Memory scheduling method, device and readable storage medium

Abstract

An embodiment of the present invention provides a memory scheduling method, device and readable storage medium. In the method embodiment of the present invention, based on the main cause corresponding to the degraded cell work order, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined, wherein the main cause corresponding to the degraded cell work order refers to the main reason causing the degradation of the network performance indicator corresponding to the degraded cell. Based on the global priority scheduling order of the data to be accessed corresponding to each main cause, the memory data is scheduled. By determining the global priority scheduling order of the data to be accessed corresponding to each main cause based on the main cause corresponding to the degraded cell work order, the memory data is scheduled based on the global priority scheduling order of the data to be accessed corresponding to each main cause. This can not only reduce the memory space for caching irrelevant data, but also improve the hit rate of memory data.

Description

Technical Field

[0001] The present invention relates to the field of IT applications, and in particular to a memory scheduling method, device and readable storage medium. Background Art

[0002] The intelligent centralized optimization platform provides support for the centralized optimization work of provinces and cities, handles problems in work orders for degraded cells, improves network quality, enables optimization personnel to analyze wireless network problems in the cells, completes the execution of some solutions in daily network optimization and special optimization, and completes the handling of work orders for degraded cells.

[0003] The intelligent centralized optimization platform is a big data analysis platform. Due to the large number of data sources involved and the large amount of data, the original analysis data is usually stored on the Hadoop / Hbase (Hadoop database) big data platform, and the analysis result data is usually stored in the MPP (Massively Parallel Processing) database, such as the popular GreenPlum database or Gbase database. When users query the system analysis results, they usually extract the analysis data from the MPP database through lightweight J2EE system frameworks such as SSM (spring MVC, spring, mybatis), and then use the front-end framework to render the data into multimedia pages such as tables or maps for users to query.

[0004] In order to assist network optimization personnel in diagnosing network problems, the system needs to extract analysis data from the database multiple times for page display. Due to the large amount of analysis result data, time-consuming database search, network transmission speed and other factors, the page query data speed is usually slow. The single page response time is sometimes more than 30 seconds, which seriously affects the efficiency of network problem diagnosis.

[0005] Existing memory scheduling methods basically store recent data in memory. Common memory management strategies include FIFO (First Input First Output), LRU (Least Recently Used), random strategy, etc. FIFO means that when the memory usage reaches a certain threshold, the data that first enters the memory will be removed, and LRU means that the data that has been least used in the recent period will be removed. The random strategy randomly selects a part of the data to remove from the memory, thereby freeing up memory and transferring the data required by the system into the memory.

[0006] Although the above memory management strategy can speed up the system response to a certain extent, large-scale data analysis systems such as intelligent centralized optimization platforms usually need to cache 7 days of full network analysis data. If traditional memory scheduling methods such as FIFO, LRU, and random scheduling are used, the cache amount required is huge. In addition, since some analysis dimension data requires a full display of 7 days of indicators, using FIFO to remove the first day's data from the memory will quickly need to be loaded back into the memory, causing cache data jitter. Therefore, the above memory scheduling method not only requires a considerable amount of physical memory, but also cannot efficiently call in and out data in the centralized network optimization system. Summary of the invention

[0007] In view of the above technical problems existing in the prior art, an embodiment of the present invention provides a memory scheduling method, device and readable storage medium.

[0008] In a first aspect, an embodiment of the present invention provides a memory scheduling method, comprising:

[0009] Based on the main causes corresponding to the degraded cell work orders, determine the global priority scheduling order of the to-be-accessed data corresponding to the main causes, wherein the main causes corresponding to the degraded cell work orders refer to the main causes that cause the degradation of the network performance indicators corresponding to the degraded cells;

[0010] The memory data is scheduled based on the global priority scheduling order of the to-be-accessed data corresponding to the main causes.

[0011] Optionally, according to a memory scheduling method of an embodiment of the present invention, determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order specifically includes:

[0012] Based on the main causes corresponding to the degraded cell work order, determine the to-be-accessed data corresponding to each main cause;

[0013] Determine the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period;

[0014] Among them, the main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on the sub-factors corresponding to the degraded cell work order and their influence weights, wherein the influence weight is used to characterize the degree of influence of the sub-factors corresponding to the degraded cell work order on the cell network performance indicators.

