_mcts_i is the probability distribution of the sample cell pairing results in the first sample, P_nn_i is the probability distribution of the predicted cell matching results, value_nn_i is the predicted reward value corresponding to the updated state of the grid environment, and value_mcts_i is the reward value corresponding to the updated state of the grid environment, wherein the probability distribution is based on the reward value obtained after determining the sample pairing cell for each cell in the first sample, and the updated state of the grid environment.

[0104] It can be understood that, in each process of repeating the training of the cell pairing strategy prediction model and the value network, the first sample and the second sample used are sample data obtained in the same iteration process of the above step S303.

[0105] In an optional implementation, the resource sculpture device can also determine the sample reference cell pairing result of the current iteration in the replica grid of the grid environment based on the greedy strategy; then, in the sample matching score values ​​corresponding to the multiple cell groups, determine the total sample reference matching score corresponding to each cell and the sample reference paired cell in the sample reference cell pairing result to obtain the reference return value of the current iteration; further, determine the difference between the return value and the reference return value to obtain the return value difference, and determine the return value difference and the return value of the current iteration as the updated return value of the current iteration. The cell pairing strategy prediction model can be optimized and trained based on the sample reference cell pairing result determined by the greedy strategy, so that the trained cell pairing strategy prediction model can determine the cell pairing result that optimizes resource utilization in actual applications.

[0106] In an optional implementation, the resource scheduling device determines the cell pairing result based on a matching score value set consisting of matching score values ​​of two cells in each cell group during a target time period and a cell pairing strategy prediction model. The process may include: constructing a matching score value matrix through the matching score value set; then, constructing the current grid environment based on the matching score value matrix; further, inputting the current state of the current grid environment into the cell pairing strategy prediction model to obtain the cell pairing result.

[0107] It should be noted that the process of constructing a matching score value matrix by the resource scheduling device is similar to the process of constructing a sample matching score value matrix, and the process of constructing the current grid environment based on the matching score value matrix is ​​similar to the process of constructing the current grid environment based on the matching score value matrix. The present embodiment of the disclosure will not elaborate on this.

[0108] In an optional implementation, before performing resource scheduling in a target time period based on the cell pairing result, the resource scheduling device may also: search for a target matching score value associated with each cell and the paired cell in the cell pairing result in the matching score value set, and determine a sum of multiple target matching score values ​​to obtain a cell pairing result score value associated with the cell pairing result; then, determine a reference cell pairing result based on a greedy strategy and the matching score value set, wherein the reference cell pairing result includes a reference paired cell determined for each cell; and search for a reference matching score value associated with each cell and the reference paired cell in the reference cell pairing result in the matching score value set, and determine a sum of multiple reference matching score values ​​to obtain a reference cell pairing result score value associated with the reference cell pairing result; further, if the reference cell pairing result score value is greater than the cell pairing result score value, the reference cell pairing result is determined as the updated cell pairing result. A reference cell pairing result can be determined based on a greedy strategy. After determining that the reference cell pairing result is better than the cell pairing result determined based on the cell pairing strategy prediction model, the reference cell pairing result is determined as the final cell pairing result, so as to facilitate a cell matching result that maximizes resource utilization.

[0109] In an optional implementation, the process of performing resource scheduling during a target time period by the resource scheduling device based on the cell pairing result may include: during the target time period, for each cell, according to the paired cell determined for the cell in the cell pairing result, when the cell load is greater than the paired cell, providing the resources of the paired cell to the cell.

[0110] An exemplary embodiment of the present disclosure provides a resource scheduling device, which may be a server or a chip applied to a server. Figure 4 FIG. 2 shows a schematic block diagram of functional modules of a resource scheduling device according to an exemplary embodiment of the present disclosure. Figure 4 As shown, the resource scheduling device 400 includes:

[0111] The acquisition module 401 is configured to acquire a plurality of historical load index values ​​respectively associated with a plurality of load indexes in a historical period for a plurality of cells;

[0112] The first determination module 402 is configured to input a plurality of historical load index values ​​associated with each load index of each cell in a historical period into a load index prediction model, respectively, to obtain a target load index value associated with each load index of each cell in a target period;

[0113] The second determination module 403 is configured to determine the matching score value of two cells in each cell group in the target time period based on the target load index value and the load index threshold value associated with each load index of each cell in the target time period, wherein the multiple cell groups are obtained after matching two cells in the multiple cells, and the matching score value is used to characterize the degree of load difference between the two cells;

[0114] The resource scheduling module 404 is configured to determine a cell pairing result based on a matching score value set composed of matching score values ​​of the two cells in each cell group during the target time period and a cell pairing strategy prediction model, and perform resource scheduling during the target time period based on the cell pairing result, wherein the cell pairing result includes a paired cell determined for each cell for resource scheduling.

