WO2022257743A1 - 流量抑制预测方法、电子设备、存储介质 - Google Patents
流量抑制预测方法、电子设备、存储介质 Download PDFInfo
- Publication number
- WO2022257743A1 WO2022257743A1 PCT/CN2022/094469 CN2022094469W WO2022257743A1 WO 2022257743 A1 WO2022257743 A1 WO 2022257743A1 CN 2022094469 W CN2022094469 W CN 2022094469W WO 2022257743 A1 WO2022257743 A1 WO 2022257743A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- network
- value
- parameter
- suppression
- prediction
- Prior art date
Links
- 230000001629 suppression Effects 0.000 title claims abstract description 190
- 238000000034 method Methods 0.000 title claims abstract description 86
- 238000003860 storage Methods 0.000 title claims abstract description 15
- 230000005540 biological transmission Effects 0.000 claims abstract description 29
- 238000013507 mapping Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims description 91
- 238000012360 testing method Methods 0.000 claims description 55
- 238000010606 normalization Methods 0.000 claims description 17
- 238000013486 operation strategy Methods 0.000 claims description 15
- 230000015654 memory Effects 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 210000004027 cell Anatomy 0.000 description 25
- 230000008569 process Effects 0.000 description 12
- 238000004458 analytical method Methods 0.000 description 9
- 230000005764 inhibitory process Effects 0.000 description 9
- 238000013480 data collection Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 238000007726 management method Methods 0.000 description 6
- 230000007423 decrease Effects 0.000 description 5
- 230000008450 motivation Effects 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 238000012795 verification Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000003062 neural network model Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
Definitions
- the present application relates to but not limited to the technical field of wireless communication, and in particular relates to a flow suppression prediction method, electronic equipment, and a storage medium.
- Embodiments of the present application provide a traffic suppression prediction method, electronic equipment, and a storage medium.
- the embodiment of the present application provides a flow suppression prediction method, including: determining the flow value of the suppression point according to a preset network flow model, the network flow model representing the mapping between the value of the network parameter of the transmission network and the flow value Relationship, the traffic value of the suppression point is the traffic threshold of the transmission network under the current operation strategy; determine the suppression reference value of the target network parameter corresponding to the traffic value of the suppression point; obtain the corresponding value of the target network parameter A parameter prediction value of , and determine a traffic suppression prediction result according to the parameter prediction value and the suppression reference value.
- the embodiment of the present application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and operable on the processor, when the processor executes the computer program, the following In one aspect, the traffic suppression prediction method.
- FIG. 1 is a flow chart of a traffic suppression prediction method provided by an embodiment of the present application
- FIG. 2 is a flow chart of training a network traffic model based on historical data provided by another embodiment of the present application
- FIG. 3 is a flow chart of training a network traffic model provided by another embodiment of the present application.
- FIG. 4 is a flowchart of a rasterized network parameter sample set provided by another embodiment of the present application.
- Fig. 5 is a flow chart of determining the flow value of the suppression point provided by another embodiment of the present application.
- FIG. 6 is a flow chart of data preprocessing provided by another embodiment of the present application.
- FIG. 7 is a flow chart of obtaining parameter prediction values provided by another embodiment of the present application.
- Fig. 8 is a flowchart of a training parameter prediction model provided by another embodiment of the present application.
- FIG. 9 is a flow chart for determining the degree of traffic suppression provided by another embodiment of the present application.
- Fig. 10 is a flow chart of determining root cause network parameters provided by another embodiment of the present application.
- Fig. 11 is a flow chart of obtaining history provided by another embodiment of the present application.
- Fig. 12 is a structural diagram of a network traffic management system provided by another embodiment of the present application.
- Fig. 13 is a flowchart of Example 1 provided by another embodiment of the present application.
- FIG. 14 is a schematic diagram of a one-dimensional mapping relationship of a network traffic model provided by another embodiment of the present application.
- Fig. 15 is a flowchart of Example 2 provided by another embodiment of the present application.
- FIG. 16 is a schematic diagram of a two-dimensional mapping relationship of a network traffic model provided by another embodiment of the present application.
- Fig. 17 is a flowchart of Example 3 provided by another embodiment of the present application.
- Fig. 18 is a structural diagram of an electronic device provided by another embodiment of the present application.
- the present application provides a traffic suppression prediction method, electronic equipment, and a storage medium.
- the traffic suppression prediction method includes: determining the traffic value of the suppression point according to a preset network traffic model, and the network traffic model represents the value of the network parameter of the transmission network and The mapping relation of the flow value, the flow value of the suppression point is the flow threshold of the transmission network under the current operation strategy; determine the suppression reference value of the target network parameter corresponding to the flow value of the suppression point; obtain the A parameter prediction value of the target network parameter, determining a traffic suppression prediction result according to the parameter prediction value and the suppression reference value.
- the traffic suppression prediction result corresponding to the target network parameters can be predicted before traffic suppression occurs, providing data basis for network optimization in advance, and effectively improving user experience.
- FIG. 1 is a flow chart of a traffic suppression prediction method provided by an embodiment of the present application.
- the traffic suppression method includes but is not limited to step S110 , step S120 and step S130 .
- Step S110 determine the flow value of the suppression point according to the preset network flow model, the network flow model represents the mapping relationship between the value of the network parameter of the transmission network and the flow value, and the flow value of the suppression point is the flow threshold of the transmission network under the current operation strategy.
- the network traffic model can be a pre-trained offline model, or it can be obtained through online training after collecting data. It should be noted that when the network traffic model is obtained by online training, the threshold value of the number of training times can also be set in advance. Execute to avoid false traffic suppression prediction when the network traffic model is inaccurate. Those skilled in the art know how to adjust the threshold value of the number of training times to ensure that the network traffic model can be used normally, so I won’t go into details here.
- the network parameters can be any parameters that can affect the traffic of the cell, such as the number of RRC radio resource control (Radio Resource Control, RRC) connected users, the number of active users, the utilization rate of traffic channel resources, the utilization rate of control channel resources, the frequency spectrum Efficiency (Spectrum Efficiency, SE), cell traffic, average modulation and coding scheme (Modulation and Coding scheme, MCS), average channel quality indicator (Channel Quality Indicator, CQI), average signal to noise ratio (Signal to Interference plus Noise Ratio, SINR ), etc., which are not limited in this embodiment.
- RRC Radio Resource Control
- RRC Radio Resource Control
- SE Frequency Efficiency
- MCS frequency spectrum Efficiency
- MCS average modulation and coding scheme
- CQI Channel Quality Indicator
- SINR Signal to Interference plus Noise Ratio
- the network traffic model can be the rule that the network traffic and network parameters obey. Since the suppression point represents the inflection point where the traffic begins to decline, the traffic threshold can be the average maximum traffic value that the transmission network can achieve. Therefore, One or more of the above network parameters can be selected for traffic suppression prediction. When a network parameter is selected for traffic prediction, for example, the number of users is selected as a network parameter, the relationship between the number of active users and network traffic can be determined according to the network traffic model.