[0015] Optionally, according to a memory scheduling method of an embodiment of the present invention, determining the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period specifically includes:

[0016] Based on the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined.

[0017] Optionally, according to a memory scheduling method of an embodiment of the present invention, the scheduling of memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each main cause specifically includes:

[0018] Based on the global priority scheduling order of the data to be accessed corresponding to the main reasons, the data to be accessed are loaded into the memory in descending order of priority, and / or removed from the memory in descending order of priority.

[0019] Optionally, according to a memory scheduling method of an embodiment of the present invention, the impact weight is determined based on a preset correction coefficient and probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators, and the preset correction coefficient is determined based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators.

[0020] Optionally, according to the memory scheduling method of an embodiment of the present invention, after determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause, the method further includes:

[0021] Based on the access frequency of the data to be accessed corresponding to each main cause within a preset time period after the global priority scheduling order of the data to be accessed corresponding to each main cause is determined, the global priority scheduling order of the data to be accessed corresponding to each main cause is updated.

[0022] Optionally, according to the memory scheduling method of an embodiment of the present invention, the conditional probability that the main cause corresponding to the historical degraded cell work order causes the degradation of the cell network performance index is calculated by the Bayesian formula based on the ratio of the cell network performance index degradation caused by the main cause in all historical degraded cell work orders, the ratio of degraded cell work orders caused by the main cause in all historical degraded cell work orders, and the ratio of the cell network performance index degradation in all historical degraded cell work orders.

[0023] In a second aspect, an embodiment of the present invention further provides a memory scheduling device, including:

[0024] A priority scheduling order determination module is used to determine the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order, wherein the main cause corresponding to the degraded cell work order refers to the main cause causing the degradation of the network performance indicator corresponding to the degraded cell;

[0025] The memory scheduling module is used to schedule the memory data based on the global priority scheduling order of the to-be-accessed data corresponding to the main causes.

[0026] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method provided in the first aspect are implemented.

[0027] In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the first aspect above.

[0028] The memory scheduling method, device and readable storage medium provided in the embodiments of the present invention determine the global priority scheduling order of the data to be accessed corresponding to each main cause based on the main cause corresponding to the degraded cell work order, and schedule the memory data based on the global priority scheduling order of the data to be accessed corresponding to each main cause, which can not only reduce the memory space for caching irrelevant data, but also improve the memory data hit rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 It is a flowchart of a memory scheduling method provided by an embodiment of the present invention;

[0031] Figure 2 It is a detailed flow chart of a memory scheduling method provided by an embodiment of the present invention;

[0032] Figure 3 It is a structural schematic diagram of a memory scheduling device provided by an embodiment of the present invention;

[0033] Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] The existing memory scheduling method not only requires a relatively large physical memory, but also cannot efficiently load and unload data. To this end, an embodiment of the present invention provides a memory scheduling method. Figure 1 A flowchart of a memory scheduling method provided by an embodiment of the present invention is shown as follows: Figure 1 As shown, the method includes:

[0036] Step 110, based on the main causes corresponding to the degraded cell work order, determine the global priority scheduling order of the to-be-accessed data corresponding to each main cause, wherein the main cause corresponding to the degraded cell work order refers to the main cause causing the degradation of the network performance indicator corresponding to the degraded cell.

[0037] Specifically, in the centralized network optimization process, the degraded cell work order corresponds to multiple main causes, that is, there is more than one main cause that causes the degradation of network performance indicators. The main cause corresponding to the degraded cell work order is determined from multiple sub-factors, and the sub-factors are all possible causes that may lead to the degradation of cell network performance indicators based on existing experience, including fault alarms, coverage, interference, resource usage, parameter configuration and modification, neighboring cells and other dimensions.

[0038] Network optimization engineers usually use the main causes output by the centralized network optimization system as the main basis for diagnosing network problems to query the corresponding analysis data, that is, the data to be accessed. Based on this, the global priority scheduling order of the data to be accessed corresponding to each main cause can be determined according to the main cause corresponding to the degraded cell work order, so that network optimization engineers can access it.

[0039] Step 120 , scheduling the memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each of the main causes.

