CN109995592A - Quality of service monitoring method and equipment - Google Patents
Quality of service monitoring method and equipment Download PDFInfo
- Publication number
- CN109995592A CN109995592A CN201910280368.0A CN201910280368A CN109995592A CN 109995592 A CN109995592 A CN 109995592A CN 201910280368 A CN201910280368 A CN 201910280368A CN 109995592 A CN109995592 A CN 109995592A
- Authority
- CN
- China
- Prior art keywords
- data
- moment
- flows
- network element
- target network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/50—Testing arrangements
- H04L43/55—Testing of service level quality, e.g. simulating service usage
Abstract
The present invention provides a kind of quality of service monitoring method and equipment.This method, comprising: obtain data on flows of the target network element at N number of moment in past, the data on flows includes: the flow for flowing into the target network element and at least one of the flow for flowing out the target network element, wherein N is the integer more than or equal to 1;Prediction model trained according to the data on flows at N number of moment in the past and in advance predicts the target network element in the data on flows at following N number of moment;According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the quality of service of the target network element is determined.Artificial setting alarm baseline is compared in the prior art, and the data on flows and business feature that the present embodiment is predicted by prediction model more match, therefore higher come the quality of service accuracy that determines according to the data on flows that prediction model is predicted.
Description
Technical field
The present invention relates to big data field more particularly to a kind of quality of service monitoring methods and equipment.
Background technique
With the continuous development of Internet technology and mobile terminal technology, constantly expanded using the userbase of mobile data
Greatly, the mobile data flow of Operator Core Network is caused rapidly to increase, datacom device link load Continued.Datacom device chain
Road load and the quality of service of mobile data are closely bound up, and being monitored in real time to quality of service has become core net O&M work
The emphasis of work.
In the prior art, generally use the quality of service as under type logarithm leads to equipment to be monitored: staff is manual
One fixed alarm baseline is set, when the flow of datacom device is more than the alarm baseline, Real-time Feedback to network management system, network management
System judges the quality of service of the datacom device according to the information that datacom device is fed back.However, staff be often based on
Above-mentioned alarm baseline is arranged in past experience, not high using quality of service accuracy obtained by the above method.
Summary of the invention
The present invention provides a kind of quality of service monitoring method and equipment, quasi- for solving the quality of service that the prior art determines
The not high problem of exactness.
In a first aspect, the present invention provides a kind of quality of service monitoring method, comprising:
Data on flows of the target network element at N number of moment in past is obtained, the data on flows includes: to flow into the target network element
Flow and flow out at least one of the flow of the target network element, wherein N is the integer more than or equal to 1;
Prediction model trained according to the data on flows at N number of moment in the past and in advance, predicts that the target network element exists
The data on flows at following N number of moment;
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target is determined
The quality of service of network element.
Optionally, the data on flows according to N number of moment in the past and prediction model trained in advance, described in prediction
Target network element is before the data on flows at following N number of moment, further includes:
The prediction model is obtained, the prediction model is used to indicate the target network element in the flow at N number of moment in past
The relationship of data and the data on flows at following N number of moment.
It is optionally, described to obtain the prediction model, comprising:
The target network element is extracted from database in the sample data at M moment of past;
According to the sample data, training obtains the prediction model.
Optionally, described according to the sample data, training obtains the prediction model, comprising:
According to the sample data, the flux deepness learning model based on time series is established;
The flux deepness learning model is trained, the prediction model is obtained.
Optionally, described that the flux deepness learning model is trained, obtain the prediction model, comprising:
Initialize the model parameter of the flux deepness learning model;
According to the sample data, the model parameter is adjusted;
When the precision of prediction of the flux deepness learning model reaches preset threshold, determined according to corresponding model parameter
The prediction model.
Optionally, described according to the sample data, adjust the model parameter, comprising:
According to the sample data, the extent of deviation of the flux deepness learning model calculated result is calculated;
According to the extent of deviation of the calculated result, the model parameter is adjusted.
Optionally, the actual monitoring data of the data on flows according to N number of moment in future and following N number of moment,
Determine the quality of service of the target network element, comprising:
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target is determined
The link load state of network element;
According to the link load state of the target network element, the quality of service of the target network element is determined.
Second aspect, the present invention provide a kind of quality of service monitoring device,
Module is obtained, for obtaining target network element in the data on flows at N number of moment in past, the data on flows includes: stream
Enter the flow of the target network element and flow out at least one of the flow of the target network element, wherein N be more than or equal to
1 integer;
Prediction module predicts institute for prediction model trained according to the data on flows at N number of moment in the past and in advance
Target network element is stated in the data on flows at following N number of moment;
Determining module, for the data on flows and the actual monitoring number at following N number of moment according to N number of moment in future
According to determining the quality of service of the target network element.
Optionally, the acquisition module, is also used to:
The target network element is extracted from database in the sample data at M moment of past;
According to the sample data, training obtains the prediction model.