[0115] Optionally, the second determining module 403 is configured to:

[0116] For each cell group, according to the load index threshold value associated with each load index and the comparison result of the target load index value associated with each load index of the first cell in the cell group during the target period, a first modified target load index value associated with each load index of the first cell during the target period is obtained;

[0117] According to the load index threshold value associated with each load index and the comparison result of the target load index value associated with each load index of the second cell in the cell group within the target time period, a second modified target load index value associated with each load index of the second cell within the target time period is obtained;

[0118] Determine a first modified target load index value and a second modified target load index value associated with each load index in a target period for the first cell and the second cell in each cell group, and obtain a load difference amount associated with each load index in the target period for the first cell and the second cell in each cell group;

[0119] The load difference amounts respectively associated with multiple load indicators of the first cell and the second cell in each cell group during the target period are weighted summed to obtain the matching score value of the two cells in each cell group during the target period.

[0120] Optionally, the device further includes an updating module 405 configured to:

[0121] In the matching score value set, searching for a target matching score value associated with each cell and the paired cell in the cell pairing result, and determining a sum of a plurality of the target matching score values ​​to obtain a cell pairing result score value associated with the cell pairing result;

[0122] Determine a reference cell pairing result based on a greedy strategy and the matching score value set, wherein the reference cell pairing result includes a reference paired cell determined for each cell;

[0123] In the matching score value set, searching for a reference matching score value associated with each cell in the reference cell matching result and the reference matching cell, and determining a sum of a plurality of the reference matching score values ​​to obtain a reference cell matching result score value associated with the reference cell matching result;

[0124] If the reference cell pairing result score value is greater than the cell pairing result score value, the reference cell pairing result is determined as the updated cell pairing result.

[0125] Optionally, the device further includes a model training module 406 configured to:

[0126] Acquire a sample data set, wherein the sample data set includes a plurality of sample historical load index values ​​respectively associated with the plurality of load indexes for each cell in a sample historical period, and a load index label value respectively associated with the plurality of load indexes for each cell in a sample target period;

[0127] Using the sample data set to train a time series network, after the time series network converges, the load index prediction model is obtained;

[0128] Inputting a plurality of sample historical load index values ​​associated with each load index of each cell in the sample historical period into the load index prediction model respectively, and obtaining a sample load index prediction value associated with each load index of each cell in the sample target period;

[0129] Based on the sample load index prediction value and load index threshold associated with each load index of each cell during the sample target time period, the sample matching score value of the two cells in each cell group during the sample target time period is determined, and based on the sample matching score value of the two cells in each cell group during the sample target time period, the cell pairing strategy prediction model is trained.

[0130] Optionally, the model training module 406 is configured to:

[0131] Constructing a sample matching score value matrix according to the sample matching score values ​​of the two cells in each cell group during the sample target period;

[0132] Determine a plurality of cell groups as an action space, and construct a grid environment based on the sample matching score value matrix, wherein the grid environment includes the plurality of cell groups and a target matching score value corresponding to each cell group, and the target matching score value is determined based on the sample matching score value;

[0133] Repeating the process of determining a sample pairing cell for each cell in the action space from the current state of the grid environment, obtaining the sample cell pairing result of the current iteration process, the updated state of the grid environment, and the reward value corresponding to the updated state of the grid environment, until the iteration stop condition is met, and obtaining a first target sample data set and a second target sample data set, wherein the first target sample data set includes the current state of the grid environment in each iteration process, the sample cell pairing result, the reward value obtained after determining the sample pairing cell for each cell, and the updated state of the grid environment, and the second target sample data set includes the updated state of the grid environment in each iteration process, and the reward value corresponding to the updated state of the grid environment;

[0134] The cell pairing strategy prediction model is jointly iteratively trained based on the first target sample data set, and the value network is jointly iteratively trained based on the second target sample data set, and the cell pairing strategy prediction model is obtained after the cell pairing strategy prediction model and the value network converge.

[0135] Optionally, the model training module 406 is configured to:

[0136] Initializing a root node based on a current state of the grid environment;

[0137] Repeating the process of searching for child nodes from the root node, and when determining that the current child node has not been expanded, expanding all child nodes of the current child node, and, based on a confidence interval upper bound algorithm, selecting a next child node from all child nodes of the current child node for searching until a search termination condition is met, to obtain multiple optional cell pairing results;

[0138] Determine the optional cell pairing result associated with the search path with the most visits among the search paths corresponding to the multiple optional cell pairing results as the sample cell pairing result of the current iteration process;

[0139] The updated state of the grid environment associated with the matching result of the sample optional cell is determined as the updated state of the grid environment of the current iteration, and the total value of the sample matching score values ​​corresponding to each cell and the sample paired cell in the sample cell pairing result is determined based on the sample matching score values ​​corresponding to multiple cell groups, so as to obtain a reward value corresponding to the updated state of the grid environment.