[0040] Specifically, the memory scheduling device schedules the memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each main cause, and stores or removes the to-be-accessed data corresponding to each main cause in the memory according to the priority scheduling order, thereby avoiding the caching of irrelevant data and enabling network optimization engineers to quickly access data related to wireless network problems. The to-be-accessed data corresponding to each main cause can be stored in the form of a data table for easy access. Of course, it can also be stored in other forms according to actual needs, and the embodiment of the present invention does not specifically limit this.

[0041] The method provided by an embodiment of the present invention determines the global priority scheduling order of the data to be accessed corresponding to each main cause based on the main cause corresponding to the degraded cell work order, and schedules the memory data based on the global priority scheduling order of the data to be accessed corresponding to each main cause, which can not only reduce the memory space for caching irrelevant data, but also improve the memory data hit rate.

[0042] Based on the above embodiment, the method of determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order specifically includes:

[0043] Based on the main causes corresponding to the degraded cell work order, determine the to-be-accessed data corresponding to each main cause;

[0044] Determine the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period;

[0045] Among them, the main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on the sub-factors corresponding to the degraded cell work order and their influence weights, wherein the influence weight is used to characterize the degree of influence of the sub-factors corresponding to the degraded cell work order on the cell network performance indicators.

[0046] Specifically, the data to be accessed corresponding to each sub-factor is pre-stored in the database in the form of a data table. Table 1 shows the three-level data constructed according to page-data extraction URL-data table involved in each URL. The table lists the data to be accessed corresponding to the four sub-factors of coverage, neighboring area, resource, and parameter. However, it can be understood that in the actual database, the data to be accessed is not limited to the data to be accessed corresponding to the above four sub-factors, but includes the data to be accessed corresponding to all sub-factors.

[0047] Table 1 Examples of data to be accessed corresponding to sub-factors

[0048]

[0049]

[0050] The memory scheduling device can determine the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order. The main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on the sub-factors corresponding to the degraded cell work order and their influence weights. The greater the influence weight, the higher the degree of influence of the corresponding sub-factor on the cell network performance index. The main cause corresponding to the degraded cell work order and the priority of the main cause can be determined based on the order of the influence weights corresponding to the sub-factors. The greater the influence weight, the higher the priority. As for the number of main causes, it can be determined in advance based on experience, that is, the first few sub-factors with the highest priority are taken as the main causes.

[0051] In order to ensure that the data to be accessed that is loaded into the memory is the data that the network optimization engineer actually wants to access, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within the preset time length. According to experience, the higher the priority of the main cause corresponding to the degraded cell work order, the higher the probability that the corresponding data to be accessed will be accessed, and the priority order of each data to be accessed corresponding to a certain main cause can be obtained according to the historical access frequency. Based on this, combined with the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within the preset time length, the global priority scheduling order of the data to be accessed corresponding to each main cause can be determined. As for the time length corresponding to the preset time length, it can be set according to the experience of the network optimization engineer, and the embodiment of the present invention does not make specific limitations on this.

[0052] The method provided by an embodiment of the present invention determines the data to be accessed corresponding to each main cause through the main cause corresponding to the degraded cell work order, and determines the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period. This can ensure the memory data hit rate on the basis of reducing the memory space for caching irrelevant data.

[0053] Based on the above embodiment, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period, specifically including:

[0054] Based on the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined.

[0055] Specifically, according to historical experience, corresponding weights are assigned to the priority of the main cause and the historical access frequency of the data to be accessed, so as to further ensure the accuracy of the global priority scheduling order of the data to be accessed corresponding to each main cause, thereby ensuring the memory data hit rate. Assuming that the main causes corresponding to the degraded cell work order obtained by the memory scheduling device are 4, the frequency of access of the corresponding data table to be accessed by the network optimization engineer within a period of time can be determined according to the main cause, and the global priority scheduling order of the data to be accessed corresponding to each main cause is determined according to the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency. As shown in Table 2, it is an example of the global priority scheduling order of the data to be accessed determined according to the 4 main causes of coverage-neighborhood-resource-parameter in descending order of priority.