- the average maximum flow value is the flow value of the suppression point;
- the average maximum flow value is the flow value of the suppression point, and the flow value of the suppression point at this time is associated with the four network parameters at the same time, so that the maximum flow that the cell can achieve under the influence of the above four network parameters can be determined.
- Step S120 determining the suppression reference value of the target network parameter corresponding to the traffic value of the suppression point.
- the number of target network parameters can be arbitrary. Therefore, in order to realize the prediction of traffic suppression, it is necessary to determine the suppression reference values of all target network parameters corresponding to the traffic value of the suppression point, for example, When the target network parameter is one of them, taking the number of users as an example, it is necessary to determine the suppression reference value corresponding to the number of users; Data, traffic channel resource utilization, control channel resource utilization, and channel quality suppression reference values provide accurate data basis for traffic suppression prediction.
- step S130 the parameter prediction value of the target network parameter is obtained, and the traffic suppression prediction result is determined according to the parameter prediction value and the suppression reference value.
- the parameter prediction value may be obtained through model prediction, for example, a parameter prediction model is preset, so as to obtain a parameter prediction value representing the future traffic state of the cell.
- the input data can be either a time series feature vector at a certain time granularity, or a scalar that characterizes the current state, for example, the 15-minute granularity of the past week
- the data sequence of the degree is the feature vector as the input of the prediction model to predict the parameter value of the granular cell in 15 minutes in the next day.
- the traffic suppression prediction result can be obtained by directly comparing the parameter prediction value with the suppression reference value, for example , there is only one target network parameter.
- the predicted value of the parameter is greater than the suppression reference value, it can be determined that as the value of the target network parameter increases, the network flow is limited and less than the flow value of the suppression point, and flow suppression occurs; another example, the target network parameter
- the three parameters of traffic channel resource utilization, control channel resource utilization and channel quality are used to quantify the relationship model between the number of users and network traffic through the section; and then according to the traffic channel resources of the current network Utilization, control channel resource utilization and channel quality status, determine the currently used inflection point model, judge whether suppression occurs according to a single parameter method, and obtain the corresponding suppression amount when suppression occurs, for subsequent prediction.
- the parameter prediction model and the network traffic model can be trained through the same sample set, and the corresponding training set and test set can be obtained from the sample set according to the specific model type selected. This embodiment does not make more limitations on this.
- step S110 in the embodiment shown in FIG. 1 , it also includes but is not limited to the following steps S210 to S220 .
- Step S210 acquiring a network parameter sample set from historical data of the transmission network, where the network parameter sample set includes network parameter samples.
- Step S220 training a network traffic model according to the network parameter sample set.
- the network traffic model is the law that obeys the relationship between network parameters and network traffic, therefore, obtaining a sample set of network parameters from the historical data of the transmission network and training the network traffic model can make the network traffic model reflect The historical operation status of the transmission network.
- the mathematical characteristics of the network flow model can be used to quickly determine the flow value of the suppression point. For example, the average maximum flow value in the historical operation process can be obtained as Suppressing point flow values can improve the efficiency of prediction.
- the suppression point of the transmission network is usually located in an approximate position, and the suppression point in the historical operation process can be determined as the suppression point of the current network.
- the flow value of the suppression point is predicted
- the method uses the historical data to train the network traffic model and then uses the mathematical characteristics of the model to determine the suppression point, which can effectively improve the accuracy of the suppression point, thereby improving the accuracy of traffic forecasting.
- the step S220 in the embodiment shown in FIG. 2 also includes but not limited to the following steps S310 to S330 .
- Step S310 obtaining a first training set and a first testing set according to the network parameter sample set.
- Step S320 normalize the first training set and the first test set.
- Step S330 acquiring a network traffic model, training the network traffic model according to the first training set, and verifying the trained network traffic model according to the first testing set.
- preprocessing such as data cleaning may also be performed on the network parameter sample set, so as to improve data integrity and usability, which will not be described in detail in this embodiment.
- the network traffic model can be a common machine learning model, such as a fully connected neural network model.
- machine learning models such as decision trees and regression models can also be used. model type.
- the network traffic model includes an input layer, an output layer and multiple hidden layers.
- the input layer is used to input the network parameters in the network parameter sample set
- the output layer is used to output the prediction results , such as SE or cell traffic, and those skilled in the art know how to adjust the number of hidden layers and the number of neurons according to the actual situation, so that the trained model can meet the requirements, which is not limited in this embodiment.
- the specific inverse normalization method may correspond to the above normalization method, which will not be described in detail in this embodiment.
- the step S310 in the embodiment shown in FIG. 3 also includes but not limited to the following steps S410 to S440 .
- Step S410 determining the value range of the network parameter sample set.
- Step S420 divide the value interval into several grid intervals according to the preset grid step size.
- step S430 each sample of the network parameter is filled into the grid interval according to the principle of value correspondence, and a grid dataset is obtained.
- Step S440 according to a preset screening strategy, filter out a first training set and a first test set from the raster dataset.
- determining the numerical interval of the network parameter sample set can be realized by determining the maximum value of the network parameter, and details are not repeated here. If multiple types of network parameters are involved, the maximum value of each type of network parameter can be determined separately, and the same
- the specific value of the grid step size can be determined according to actual needs, and this embodiment does not limit it.
- the samples of each network parameter can be filled into the corresponding grid interval according to the value of the sample of each network parameter as a key value.
- the grid of ⁇ MeshSampleMin, the average value of randomly selected 1/2 samples in the grid data is the first training set of the grid, and the average value of the remaining 1/2 samples is the first test set of the grid ;
- the step S120 in the embodiment shown in FIG. 1 also includes but not limited to the following steps S510 to S520 .
- Step S510 acquiring a preset threshold value, determining the first-order difference of the network parameter samples, and determining the network parameter samples whose first-order difference is smaller than the threshold value as target samples.
- Step S520 Determine the target flow value corresponding to the target sample according to the mapping relationship, and determine the target flow value as the suppression point flow value.
- the first-order difference is the difference between two consecutive adjacent terms in the discrete function
- the flow value of the suppression point is the average maximum flow value that can be realized by the current transmission network.
- the flow value The change begins to increase with the increase of the value of the network parameter, and the growth rate slows down before reaching the suppression point, and finally stops increasing after reaching the suppression point, and then decreases with the increase of the value of the network parameter. Therefore, theoretically, when the adjacent If the first-order difference between the two samples is zero, it can be determined that flow suppression has occurred, and the corresponding flow value is the flow value of the suppression point.