[0140] Optionally, the model training module 406 is further configured to:

[0141] Determining a sample reference cell pairing result of a current iteration in a replica grid of the grid environment based on a greedy strategy;

[0142] Determine the total value of the sample reference matching scores corresponding to each cell and the sample reference paired cell in the sample reference cell pairing result among the sample matching score values ​​corresponding to the multiple cell groups, and obtain the reference return value of the current iteration;

[0143] The difference between the reward value and the reference reward value is determined to obtain a reward value difference, and the reward value difference and the reward value of the current iteration are determined as the updated reward value of the current iteration.

[0144] The exemplary embodiment of the present disclosure also provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor in communication. The memory stores a computer program that can be executed by the at least one processor, and the computer program is used to cause the electronic device to perform the method according to the embodiment of the present disclosure when executed by the at least one processor.

[0145] The exemplary embodiments of the present disclosure also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to perform the method according to the embodiments of the present disclosure.

[0146] like Figure 5 As shown, the exemplary embodiment of the present disclosure further provides a computer program product 500, including a computer program 501, wherein the computer program 501 is used to enable the computer to perform the method according to the embodiment of the present disclosure when being executed by a processor of a computer.

[0147] refer to Figure 6, a block diagram of an electronic device 600 that can be used as a server or client of the present disclosure will now be described, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0148] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0149] A plurality of components in the electronic device 600 are connected to the I / O interface 605, including: an input unit 606, an output unit 607, a storage unit 608, and a communication unit 609. The input unit 606 may be any type of device capable of inputting information to the electronic device 600, and the input unit 606 may receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 607 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 608 may include, but is not limited to, a disk, an optical disk. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth™ device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0150] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 601 performs the various methods and processes described above. For example, in some embodiments, the method of the exemplary embodiment of the present disclosure may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 600 via the ROM 602 and / or the communication unit 609. In some embodiments, the computing unit 601 may be configured to perform the method of the exemplary embodiment of the present disclosure by any other appropriate means (e.g., by means of firmware).

[0151] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0152] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0153] As used in this disclosure, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0154] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0155] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0156] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0157] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instruction is loaded and executed on a computer, the process or function described in the embodiment of the present disclosure is executed in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, a terminal, a user device or other programmable device. The computer program or instruction may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer program or instruction may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server, data center, etc. that integrates one or more available media. The available medium may be a magnetic medium, for example, a floppy disk, a hard disk, a tape; it may also be an optical medium, for example, a digital video disc (DVD); it may also be a semiconductor medium, for example, a solid state drive (SSD).

[0158] Although the present disclosure has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present disclosure. Accordingly, this specification and the drawings are merely exemplary illustrations of the present disclosure as defined by the appended claims, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present disclosure. Obviously, those skilled in the art may make various modifications and variations to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A resource scheduling method, characterized in that: include: Acquire multiple historical load index values ​​respectively associated with multiple load indexes in multiple cells within a historical period; Inputting a plurality of historical load index values ​​associated with each load index in each cell in a historical period into a load index prediction model respectively, to obtain a target load index value associated with each load index in each cell in a target period; Based on the target load index value and the load index threshold value associated with each load index of each cell in the target time period, determining the matching score value of two cells in each cell group in the target time period, wherein the multiple cell groups are obtained after matching two cells in the multiple cells, and the matching score value is used to characterize the degree of load difference between the two cells; Based on a matching score value set consisting of matching score values ​​of the two cells in each cell group in the target time period and a cell pairing strategy prediction model, a cell pairing result is determined, and based on the cell pairing result, resource scheduling is performed in the target time period, wherein the cell pairing result includes a paired cell determined for each cell for resource scheduling.

2. The resource scheduling method according to claim 1, characterized in that: Determining a matching score value of two cells in each cell group in the target time period based on a target load index value and a load index threshold value associated with each load index in each cell in the target time period includes: For each cell group, according to the load index threshold value associated with each load index and the comparison result of the target load index value associated with each load index of the first cell in the cell group during the target period, a first modified target load index value associated with each load index of the first cell during the target period is obtained; According to the load index threshold value associated with each load index and the comparison result of the target load index value associated with each load index of the second cell in the cell group within the target time period, a second modified target load index value associated with each load index of the second cell within the target time period is obtained; Determine a first modified target load index value and a second modified target load index value associated with each load index in a target period for the first cell and the second cell in each cell group, and obtain a load difference amount associated with each load index in the target period for the first cell and the second cell in each cell group; The load difference amounts respectively associated with multiple load indicators of the first cell and the second cell in each cell group during the target period are weighted summed to obtain the matching score value of the two cells in each cell group during the target period.