[0056] Table 2 Example of global priority scheduling order

[0057]

[0058]

[0059] The priority of each data table to be accessed is equal to C (main cause priority weight) * main cause priority + D (access frequency weight) * table access frequency statistics (C and D can be adjusted. In the above table, in order to illustrate the calculation process, C is 100 and D is 1. Since there are 4 main causes, the main cause with the highest priority has a priority of 4, and the priorities of other main causes decrease in sequence with a span of 1). In actual calculations, C takes the total number of data tables corresponding to all sub-factors, and D can be set to the total number of data tables divided by the total number of data table accesses to ensure sufficient priority differentiation. The larger the value corresponding to the scheduling order in Table 2, the higher the global priority of the corresponding data to be accessed.

[0060] The method provided in an embodiment of the present invention determines the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency, thereby ensuring the accuracy of the global priority scheduling order of the data to be accessed corresponding to each main cause, thereby ensuring the memory data hit rate.

[0061] Based on the above embodiment, the scheduling of memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each main cause specifically includes:

[0062] Based on the global priority scheduling order of the data to be accessed corresponding to the main reasons, the data to be accessed are loaded into the memory in descending order of priority, and / or removed from the memory in descending order of priority.

[0063] Specifically, assuming that the total number of sub-factors is 6, and based on the pre-setting of 4 main factors corresponding to the degraded cell work order, the memory scheduling device can generate a global priority scheduling order table of the to-be-accessed data corresponding to each main factor sorting scenario according to the sorting of the 4 main factors in advance. Because there are 6 dimensions of sub-factors, the global data priority scheduling order table corresponding to the 4 main factors has a total of 6*5*4*3=360 scenario directories, and each directory is identified by the main factor scenario corresponding to R1-R2-R3-R4, such as coverage-resources-neighborhood-parameters. Table 3 shows an example of a global priority scheduling order table corresponding to the main factor scenario.

[0064] Table 3 Example of global priority scheduling order corresponding to the main cause scenario

[0065]

[0066] When the memory scheduling device is started, the memory management program scans all work orders, and searches the global data priority scheduling order table according to the four dimensions involved in the main cause of each work order and the R1-R2-R3-R4 main cause scenario identifiers to obtain the global data priority scheduling order table corresponding to the main cause scenario, and then loads the required tables for the work order into the memory in sequence according to the priority in the table until the memory usage reaches the usage threshold.

[0067] When the data accessed by the network optimization engineer is not in the memory, the data tables involved in the dispatched work orders are first moved out of the memory, and then the data tables with lower priorities for the undispatched work orders are moved out of the memory, such as moving tables with a priority less than 200 out of the memory, or moving tables with a priority less than 320 out of the memory, etc., so as to free up enough memory to load the data from the hard disk into the memory.

[0068] The method provided by an embodiment of the present invention can improve the memory data hit rate on the basis of reducing the memory space of cache irrelevant data by loading the data to be accessed into the memory in descending order of priority based on the global priority scheduling order of the data to be accessed corresponding to the main causes, and / or removing the data to be accessed from the memory in descending order of priority.

[0069] Based on the above embodiment, the impact weight is determined based on a preset correction coefficient and the probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators. The preset correction coefficient is determined based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators.

[0070] Specifically, the sub-factor corresponding to the degraded cell work order is denoted as R. In view of the fact that different sub-factors R have different degradation effects on different cell network performance indicators, in the actual operation and maintenance environment, the cell network performance indicators used as the evaluation benchmark can be predetermined. The cell network performance indicators can be denoted as K, and the impact weights are used to characterize the degree of influence of the above-mentioned degradation effects. The impact weights are determined based on a preset correction coefficient and the probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators. The preset correction coefficient is determined based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators. The main cause can be determined based on the sub-factor R and its corresponding impact weight. For example Figure 2 The figure shows a detailed flow chart of a memory scheduling method provided by an embodiment of the present invention. Figure 2 The process of determining the main cause is introduced in detail:

[0071] 1) Quantify and standardize the degradation degree of indicator K

[0072] Network performance KPI is usually used to reflect network quality. If the network quality deteriorates, the system can obtain the sampling time series of each network performance indicator according to the time series (such as sampling indicators at hourly granularity during the degradation period), and then standardize it according to the following formula, and assign the degree of degradation according to the standardized degradation range:

[0073]

[0074] Among them, K jn is the indicator value in the hourly time period, K is the problem threshold value, K0 is the value without any problem, after normalizing multiple indicator sampling sequences, the assignment matrix of K can be obtained as