- step S510 can be performed separately for each network parameter to determine the target sample , which will not be elaborated here.
- the network traffic model can represent the mapping relationship between the value of the network parameters and the network traffic, after the target sample is determined, the corresponding target traffic value can be determined through the value of the target sample, so as to be further determined as the suppression point traffic value, and will not be repeated here.
- the step S210 in the embodiment shown in FIG. 2 also includes but not limited to the following steps S610 to S630 .
- step S610 a preset collection granularity and a collection period are acquired, wherein the collection granularity represents the frequency of collecting network parameters within one collection period.
- Step S620 collect samples of network parameters from historical data according to the collection granularity and collection period.
- Step S630 performing data preprocessing on the collected network parameter samples to obtain a network parameter sample set.
- the specific values of the collection granularity and the collection period can be determined according to actual needs. For example, if the collection granularity is one minute and the collection period is one day, then the network parameters can be collected once at the minute granularity to obtain samples. Those skilled in the art have The motivation is adjusted according to actual needs, so I won’t limit it here.
- the data set can be preprocessed, such as data time axis completion, data time axis deduplication and data completion.
- data time axis completion data time axis completion
- data time axis deduplication data time axis deduplication
- the deduplication rule can be to retain the data that appears for the first time in the data set and delete the subsequent data. Repeat the data and select specific rules according to the actual situation.
- the lack of data in the collection process can be resolved.
- the empty data at the head can be filled as the first non-empty data from the head;
- the empty data at the tail can be filled with Starting from the first non-empty data, when the missing data is in the middle of the data set, you can search for the first non-empty data forward and backward respectively, and perform linear interpolation filling.
- step S130 in the embodiment shown in FIG. 1 also includes, but is not limited to, the following steps S710 to S740 .
- step S710 the collection time of each collection period is determined according to the collection granularity.
- step S720 a prediction sample set is obtained from the network parameter sample set, and the sample data of the prediction sample set includes the collection time and the parameter value of the network parameter corresponding to the collection time.
- step S730 a parameter prediction model is trained according to the prediction sample set.
- Step S740 obtaining predicted feature samples corresponding to the target network parameters, and inputting the predicted feature samples into the parameter prediction model to obtain parameter predicted values.
- the collection granularity usually represents the collection frequency, and the collection time of the collection cycle is determined according to the collection frequency.
- the collected prediction sample set can refer to Table 1.
- Table 1 is the sample data corresponding to one network parameter, and the data corresponding to the collection time can be added correspondingly for various network parameters, and will not be repeated here.
- N is the data length of the network parameter sample set
- each row in the table is a sample of network parameters
- data_i represents the parameter value of the i-th sample
- time_i is the collection time of data_i.
- Table 1 Example table of forecast sample set data
- the sample data of the prediction sample set can be K+L+M+1 dimensional, where K is the parameter value at the same acquisition time in the previous K period, L is the parameter value at the previous L time, and M is the current data Time information, such as week, hour, minute, whether it is a holiday, etc.; at the same time, K+L+M before each sample is the feature of the sample, and the last dimension is the state value at the current moment, which is used to represent the label of the sample, which is convenient different samples.
- the first K+L dimensions of each sample may include all types of network parameters, and details will not be repeated here.
- the prediction feature sample can be obtained by selecting the time dimension. For example, if it is necessary to predict the parameter prediction value of K periods and L moments in the future, K+L+M dimensional data can be obtained from the prediction sample set as the prediction feature sample , those skilled in the art have the motivation to adjust the dimension of the predicted feature samples according to the actual forecasting requirements, which will not be limited here.
- step S730 in the embodiment shown in FIG. 7 also includes, but is not limited to, the following steps S810 to S850 .
- Step S810 performing normalization on the data of the prediction sample set.
- Step S820 Divide the normalized prediction sample set into a second training set and a second testing set according to a preset division ratio.
- Step S830 acquiring an initial parameter prediction model, and determining training parameters of the initial parameter prediction model.
- Step S840 training the initial parameter prediction model according to the second training set, and checking the trained initial parameter prediction model according to the second test set.
- step S850 the verified initial parameter prediction model is determined as the parameter prediction model.
- the data of the samples in each prediction sample set can also be normalized, such as the collection time, number of users, and traffic channel resources of a sample
- the utilization rate, the resource utilization rate of the control channel and the channel quality are normalized, and then the normalized prediction sample set is divided by a preset division ratio, so as to obtain a second training set and a second test set.
- the initial parameter prediction model can use a long-short-term memory (Long Short-Term Memory, LSTM) neural network, and can also be other alternative machine learning model decision trees, regression models, etc. Select a specific model for actual needs, so I won’t go into details here.
- LSTM Long Short-Term Memory
- the weights and biases of the neural network can be initialized randomly, the activation function can be selected, and the maximum number of iterative training of the state prediction model and the model error can be defined, where the model error is defined as the mean absolute error.
- the absolute error between the parameter prediction value state_predict and the real value state_true can be obtained by the following formula.
- N1 is the output feature dimension and N2 is the number of samples.
- the feature predicted samples can also be normalized, and the normalized parameters are the same as the normalized parameters of the predicted sample set, and the normalization method is also same to ensure data consistency.
- the step S210 in the embodiment shown in FIG. 2 also includes but not limited to the following steps S910 to S920 .
- Step S910 determining target historical data, which is historical data obtained by the transmission network according to the latest operation strategy.
- Step S920 acquiring network parameter samples from target historical data.
- the decision-making module usually judges the root cause of the traffic suppression based on the characteristic data of the suppression point, so as to obtain a network optimization strategy, so that the traffic of the cell is improved. Therefore, when the operation strategy changes, the suppression point will also change accordingly.
- the network traffic model and parameter prediction model trained by the historical data of the previous operation strategy cannot accurately predict the suppression point under the current operation strategy.
- the latest historical data of the operation policy is determined as the target historical data, and network parameter samples are reacquired from the target historical data.
- the update of the network traffic model and parameter prediction model can not only respond to changes in the operation strategy, but also pre-set the timer and update cycle, so that the model can be updated regularly with the latest data, effectively improving the network traffic model. and the accuracy of the parameter prediction model.
- step S130 in the embodiment shown in FIG. 1 is executed, it also includes but is not limited to the following steps S1010 to S1020 .
- Step S1010 when the traffic suppression prediction result indicates that traffic suppression occurs, input the predicted value of the parameter into the network traffic model to obtain the predicted traffic value.
- Step S1020 determining the flow suppression degree according to the flow prediction value and the suppression reference value.