3. The resource scheduling method according to claim 1, characterized in that: Before performing resource scheduling in a target time period based on the cell pairing result, the method further includes: In the matching score value set, searching for a target matching score value associated with each cell and the paired cell in the cell pairing result, and determining a sum of a plurality of the target matching score values ​​to obtain a cell pairing result score value associated with the cell pairing result; Determine a reference cell pairing result based on a greedy strategy and the matching score value set, wherein the reference cell pairing result includes a reference paired cell determined for each cell; In the matching score value set, searching for a reference matching score value associated with each cell in the reference cell matching result and the reference matching cell, and determining a sum of a plurality of the reference matching score values ​​to obtain a reference cell matching result score value associated with the reference cell matching result; If the reference cell pairing result score value is greater than the cell pairing result score value, the reference cell pairing result is determined as the updated cell pairing result.

4. The resource scheduling method according to claim 1, characterized in that: The method further comprises: Acquire a sample data set, wherein the sample data set includes a plurality of sample historical load index values ​​respectively associated with the plurality of load indexes for each cell in a sample historical period, and a load index label value respectively associated with the plurality of load indexes for each cell in a sample target period; Using the sample data set to train a time series network, after the time series network converges, the load index prediction model is obtained; Inputting a plurality of sample historical load index values ​​associated with each load index of each cell in the sample historical period into the load index prediction model respectively, and obtaining a sample load index prediction value associated with each load index of each cell in the sample target period; Based on the sample load index prediction value and load index threshold associated with each load index of each cell during the sample target time period, the sample matching score value of the two cells in each cell group during the sample target time period is determined, and based on the sample matching score value of the two cells in each cell group during the sample target time period, the cell pairing strategy prediction model is trained.

5. The resource scheduling method according to claim 4, characterized in that: The training of the cell pairing strategy prediction model based on the sample matching score values ​​of the two cells in each cell group in the sample target time period includes: Constructing a sample matching score value matrix according to the sample matching score values ​​of the two cells in each cell group during the sample target period; Determine a plurality of cell groups as an action space, and construct a grid environment based on the sample matching score value matrix, wherein the grid environment includes the plurality of cell groups and a target matching score value corresponding to each cell group, and the target matching score value is determined based on the sample matching score value; Repeating the process of determining a sample pairing cell for each cell in the action space from the current state of the grid environment, obtaining the sample cell pairing result of the current iteration process, the updated state of the grid environment, and the reward value corresponding to the updated state of the grid environment, until the iteration stop condition is met, and obtaining a first target sample data set and a second target sample data set, wherein the first target sample data set includes the current state of the grid environment in each iteration process, the sample cell pairing result, the reward value obtained after determining the sample pairing cell for each cell, and the updated state of the grid environment, and the second target sample data set includes the updated state of the grid environment in each iteration process, and the reward value corresponding to the updated state of the grid environment; The cell pairing strategy prediction model is jointly iteratively trained based on the first target sample data set, and the value network is jointly iteratively trained based on the second target sample data set, and the cell pairing strategy prediction model is obtained after the cell pairing strategy prediction model and the value network converge.

6. The resource scheduling method according to claim 5, characterized in that: The step of determining a sample pairing cell for each cell in the action space from the current state of the grid environment, obtaining a sample cell pairing result of the current iteration process, an updated state of the grid environment, and a reward value corresponding to the updated state of the grid environment includes: Initializing a root node based on a current state of the grid environment; Repeating the process of searching for child nodes from the root node, and when determining that the current child node has not been expanded, expanding all child nodes of the current child node, and, based on a confidence interval upper bound algorithm, selecting a next child node from all child nodes of the current child node for searching until a search termination condition is met, to obtain multiple optional cell pairing results; Determine the optional cell pairing result associated with the search path with the most visits among the search paths corresponding to the multiple optional cell pairing results as the sample cell pairing result of the current iteration process; The updated state of the grid environment associated with the matching result of the sample optional cell is determined as the updated state of the grid environment of the current iteration, and the total value of the sample matching score values ​​corresponding to each cell and the sample paired cell in the sample cell pairing result is determined based on the sample matching score values ​​corresponding to multiple cell groups, so as to obtain a reward value corresponding to the updated state of the grid environment.

7. The resource scheduling method according to claim 6, characterized in that: The method further comprises: Determining a sample reference cell pairing result of a current iteration in a replica grid of the grid environment based on a greedy strategy; Determine the total value of the sample reference matching scores corresponding to each cell and the sample reference paired cell in the sample reference cell pairing result among the sample matching score values ​​corresponding to the multiple cell groups, and obtain the reference return value of the current iteration; The difference between the reward value and the reference reward value is determined to obtain a reward value difference, and the reward value difference and the reward value of the current iteration are determined as the updated reward value of the current iteration.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the resource scheduling method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the resource scheduling method described in any one of claims 1 to 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the resource scheduling method described in any one of claims 1 to 7 is implemented.

Images