[0075]

[0076] 2) Quantify the sub-factor R:

[0077] According to the empirical formula, the sub-factor R that causes network index degradation can be calculated and quantified, and hourly values ​​can be assigned according to severity to obtain a degradation time assignment table for sub-factor R, and whether there is a problem can be marked;

[0078] The elements of the quality difference dimension are divided into three categories. The first type of quality difference dimension presents a switch feature in a heartbeat cycle, that is, there are two states: there is a problem with this dimension and there is no problem with this dimension. The elements in this type of quality difference dimension include: alarm, parameter, and neighboring area. The second type of quality difference dimension presents a three-state feature, including severe quality difference, weak quality difference, and no problem. The elements of this dimension include resources, interference, and coverage; the third type of quality difference dimension presents a one-click judgment feature, that is, this dimension is a single feature within the degradation cycle, with or without problems.

[0079] According to the above R quality difference dimension feature states, the R dimension hourly value is assigned (the three-state value assignment rule of the R quality difference dimension is: severe quality difference is assigned to 4, weak quality difference is assigned to 2, and no problem is assigned to 0. The switch feature and one-key judgment feature assignment are: problematic is assigned to 2, and no problem is assigned to 0);

[0080] By calculating statistics, the assignment matrix of N-hour granularity within the degradation period T is obtained;

[0081]

[0082] 3) Screening of correlation matrix C:

[0083] By counting a large amount of work order data and expert experience, we count whether the sub-factors are related to the indicators, that is, whether the sub-factors will cause the indicators to deteriorate. If there is a correlation, it is marked as 1, and if there is no correlation, it is marked as 0. We get the matrix list C of whether the indicator R affects K, as shown below:

[0084]

[0085] The sub-factor R is associated with the correlation coefficient matrix C, and the correlation between the sub-factor R and the indicator degradation is screened, and the irrelevant reasons R are eliminated to obtain the relevant wireless reasons R.

[0086] 4) Calculation of influence weights:

[0087] The impact weight of the ranking of the main factors is calculated according to the following formula:

[0088]

[0089] The three items in the above brackets correspond to probability association information, heartbeat association information and Pearson association information respectively. The probability association information corresponds to the probability association method, the heartbeat association information corresponds to the heartbeat association method, and the Pearson association information corresponds to the Pearson association method.

[0090] In terms of the probability association method, the number of times the sub-factor R corresponding to the degraded cell work order affects the cell network performance index K can be counted. This number of times is the probability association information, which can be recorded as x. In the actual calculation process, x is a probability constant.

[0091] As for the heartbeat correlation method, the sub-factor R and the cell index K may be combined first to obtain the correlation direction.

[0092] When the correlation direction is positive, the stronger the R element is, the more serious the degradation of the K element is, which means that the R element has a stronger influence on the K element. When the correlation direction is negative, the R element is strong but the K element is not affected by it, that is, the possibility that the degradation of R leads to the degradation of K is smaller.

[0093] The calculation formula for heartbeat correlation within each hour can be expressed as follows:

[0094] R i *K j

[0095] Among them, K j Represents the normalized index value at hourly granularity; R i It indicates the quality of the cell network performance indicator, and the value can be 0, 1 or 2; i and j represent the sequence number.

[0096] Next, a comprehensive weighted calculation may be performed in the time domain to obtain heartbeat correlation information.

[0097] Among them, the heartbeat related information can be recorded as:

[0098]

[0099] Among them, f(k, r) represents the heartbeat correlation information, K represents the cell network performance index, R represents the sub-factor, and n represents the number of hours of degradation time. The weighted K and R correlation relationships can be screened from the correlation matrix C.

[0100] As for the Pearson correlation method, given that the cell network performance index K may show different changes for the change of sub-factor R, by using the Pearson correlation coefficient, the variance correlation of the changes in the cell network performance index in the same period when different sub-factors change can be explored. At the same time, positive correlation and negative correlation are introduced to describe the impact of sub-factor changes on the cell network performance index, which is shown as follows:

[0101]

[0102] Where T represents the hourly statistical data during the time degradation period, Represent the average values ​​of R and K at different times.