- the traffic prediction corresponding to the parameter prediction value can be determined through the mapping relationship of the network traffic model. value.
- the flow loss can also be determined according to actual needs, for example, the absolute amount of flow loss and the relative degree of flow loss, the absolute amount of flow loss is Max(CellBestThp-CellPredictThp, 0), where , CellBestThp is the flow value of the suppression point, CellPredictThp is the flow prediction value; the relative degree of flow loss is Max(CellBestThp-CellPredictThp, 0)/CellBestThp, those skilled in the art also have motivation to select other parameters that can characterize flow loss according to actual needs, here Not much to limit.
- the analysis can only be performed on this network parameter; Plane method, to obtain the relationship between the number of users and network traffic.
- the network traffic model includes at least two different types of target network parameters. Referring to FIG. 11 , after step S1020 in the embodiment shown in FIG. 10 is executed, the following steps are also included but not limited to: S1111 to step S1120.
- Step S1111 determining the traffic suppression degree corresponding to each target network parameter.
- Step S1120 determining the root cause network parameter that causes traffic suppression, where the root cause network parameter is the target network parameter corresponding to the traffic suppression degree with the largest numerical value.
- the target network parameters of .
- the following takes four target network parameters as examples to illustrate the technical solution of this embodiment, wherein the four target network parameters are the number of users, traffic channel resource utilization, control channel resource utilization, and channel quality.
- each first component slice can be obtained
- the average value of the first-order difference in the first part is recorded as the first average value.
- the first half of the data of each first component slice is obtained, and the specific data range can be determined according to the actual data distribution; Slicing operation, so as to obtain the second component slice of the traffic prediction value of each target network parameter, and obtain the second average value in the same way as the first component slice; calculate the degree of suppression based on the average value of the first-order difference, and use the number of users As an example, the degree of suppression of the number of users satisfies the following relationship.
- UserNum_inhibition_ratio (diff_UserNum_best ⁇ diff_UserNum)/diff_UserNum_best, where UserNum_inhibition_ratio is the degree of inhibition, diff_UserNum_best is the first average value, and diff_UserNum is the second average value.
- the degree of traffic suppression corresponding to each target network parameter, the greater the degree of traffic suppression, the higher the degree of suppression, that is, the easier it is for traffic suppression to occur. Therefore, according to the numerical value of the degree of traffic suppression Determine the root cause network parameters, and then determine the main factors of traffic suppression, and optimize the network for the main factors, which can relieve traffic suppression to a greater extent.
- the traffic value of the suppression point is the average maximum traffic value of the transmission network under the current operation strategy.
- the traffic threshold can be the average maximum traffic value of the transmission network under the current operating strategy, for example, multiple maximum traffic values obtained through the training of the network traffic model, and then according to Multiple maximum flow values are used to calculate the average maximum flow value, thereby improving the prediction reference value of the flow value at the suppression point.
- the embodiment of the present application also provides a network traffic management system, including a data processing module 1210 , a network traffic model 1220 , a parameter prediction model 1230 and a traffic suppression analysis module 1240 .
- the data processing module 1210 is configured to collect data to obtain a sample set of network parameters, and is also configured to perform data preprocessing, including but not limited to data deduplication and data completion, wherein the methods of data deduplication and data completion can be Refer to the description of the embodiment shown in FIG. 6 , and details are not repeated here.
- the network traffic model 1220 is set to divide the network parameter sample set into grids, and obtain the first training set and the second test set according to the screening strategy.
- the specific principles and methods can refer to the description of the embodiment shown in FIG. 4 , and at the same time, in After the model is initialized, iterative training and model verification are performed to obtain a trained network traffic model.
- the specific principles and methods can refer to the description of the embodiment shown in FIG. 3 , and will not be repeated here.
- the parameter prediction model 1230 is configured to divide the preprocessed data to obtain a second training set and a second test set, and complete the initialization, training and testing of the parameter prediction model through the second training set and the second test set, For details, refer to the description of the embodiment shown in FIG. 8; and, it is also set to obtain the future parameter prediction value of the cell after inputting the prediction feature samples. For details, refer to the description of the embodiment shown in FIG. 7.
- the traffic suppression analysis module 1240 is configured to determine the result of traffic suppression, determine the degree of suppression and identify the cause of the suppression in the event of traffic suppression, for details, refer to the description of the embodiment shown in Figure 10 and Figure 11 , not much here For repeat.
- Example 1 In this example, one-dimensional modeling is adopted, that is, a network parameter is used as the network parameter, and the number of users is taken as an example. At the same time, the network traffic model and parameter prediction model are obtained through online training, and the parameter prediction value is obtained by The model prediction is obtained, and the network traffic model and the parameter prediction model are regularly updated. Referring to FIG. 13 , this example includes but is not limited to the following steps S1310 to S1350.
- step S1310 it is determined that both the network traffic model and the parameter prediction model are generated online, and it is judged whether the repeated execution times of network traffic model training failures have reached the first training threshold, and if not, execute step S1320, and if so, end.
- Step S1321 collect data according to the collection granularity and collection period, and preprocess the data.
- the collected data is historical data from modules such as network management, network elements, or data services, including the number of RRC connected users or activated users and cell traffic.
- Preprocessing includes data time extraction and completion, data time extraction Reload and data completion.
- the data time axis is completed online. During the data collection process, there may be omissions within a certain granularity. If some data information on the time axis is missing, the time axis will be filled and the corresponding data will be empty.
- Deduplication of data time axis is carried out online. In the process of data collection, repeated collection may occur within a certain granularity. It is necessary to deduplicate the data collected repeatedly. The deduplication rule is to keep the data that appears for the first time in the data set and delete other data. After the repeated data.
- the filled data set is the network parameter sample set.
- Step S1322 obtain the maximum number of users in the network parameter sample set, divide the user number grid into several grids with the grid step size, and deliver the samples in the network parameter sample set to the corresponding user number grids.
- the maximum number of users is UserNumMax
- the grid of user numbers is [0,UserNumMax]
- the grid step is UserNumMeshStep
- the data in the network parameter sample set is delivered to the corresponding grid by the number of users.
- Step S1323 normalize the data in the user number grid, and obtain the first training set and the first test set according to the screening strategy.
- the effective grid number threshold MeshSampleMin can be set in advance.
- the average value of 1/2 samples randomly selected in the user number grid data is The grid’s candidate training set, the average value of the remaining 1/2 samples is the grid’s candidate test set;
- the number of user grid data is less than MeshSampleMin, discard the grid data;
- the number of grids is between MeshSampleMin and 2 ⁇ MeshSampleMin, and the average value of the grid data is used as a candidate training set. Then perform online normalization on the candidate training set and the candidate test set respectively to obtain the first training set and the first test set.