[0103] D is the correction coefficient, which reflects the possibility that the sub-factor causes the deterioration of the K indicator. The correction coefficient D is obtained by statistically analyzing the real main cause R that causes the deterioration of the indicator K reported by network optimization engineers and calculating the linear function of the conditional probability f(P(R|K)) using the Bayesian formula.

[0104] 5) Adjustment of correction coefficient D

[0105] The Bayesian statistical theory is used to adjust the correction coefficient. The network optimization personnel provide manual feedback on the correctness of the main cause inference generated by the memory scheduling device. The data involved in the adjustment of the correction coefficient D for each work order includes the work order degradation indicator set K[K1, K2, K3.., K n ], the main reason ranking of manual feedback R XFirst, count the work order degradation index set K[K1, K2, K3.., K n ]:

[0106] P(K|Rx) counts the ratio of indicator K degradation caused by the main cause Rx reported by network optimization engineers of all historical work orders;

[0107] P(Rx) counts the ratio of deteriorated work orders caused by the main reason Rx in all historical work orders;

[0108] P(K) counts the ratio of indicator K degradation in all work orders;

[0109] Rx refers to the main cause of the impact of Rx on the degradation indicator set K reported by network optimization engineers.

[0110] Then use Bayes' formula to calculate the conditional probability:

[0111] P(Rx|K)=P(K|Rx)*P(Rx) / P(K)

[0112] The Bayesian probability estimate of the work order K being caused by the main cause Rx is obtained. The adjusted linear function of P(Rx|K) is used as the correction coefficient D for the next calculation of the impact weight.

[0113] 6) Main cause output:

[0114] The priority order of the sub-factors is determined according to the impact weights to output the main factors (for example, the top 4 sub-factors are taken as the main factors for output). It is worth noting that the interference and alarms of this cell are directly output regardless of priority, because interference and alarms are the direct factors leading to degradation. This is done to avoid the interference and alarms not being output normally due to the low priority.

[0115] The method provided by the embodiment of the present invention determines the influence weight based on the preset correction coefficient and the probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators, and determines the preset correction coefficient based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators. This can ensure the accuracy of the influence weight, ensure the accuracy of the priority of each main cause, and further ensure the accuracy of the global priority scheduling order of the data to be accessed.

[0116] Based on the above embodiment, after determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause, the method further includes:

[0117] Based on the access frequency of the data to be accessed corresponding to each main cause within a preset time period after the global priority scheduling order of the data to be accessed corresponding to each main cause is determined, the global priority scheduling order of the data to be accessed corresponding to each main cause is updated.

[0118] Specifically, Figure 2 As shown, the access priority of each table in the global data priority scheduling order table is counted and dynamically maintained based on the number of times the network optimization engineer actually accesses the database table when using it. The memory scheduling device counts the number of times the to-be-accessed data table is actually accessed by the network optimization engineer when using it, and counts the frequency of the data table actually accessed by the user in each R1-R2-R3-R4 main cause scenario, and updates the global data priority scheduling order table for use in the memory scheduling process, which can continuously improve the memory data hit rate.

[0119] The method provided by an embodiment of the present invention can continuously improve and enhance the memory data hit rate by updating the global priority scheduling order of the data to be accessed corresponding to each main cause based on the access frequency of the data to be accessed corresponding to each main cause within a preset time period after the global priority scheduling order of the data to be accessed corresponding to each main cause is determined.

[0120] Based on the above embodiment, the conditional probability that the main cause corresponding to the historical degraded cell work order causes the degradation of the cell network performance index is calculated by the Bayesian formula based on the ratio of the cell network performance index degradation caused by the main cause in all historical degraded cell work orders, the ratio of the degraded cell work orders caused by the main cause in all historical degraded cell work orders, and the ratio of the cell network performance index degradation in all historical degraded cell work orders.

[0121] Specifically, in the calculation method of the main cause, the correction coefficient D is also dynamically adjusted based on the feedback from the network optimization engineer on the accuracy of the main cause judgment algorithm of the memory scheduling device. During operation, the memory scheduling device records the correlation between each work order degradation index and the inferred main cause and the main cause feedback from the network optimization engineer; records the number of times the network optimization engineer accesses the database table in each R1-R2-R3-R4 main cause scenario, and uses this information to update the correction coefficient D:

[0122] Use the daily work order feedback data to recalculate the probability distribution of P(K|Rx), P(Rx), and P(K) in all historical work orders. Rx refers to the main factor affecting the degradation index set K reported by the network optimization engineer, and K is the degradation index set for each work order. Then use the Bayesian formula to calculate the conditional probability:

[0123] P(Rx|K)=P(K|Rx)*P(Rx) / P(K)

[0124] The adjusted P(Rx|K) linear function is used as the correction coefficient D for the next calculation of the influence weight.

[0125] The method provided in the embodiment of the present invention can continuously improve the accuracy of the main cause judgment and thereby improve the memory data hit rate by dynamically adjusting the correction coefficient D through Bayesian formula calculation based on the ratio of the cell network performance index degradation caused by the main cause in all historical degraded cell work orders, the ratio of the degraded cell work orders caused by the main cause in all historical degraded cell work orders, and the ratio of the cell network performance index degradation in all historical degraded cell work orders.

[0126] Based on any of the above embodiments, Figure 3 A schematic diagram of a memory scheduling device provided by an embodiment of the present invention, such as Figure 3 As shown, the device comprises:

[0127] The priority scheduling order determination module 310 is used to determine the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order, wherein the main cause corresponding to the degraded cell work order refers to the main reason causing the degradation of the network performance indicator corresponding to the degraded cell.

[0128] Specifically, network optimization engineers generally use the main causes output by the centralized network optimization system as the main basis for diagnosing network problems to query the corresponding analysis data, i.e., the data to be accessed. Based on this, the priority scheduling order determination module 310 can determine the global priority scheduling order of the data to be accessed corresponding to each main cause according to the main cause corresponding to the degraded cell work order, so that the network optimization engineer can access it.

[0129] The memory scheduling module 320 is used to schedule memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each of the main causes.

[0130] Specifically, the memory scheduling module 320 schedules the memory data based on the global priority scheduling order of the data to be accessed corresponding to each main cause, and stores or removes the data to be accessed corresponding to each main cause into or out of the memory according to the priority scheduling order, thereby avoiding the caching of irrelevant data and enabling network optimization engineers to quickly access data related to wireless network problems.

[0131] The device provided by an embodiment of the present invention determines the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order through a priority scheduling order determination module, and the memory scheduling module schedules the memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each main cause, which can not only reduce the memory space for caching irrelevant data, but also improve the hit rate of the memory data.

[0132] Based on the above embodiment, the method of determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order specifically includes:

[0133] Based on the main causes corresponding to the degraded cell work order, determine the to-be-accessed data corresponding to each main cause;

[0134] Determine the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period;

[0135] Among them, the main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on the sub-factors corresponding to the degraded cell work order and their influence weights, wherein the influence weight is used to characterize the degree of influence of the sub-factors corresponding to the degraded cell work order on the cell network performance indicators.

[0136] Based on the above embodiment, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period, specifically including:

[0137] Based on the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined.

[0138] Based on the above embodiment, the scheduling of memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each main cause specifically includes:

[0139] Based on the global priority scheduling order of the data to be accessed corresponding to the main reasons, the data to be accessed are loaded into the memory in descending order of priority, and / or removed from the memory in descending order of priority.

[0140] Based on the above embodiment, the impact weight is determined based on a preset correction coefficient and the probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators. The preset correction coefficient is determined based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators.

[0141] Based on the above embodiment, after determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause, the method further includes:

[0142] Based on the access frequency of the data to be accessed corresponding to each main cause within a preset time period after the global priority scheduling order of the data to be accessed corresponding to each main cause is determined, the global priority scheduling order of the data to be accessed corresponding to each main cause is updated.

[0143] Based on the above embodiment, the conditional probability that the main cause corresponding to the historical degraded cell work order causes the degradation of the cell network performance index is calculated by the Bayesian formula based on the ratio of the cell network performance index degradation caused by the main cause in all historical degraded cell work orders, the ratio of the degraded cell work orders caused by the main cause in all historical degraded cell work orders, and the ratio of the cell network performance index degradation in all historical degraded cell work orders.

[0144] The memory scheduling device provided in the embodiment of the present invention can execute the above-mentioned memory scheduling method. Its specific working principle and corresponding technical effects are the same as those of the above-mentioned method embodiment, and will not be repeated here.