- Step S1324 complete the training and verification of the network traffic model according to the first training set and the first test set.
- the network traffic model is constructed as a neural network model, and the network traffic model is randomly initialized, the network traffic model is trained using the first training set, and the model is evaluated after training to the maximum number of training times.
- the training error meets the requirements according to the first training set
- a test set is used to test the model, and if the test error meets the requirements, the network traffic model is output; otherwise, the number of training failures of the network traffic model is incremented by 1, and step S1310 is executed.
- Step S1330 when the update timer of the network traffic model reaches the update period of the network traffic model, execute step S1310, otherwise execute step S1331.
- Step S1331 determining the traffic value of the suppression point and the number of suppressed users according to the network traffic model.
- FIG. 14 is a schematic diagram of the mapping relationship of the network traffic model, wherein the abscissa is one-dimensional network parameters, such as any one of the number of users, traffic channel resource utilization, control channel resource utilization, and channel quality. , the ordinate is the value of the cell flow or cell SE. It can be seen that as the abscissa increases from the origin, the cell flow or cell SE gradually increases, and then decreases after reaching the suppression point, and flow suppression occurs.
- Step S1341 when the number of repeated executions of the parameter prediction model fails to reach the second training threshold, execute S1342; otherwise, obtain the real-time parameter value of the number of users, and use the filtered value of the real-time parameter value as the predicted parameter value to execute S1350;
- Step S1342 constructing a prediction sample set based on the time series-based network parameter sample set, and obtaining a second training set and a second test set according to the prediction sample set.
- each initial data contains the collection time and the corresponding number of users
- each row of the sample data set is a prediction sample
- the dimension of each prediction sample is K+L+M+1, where K is the previous The user value at the same moment in the K period, L is the user value at the previous L time, M is the time information of the current data, including week, hour, minute, etc., and the last dimension is the user value at the current moment, which is used to represent the prediction sample Label.
- the maximum and minimum normalization can be performed on each column of the prediction sample set according to the normalization parameters, and the normalized data can be divided into the first Two training sets and the second test set, and the parameter prediction model is trained and tested through the second training set and the second test set, when the error of training and testing meets the preset parameters, the parameter prediction model is output, otherwise the parameter prediction model Step S1341 is executed after the number of failed training times is incremented by 1.
- step S1350 the predicted value of the parameter is obtained, and the traffic suppression analysis is performed according to the predicted value of the state at a future moment output by the parameter prediction model.
- step S1310 is executed.
- the feature data of the predicted feature sample can be normalized using the normalization parameter. If there is data greater than 1 after normalization, it will be uniformly updated to 1, and the predicted feature sample will be input
- the parameter prediction model obtains the intermediate prediction value of the future period, and performs normalized inverse transformation on the intermediate prediction value, which is the future parameter prediction value.
- the analysis of flow suppression can determine whether the flow will be suppressed and the degree of suppression. If the parameter prediction value is less than or equal to the suppression reference value, the traffic will not be suppressed in this period in the future; otherwise, it will be suppressed in this period in the future. It is also possible to determine the degree of traffic loss. For example, input the parameter prediction value into the network traffic model to obtain the traffic prediction value of cell traffic as CellPredictThp, then the absolute amount of traffic loss is Max(CellBestThp-CellPredictThp, 0), where CellBestThp is the suppression point traffic value , the relative degree of flow loss is Max(CellBestThp-CellPredictThp, 0)/CellBestThp.
- Example 2 In this example, multi-dimensional modeling is adopted, that is, multiple types of network parameters are used as network parameters, taking the number of users, service channel resource utilization, control channel resource utilization, and channel quality as examples. At the same time, The network traffic model and the parameter prediction model adopt an offline model, and the parameter prediction value is obtained through model prediction. Referring to FIG. 15 , this example includes but is not limited to the following steps S1510 to S1530.
- step S1510 it is determined that both the network traffic model and the parameter prediction model are generated offline.
- Step S1521 collect data according to the collection granularity and collection period, and preprocess the data.
- data preprocessing is performed offline, including completion of the data time axis. During the data collection process, there may be omissions within a certain granularity. If some data information on the time axis is missing, time The axis is filled and the corresponding data is empty.
- data preprocessing also includes data completion. If the missing data is at the head of the data set, the empty data at the head will be filled as the first non-empty data from the head; if the missing data is at the head of the data set At the end of the set, fill the empty data at the end with the first non-empty data from the end; if the missing data is in the middle of the data set, search for the first non-empty data forward and backward, and perform linear interpolation filling. After filling The data set of is the network parameter sample set.
- Step S1522 rasterize the network parameter sample set according to the maximum value of each type of network parameter, and deliver the samples in the network parameter sample set to the corresponding grid.
- the rasterization specifically includes the following steps: obtain the maximum values of the number of users, service channel resource utilization, control channel resource utilization, and channel quality respectively, and record them as UserNumMax, ShareChannelMax, ControlChannelMax, and ChannelQualityMax in the above order.
- Step S1523 normalize the rasterized data, and obtain the first training set and the first test set according to the screening strategy.
- the threshold MeshSampleMin for the number of effective grids can be preset.
- the average value of randomly selected 1/2 samples in the grid data is Alternative training set, the average value of the remaining 1/2 samples is the candidate test set of the grid; when the number of grid data is less than MeshSampleMin, discard the grid data; when the number of grid data is between MeshSampleMin and 2
- the grid of ⁇ MeshSampleMin the average value of the grid data is used as a candidate training set.
- the candidate training set and the candidate test set are normalized offline, and the first training set and the first test set of the network traffic model are constructed offline.
- Step S1524 complete the training and verification of the network traffic model according to the first training set and the first test set.
- Step S1530 determining the suppression reference value of each type of network parameter according to the network traffic model.
- Figure 16 is a schematic diagram of the mapping relationship of the network traffic model under the two-dimensional network parameters.
- the network parameters include the number of users and the utilization rate of traffic channel resources.
- the cell traffic or cell SE gradually increases, reaching the average maximum traffic value after reaching the suppression point, and then any increase in network parameters will lead to a decrease in traffic, that is, Flow suppression occurs.
- Step S1540 constructing a prediction sample set offline on the time series-based network parameter sample set, and obtaining a second training set and a second test set according to the prediction sample set.
- Each prediction sample dimension is K+L+M+1 dimension (where K is the parameter value at the same moment in the previous K period, L is the parameter value at the previous L moment, and M is the time information of the current data, including week, hour, minute etc.), K+L+M before each sample is the feature of the sample, and the last dimension is the parameter value at the current moment, which is used to represent the label of the predicted sample.