[0145] Figure 4 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication interface 420 and the memory 430 communicate with each other through the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the step flow provided in the above method embodiment.

[0146] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0147] On the other hand, an embodiment of the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the step flow provided in the above method embodiment is implemented.

[0148] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0149] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A memory scheduling method, characterized in that: include: Based on the main causes corresponding to the degraded cell work orders, determine the global priority scheduling order of the to-be-accessed data corresponding to the main causes, wherein the main causes corresponding to the degraded cell work orders refer to the main causes that cause the degradation of the network performance indicators corresponding to the degraded cells; Scheduling memory data based on the global priority scheduling order of the to-be-accessed data corresponding to the main causes; The determining of the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order specifically includes: Based on the main causes corresponding to the degraded cell work order, determine the to-be-accessed data corresponding to each main cause; Determine the global priority scheduling order of the data to be accessed corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the data to be accessed corresponding to each main cause within a preset time period; Among them, the main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on the sub-factors corresponding to the degraded cell work order and their influence weights, wherein the influence weight is used to characterize the degree of influence of the sub-factors corresponding to the degraded cell work order on the cell network performance indicators.

2. The memory scheduling method according to claim 1, characterized in that: The determining, based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the to-be-accessed data corresponding to each main cause within a preset time period, a global priority scheduling order of the to-be-accessed data corresponding to each main cause specifically includes: Based on the priority of the main cause corresponding to the degraded cell work order and the weight corresponding to the priority, the historical access frequency of the data to be accessed within a preset time period and the weight corresponding to the access frequency, the global priority scheduling order of the data to be accessed corresponding to each main cause is determined.

3. The memory scheduling method according to claim 1, characterized in that: The scheduling of memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each of the main causes specifically includes: Based on the global priority scheduling order of the data to be accessed corresponding to the main reasons, the data to be accessed are loaded into the memory in descending order of priority, and / or removed from the memory in descending order of priority.

4. The memory scheduling method according to claim 1, characterized in that: The impact weight is determined based on a preset correction coefficient and the probability correlation information, heartbeat correlation information and Pearson correlation information between the sub-factors corresponding to the degraded cell work order and the cell network performance indicators. The preset correction coefficient is determined based on the conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance indicators.

5. The memory scheduling method according to claim 1, characterized in that: After determining the global priority scheduling order of the to-be-accessed data corresponding to each main cause, the method further includes: Based on the access frequency of the data to be accessed corresponding to each main cause within a preset time period after the global priority scheduling order of the data to be accessed corresponding to each main cause is determined, the global priority scheduling order of the data to be accessed corresponding to each main cause is updated.

6. The memory scheduling method according to claim 4, characterized in that: The conditional probability that the main cause corresponding to the historical degraded cell work order leads to the degradation of the cell network performance index is calculated by the Bayesian formula based on the ratio of the cell network performance index degradation caused by the main cause in all historical degraded cell work orders, the ratio of degraded cell work orders caused by the main cause in all historical degraded cell work orders, and the ratio of the cell network performance index degradation in all historical degraded cell work orders.

7. A memory scheduling device, characterized in that: include: A priority scheduling order determination module is used to determine the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order, wherein the main cause corresponding to the degraded cell work order refers to the main cause causing the degradation of the network performance indicator corresponding to the degraded cell; A memory scheduling module, used for scheduling memory data based on the global priority scheduling order of the to-be-accessed data corresponding to each of the main causes; The priority scheduling order determination module is specifically used to determine the to-be-accessed data corresponding to each main cause based on the main cause corresponding to the degraded cell work order; determine the global priority scheduling order of the to-be-accessed data corresponding to each main cause based on the priority of the main cause corresponding to the degraded cell work order and the historical access frequency of the to-be-accessed data corresponding to each main cause within a preset time period; wherein the main cause corresponding to the degraded cell work order and the priority of the main cause are determined based on each sub-factor corresponding to the degraded cell work order and their influence weights, wherein the influence weight is used to characterize the degree of influence of the sub-factor corresponding to the degraded cell work order on the cell network performance indicator.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the memory scheduling method according to any one of claims 1 to 6 are implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the memory scheduling method as claimed in any one of claims 1 to 6 are implemented.

Images