- the normalization parameter is denoted as NormParameter
- the normalized data set is denoted as D_norm.
- the normalized data D_norm is divided into a second training set and a second test set according to the training set test set split ratio parameter alpha.
- Step S1550 complete offline training and testing of the parameter prediction model according to the second training set and the second testing set.
- step S1560 the parameter prediction value is obtained according to the parameter prediction model.
- obtaining the parameter prediction value includes the following steps: preparing data according to the data processing method, constructing the characteristic data data of the prediction sample, the data dimension is K+L+M dimension, where K is the parameter value at the same moment in the previous K period, L is the parameter value at the previous L moment, and M is the time information of the current data, including week, hour, minute, etc.
- Step S1570 loading the offline network traffic model and the offline parameter prediction model to complete the traffic suppression analysis.
- the state prediction value CellStatePredict can be input into the network traffic model, and the cell traffic in this state is CellPredictThp, then the absolute amount of traffic loss is Max(CellBestThp-CellPredictThp, 0), and the traffic loss The relative degree is Max(CellBestThp-CellPredictThp, 0)/CellBestThp.
- the analysis of the reasons for traffic suppression can be accomplished through the following steps: Obtain slices of the network traffic model corresponding to CellBest components CellBestUserNum, CellBestShareChannel, CellBestControlChannel, and CellBestChannelQuality, and take the average value of the first-order difference in the preset area of the curve, which are recorded as diff_UserNum_best, diff_ShareChannel_best, diff_ControlChannel_best, diff_ChannelQuality_best, where the preset area of the curve may be the first half close to the origin.
- the maximum value is UserNum_inhibition_ratio, the limitation of the number of users is the main factor; if the maximum value is ShareChannel_inhibition_ratio, the limitation of traffic channel resources is the main factor; if the maximum value is ControlChannel_inhibition_ratio, the limitation of control channel resources is the main factor; if the maximum value is ChannelQuality_inhibition_ratio, the channel quality limitation is the main factor.
- Example 3 In this example, two-dimensional modeling is adopted, and the network parameters are taken as examples of service channel resource utilization and control channel resource utilization. At the same time, the network traffic model adopts an online model, and the parameter prediction value adopts real-time parameters. Refer to FIG. 17 , this example includes but is not limited to the following steps S1710 to S1740.
- step S1710 it is determined that both the network traffic model and the parameter prediction model are generated online.
- Step S1721 collect data according to the collection granularity and collection period, and preprocess the data.
- the data is collected from the historical data of the network management system, with a collection granularity of 15 minutes and a collection period of 28 days, including service channel resource utilization, control channel resource utilization, and cell traffic.
- Preprocessing includes data time extraction and completion, data time extraction and deduplication, and data completion.
- the data time axis is completed online. During the data collection process, there may be omissions within a certain granularity. If some data information on the time axis is missing, the time axis will be filled and the corresponding data will be empty.
- Deduplication of data time axis is carried out online. In the process of data collection, repeated collection may occur within a certain granularity. It is necessary to deduplicate the data collected repeatedly. The deduplication rule is to keep the data that appears for the first time in the data set and delete other data. After the repeated data.
- the filled data set is the network parameter sample set.
- Step S1722 rasterize the network parameter sample set according to the maximum value of each type of network parameter, and deliver the samples in the network parameter sample set to the corresponding grid.
- the rasterization specifically includes the following steps: obtaining the maximum values of the traffic channel resource utilization rate and the control channel resource utilization rate respectively, and the distribution is recorded as ShareChannelMax and ControlChannelMax.
- Step S1723 normalize the rasterized data, and obtain the first training set and the first test set according to the screening strategy.
- the threshold MeshSampleMin for the number of effective grids can be preset.
- the average value of randomly selected 1/2 samples in the grid data is Alternative training set, the average value of the remaining 1/2 samples is the candidate test set of the grid; when the number of grid data is less than MeshSampleMin, discard the grid data; when the number of grid data is between MeshSampleMin and 2
- the grid of ⁇ MeshSampleMin the average value of the grid data is used as a candidate training set.
- the candidate training set and the candidate test set are normalized to construct a first training set and a first test set of the network traffic model.
- Step S1724 constructing the network traffic model as a neural network model and randomly initializing the network traffic model, and completing the training and verification of the network traffic model according to the first training set and the first test set.
- Step S1730 determine the flow value of the suppression point and the suppression reference value of each type of network parameter according to the network flow model.
- the suppression point obtained from the traffic suppression model is denoted as CellBestThp
- CellBestThp the traffic value of the suppression point obtained from the network traffic model
- CellBest [CellBestShareChannel, CellBestControlChannel].
- Step S1740 load the network traffic model, and complete the traffic suppression analysis.
- the real-time filtering value of the network parameter is used as the parameter prediction value, which is denoted as CellStatePredict.
- CellStatePredict When the future state CellStatePredict is less than or equal to CellBest, the traffic will not be suppressed in this period in the future; otherwise, it will be suppressed in this period in the future.
- the state prediction value CellStatePredict can be input into the network traffic model, and the cell traffic in this state is CellPredictThp, then the absolute amount of traffic loss is Max(CellBestThp-CellPredictThp, 0), and the traffic loss The relative degree is Max(CellBestThp-CellPredictThp, 0)/CellBestThp.
- the analysis of the reason for traffic suppression can be completed through the following steps: Obtain the slices of the network traffic model corresponding to the CellBest components CellBestShareChannel and CellBestControlChannel, and take the average value of the first-order difference of the preset area of the curve, which are respectively recorded as diff_ShareChannel_best, diff_ControlChannel_best, Wherein, the preset area of the curve may be the first half close to the origin. Obtain slices of ShareChannel and ControlChannel corresponding to the components of the network traffic model in CellStatePredict and take the first-order difference of each point before and after and take the average value, which are recorded as diff_ShareChannel and diff_ControlChannel respectively.
- the maximum value of UShareChannel_inhibition_ratio and ControlChannel_inhibition_ratio can be obtained, and the main reason causing traffic suppression can be determined. If the maximum value is ShareChannel_inhibition_ratio, the limitation of traffic channel resources is the main factor; if the maximum value is ControlChannel_inhibition_ratio, the limitation of control channel resources is the main factor.
- an embodiment of the present application also provides an electronic device.
- the electronic device 1800 includes: a memory 1810 , a processor 1820 , and a computer program stored in the memory 1810 and operable on the processor 1820 .
- the processor 1820 and the memory 1810 may be connected through a bus or in other ways.
- the non-transitory software programs and instructions required to realize the traffic suppression prediction method of the above-mentioned embodiment are stored in the memory 1810, and when executed by the processor 1820, the traffic suppression prediction method in the above-mentioned embodiment is executed, for example, the above-described Method step S110 to step S130 in Fig. 1, method step S210 to step S220 in Fig. 2, method step S310 to step S330 in Fig. 3, method step S410 to step S440 in Fig. 4, method step in Fig. 5 S510 to step S520, method step S610 to step S630 in Fig. 6, method step S710 to step S740 in Fig. 7, method step S810 to step S850 in Fig. 8, method step S910 to step S920 in Fig.
- the device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
- an embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by a processor or a controller, for example, by the above-mentioned Execution by a processor in the embodiment of the electronic device can cause the above-mentioned processor to execute the traffic suppression prediction method in the above-mentioned embodiment, for example, execute the method steps S110 to S130 in FIG. 1 described above, and the method steps in FIG. 2 S210 to step S220, method step S310 to step S330 in Fig. 3, method step S410 to step S440 in Fig. 4, method step S510 to step S520 in Fig. 5, method step S610 to step S630 in Fig.
- the embodiment of the present application includes: determining the flow value of the suppression point according to the preset network flow model, the network flow model represents the mapping relationship between the value of the network parameter of the transmission network and the flow value, and the flow value of the suppression point is the transmission network
- the traffic threshold under the current operation policy determine the suppression reference value of the target network parameter corresponding to the suppression point traffic value; obtain the parameter prediction value of the target network parameter, according to the parameter prediction value and the The suppression reference value determines the traffic suppression prediction result.
- the traffic value of the suppression point can be determined according to the trained network traffic model, and a data basis can be provided for traffic suppression prediction, so that the traffic suppression prediction result corresponding to the target network parameters can be predicted before traffic suppression occurs , to provide data basis for network optimization in advance and effectively improve user experience.
- Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, tape, magnetic disk storage or other magnetic storage devices, or can Any other medium used to store desired information and which can be accessed by a computer.
- communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
index | 0 | 1 | 2 | ... | i | ... | N-2 | N-1 |
time | time_0 | time_1 | time_2 | time_i | time_N-2 | time_N-1 | ||
state_data | data_0 | data_1 | data_2 | ... | data_i | ... | data_N-2 | data_N-1 |
Claims (14)
- 一种流量抑制预测方法,包括:根据预置的网络流量模型确定抑制点流量值,所述网络流量模型表征传输网络的网络参数的数值与流量值的映射关系,所述抑制点流量值为所述传输网络在当前运行策略下的流量阈值;确定与所述抑制点流量值所对应的目标网络参数的抑制参考值;获取与所述目标网络参数相对应的参数预测值,根据所述参数预测值和所述抑制参考值确定流量抑制预测结果。
- 根据权利要求1所述的方法,其中,在所述根据预置的网络流量模型确定抑制点流量值之前,所述方法还包括:从所述传输网络的历史数据获取网络参数样本集,所述网络参数样本集包括所述网络参数的样本;根据所述网络参数样本集训练所述网络流量模型。
- 根据权利要求2所述的方法,其中,所述根据所述网络参数样本集训练所述网络流量模型,包括:根据网络参数样本集得到第一训练集和第一测试集;对所述第一训练集和第一测试集进行归一化;获取网络流量模型,根据所述第一训练集训练所述网络流量模型,并根据所述第一测试集检验训练后的所述网络流量模型。
- 根据权利要求3所述的方法,其中,所述根据网络参数样本集得到第一训练集和第一测试集,包括:确定所述网络参数样本集的数值区间;根据预置的栅格步长将所述数值区间划分为若干个栅格区间;将所述网络参数的每个样本按照数值对应的原则填充至所述栅格区间,得到栅格数据集;根据预置的筛选策略,从所述栅格数据集筛选出所述第一训练集和所述第一测试集。
- 根据权利要求2所述的方法,其中,所述根据预置的网络流量模型确定抑制点流量值,包括:获取预置的门限值,确定所述网络参数的样本的一阶差分,将一阶差分小于所述门限值的所述网络参数的样本确定为目标样本;根据所述映射关系确定与所述目标样本所对应的目标流量值,将所述目标流量值确定为所述抑制点流量值。
- 根据权利要求2所述的方法,其中,所述从所述传输网络的历史数据获取网络参数样本集,包括:获取预置的采集粒度和采集周期,其中,所述采集粒度表征一个所述采集周期内采集所述网络参数的频率;根据所述采集粒度和所述采集周期从所述历史数据中采集网络参数的样本;对采集到的网络参数的样本进行数据预处理,得到所述网络参数样本集。
- 根据权利要求6所述的方法,其中,所述获取与所述目标网络参数相对应的参数预测值,包括:根据所述采集粒度确定每个采集周期的采集时刻;从所述网络参数样本集获取预测样本集,所述预测样本集的样本数据包括采集时刻和与所述采集时刻所对应的所述网络参数的参数值;根据所述预测样本集训练出参数预测模型,获取与所述目标网络参数相对应的预测特征样本,将所述预测特征样本输入至所述参数预测模型以得到所述参数预测值。
- 根据权利要求7所述的方法,其中,所述根据所述预测样本集训练出参数预测模型,包括:针对所述预测样本集的数据进行归一化;根据预置的分割比例将归一化后的所述预测样本集分割为第二训练集和第二测试集;获取初始参数预测模型,并确定所述初始参数预测模型的训练参数;根据所述第二训练集训练所述初始参数预测模型,并根据所述第二测试集检验训练后的所述初始参数预测模型;将检验后的所述初始参数预测模型确定为所述参数预测模型。
- 根据权利要求2所述的方法,其中,所述从传输网络的历史数据获取网络参数样本,包括:确定目标历史数据,所述目标历史数据为所述传输网络根据最新的运行策略所得到的历史数据;从所述目标历史数据获取网络参数样本。
- 根据权利要求1所述的方法,其中,在所述根据所述参数预测值和所述抑制参考值确定流量抑制预测结果之后,所述方法还包括:当所述流量抑制预测结果表征发生流量抑制,将所述参数预测值输入至所述网络流量模型,得到流量预测值;根据所述流量预测值和所述抑制参考值确定流量抑制程度。
- 根据权利要求10所述的方法,其中,所述网络流量模型中至少包括两种不同类型的目标网络参数,在所述根据所述流量预测值和所述抑制参考值确定流量抑制程度之后,所述方法还包括:确定与每个所述目标网络参数所对应的所述流量抑制程度;确定引起流量抑制的根因网络参数,所述根因网络参数为数值最大的所述流量抑制程度所对应的目标网络参数。
- 根据权利要求1所述的方法,其中:所述抑制点流量值为所述传输网络在当前运行策略下的平均最大流量值。
- 一种电子设备,包括:存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,其中,所述处理器执行所述计算机程序时实现如权利要求1至12中任意一项所述的流量抑制预测方法。
- 一种计算机可读存储介质,存储有计算机可执行指令,其中,所述计算机可执行指令被设置为执行如权利要求1至12中任意一项所述的流量抑制预测方法。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP22819341.3A EP4354941A1 (en) | 2021-06-11 | 2022-05-23 | Flow suppression prediction method, electronic device and storage medium |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110656628.7A CN115550195A (zh) | 2021-06-11 | 2021-06-11 | 流量抑制预测方法、电子设备、存储介质 |
CN202110656628.7 | 2021-06-11 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022257743A1 true WO2022257743A1 (zh) | 2022-12-15 |
Family
ID=84424604
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2022/094469 WO2022257743A1 (zh) | 2021-06-11 | 2022-05-23 | 流量抑制预测方法、电子设备、存储介质 |
Country Status (3)
Country | Link |
---|---|
EP (1) | EP4354941A1 (zh) |
CN (1) | CN115550195A (zh) |
WO (1) | WO2022257743A1 (zh) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117041073A (zh) * | 2023-09-05 | 2023-11-10 | 广州天懋信息系统股份有限公司 | 网络行为预测方法、系统、设备及存储介质 |
CN117319312A (zh) * | 2023-11-29 | 2023-12-29 | 凯美瑞德(苏州)信息科技股份有限公司 | 一种数据流量控制方法及装置 |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117858169A (zh) * | 2022-09-29 | 2024-04-09 | 中兴通讯股份有限公司 | 小区流量控制方法、基站及存储介质 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6272110B1 (en) * | 1997-10-10 | 2001-08-07 | Nortel Networks Limited | Method and apparatus for managing at least part of a communications network |
US20170111233A1 (en) * | 2015-10-15 | 2017-04-20 | Citrix Systems, Inc. | Systems and methods for determining network configurations using historical and real-time network metrics data |
WO2017200651A1 (en) * | 2016-05-17 | 2017-11-23 | Brocade Communications Systems, Inc. | Anomaly detection and prediction in a packet broker |
US20190028537A1 (en) * | 2017-07-18 | 2019-01-24 | Wipro Limited | Method and system for dynamically handling load on a computer network |
CA3117707A1 (en) * | 2018-10-26 | 2020-04-30 | Hughes Network Systems, Llc | Monitoring a communication network |
-
2021
- 2021-06-11 CN CN202110656628.7A patent/CN115550195A/zh active Pending
-
2022
- 2022-05-23 WO PCT/CN2022/094469 patent/WO2022257743A1/zh active Application Filing
- 2022-05-23 EP EP22819341.3A patent/EP4354941A1/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6272110B1 (en) * | 1997-10-10 | 2001-08-07 | Nortel Networks Limited | Method and apparatus for managing at least part of a communications network |
US20170111233A1 (en) * | 2015-10-15 | 2017-04-20 | Citrix Systems, Inc. | Systems and methods for determining network configurations using historical and real-time network metrics data |
WO2017200651A1 (en) * | 2016-05-17 | 2017-11-23 | Brocade Communications Systems, Inc. | Anomaly detection and prediction in a packet broker |
US20190028537A1 (en) * | 2017-07-18 | 2019-01-24 | Wipro Limited | Method and system for dynamically handling load on a computer network |
CA3117707A1 (en) * | 2018-10-26 | 2020-04-30 | Hughes Network Systems, Llc | Monitoring a communication network |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117041073A (zh) * | 2023-09-05 | 2023-11-10 | 广州天懋信息系统股份有限公司 | 网络行为预测方法、系统、设备及存储介质 |
CN117041073B (zh) * | 2023-09-05 | 2024-05-28 | 广州天懋信息系统股份有限公司 | 网络行为预测方法、系统、设备及存储介质 |
CN117319312A (zh) * | 2023-11-29 | 2023-12-29 | 凯美瑞德(苏州)信息科技股份有限公司 | 一种数据流量控制方法及装置 |
CN117319312B (zh) * | 2023-11-29 | 2024-03-08 | 凯美瑞德(苏州)信息科技股份有限公司 | 一种数据流量控制方法及装置 |
Also Published As
Publication number | Publication date |
---|---|
CN115550195A (zh) | 2022-12-30 |
EP4354941A1 (en) | 2024-04-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022257743A1 (zh) | 流量抑制预测方法、电子设备、存储介质 | |
KR101984730B1 (ko) | 서버 장애 자동 예측 시스템 및 자동 예측 방법 | |
US10223190B2 (en) | Identification of storage system elements causing performance degradation | |
CN110413227B (zh) | 一种硬盘设备的剩余使用寿命在线预测方法和系统 | |
CN109710612B (zh) | 向量索引的召回方法、装置、电子设备和存储介质 | |
CN110289994B (zh) | 一种集群容量调整方法及装置 | |
CN112001644B (zh) | 一种配电网运行可靠性检测方法、装置、终端及存储介质 | |
US11258659B2 (en) | Management and control for IP and fixed networking | |
CN114943321A (zh) | 一种针对硬盘的故障预测方法、装置及设备 | |
CN115774652A (zh) | 一种基于聚类算法的群控设备健康监测方法、设备及介质 | |
CN115996249A (zh) | 一种基于分级的数据传输方法及装置 | |
CN113673822A (zh) | 一种弹性调度方法及系统 | |
CN111026624B (zh) | 电网信息系统的故障预测方法、装置 | |
CN116295506A (zh) | 一种车辆剩余里程的预测方法、装置、设备及介质 | |
CN110807014A (zh) | 一种基于交叉验证的台区数据异常甄别方法和装置 | |
CN116318386A (zh) | 光模块的失效预测方法及其系统、存储介质 | |
CN115712834A (zh) | 一种告警误报检测方法、装置、设备及存储介质 | |
CN115115107A (zh) | 光伏功率的预测方法、预测装置和计算机设备 | |
CN114330569A (zh) | 一种检测风机组部件故障的方法、设备及存储介质 | |
CN114398235A (zh) | 基于融合学习和假设检验的内存回收趋势预警装置及方法 | |
CN116413609B (zh) | 一种电池跳水识别方法、装置、电子设备及存储介质 | |
CN111698700B (zh) | 一种小区的工作状态的判定方法及装置 | |
CN111768021A (zh) | 订单价格调整方法、装置、服务器及存储介质 | |
CN115511106B (zh) | 基于时序数据生成训练数据的方法、设备和可读存储介质 | |
CN116450485B (zh) | 一种应用性能干扰的检测方法和系统 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22819341 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18567597 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2022819341 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 2022819341 Country of ref document: EP Effective date: 20240110 |