Optionally, the acquisition module, is specifically used for:
According to the sample data, the flux deepness learning model based on time series is established;
The flux deepness learning model is trained, the prediction model is obtained.
Optionally, the acquisition module, is specifically used for:
Initialize the model parameter of the flux deepness learning model;
According to the sample data, the model parameter is adjusted;
When the precision of prediction of the flux deepness learning model reaches preset threshold, determined according to corresponding model parameter
The prediction model.
Optionally, the acquisition module, is specifically used for:
According to the sample data, the extent of deviation of the flux deepness learning model calculated result is calculated;
According to the extent of deviation of the calculated result, the model parameter is adjusted.
Optionally, the determining module, is specifically used for:
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target is determined
The link load state of network element;
According to the link load state of the target network element, the quality of service of the target network element is determined.
The third aspect, the present invention provide a kind of computer readable storage medium, are stored thereon with computer program, the meter
Calculation machine program realizes above-mentioned quality of service monitoring method when being executed by processor.
Fourth aspect, the present invention provide a kind of network management system, comprising:
Processor;And
Memory, for storing the executable instruction of the processor;
Wherein, the processor is configured to realize above-mentioned quality of service monitoring side via the executable instruction is executed
Method.
Quality of service monitoring method and equipment provided by the invention first obtain target network element in the flow at N number of moment in past
Data obtain the stream at following N number of moment then by the data on flows input at N number of moment in past prediction model trained in advance
Data are measured, the actual monitoring data of the data on flows at obtained N number of moment in the future and following N number of moment are compared, from
And determine the quality of service of target network element, it is compared with artificial setting alarm baseline in the prior art, the present invention passes through prediction model
Predict that obtained data on flows and business feature more match, therefore determine according to the data on flows that prediction model is predicted
Quality of service accuracy it is higher.
Detailed description of the invention
Fig. 1 is the application scenario diagram that quality of service monitoring method provided by the invention is related to;
Fig. 2 is the flow diagram of the embodiment one of quality of service monitoring method provided by the invention;
Fig. 3 is the flow diagram of the embodiment two of quality of service monitoring method provided by the invention;
Fig. 4 is the structural schematic diagram of quality of service monitoring device provided by the invention;
Fig. 5 is the hardware structural diagram of network management system provided by the invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Description and claims of this specification and term " first ", " second ", " third " " in above-mentioned attached drawing
The (if present)s such as four " are to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should manage
The data that solution uses in this way are interchangeable under appropriate circumstances, so that the embodiment of the present invention described herein for example can be to remove
Sequence other than those of illustrating or describe herein is implemented.
In addition, term " includes " and " having " and their any deformation, it is intended that covering non-exclusive includes example
Such as, the process, method, system, product or equipment for containing a series of steps or units those of are not necessarily limited to be clearly listed
Step or unit, but may include being not clearly listed or intrinsic for these process, methods, product or equipment other
Step or unit.
The link load of core net datacom device and the quality of service of mobile data have substantial connection.Generally, Ke Yiyong
Flow measures the link load of datacom device, such as: it can use the flow or outflow datacom device for flowing into datacom device
Flow measures the link load of datacom device.In the prior art, it generallys use under type logarithm such as and leads to the quality of service of equipment
It is monitored: staff's manual setting one fixed alarm baseline, it is real when the flow of datacom device is more than the alarm baseline
When feed back to network management system, network management system judges the quality of service of the datacom device according to the information that datacom device is fed back.
However, staff, which is often based on previous experience, is arranged above-mentioned alarm baseline, network management system is led in number to be set
Just directly determine that quality of service is poor when standby flow is more than the alarm baseline.Since staff's manual setting alerts baseline
Method can not accomplish that the precision for business feature is arranged, such as: working day and day off corresponding flow timing variations are
Different, the alarm baseline of staff's setting may be suitble to working day, but not necessarily be suitble to day off.In this case network management
The quality of service of system judgement has deviation.
Based on above-mentioned technical problem, the present invention provides a kind of quality of service monitoring method and equipment, first obtains target network element
In the data on flows at N number of moment in past, then the data on flows at N number of moment in past is inputted in prediction model trained in advance,
The data on flows at following N number of moment is obtained, by the reality of the data on flows at obtained N number of moment in the future and following N number of moment
Monitoring data compare, so that it is determined that the quality of service of target network element, is compared with artificial setting alarm baseline in the prior art,
The data on flows and business feature that the present invention is predicted by prediction model more match, therefore measure in advance according to prediction model
To data on flows it is higher come the quality of service accuracy determined.
Fig. 1 is the application scenario diagram that quality of service monitoring method provided by the invention is related to.Scene figure shown in Fig. 1 includes:
Network management system and core network element.
Wherein, core network element can be mobile management nodes (Mobility Management Entity
Function, abbreviation MME), it can be gateway (Serving GateWay, abbreviation SGW), can be strategy and advised with charging
Then functional unit (Policy and Charging Rules Function, abbreviation PCRF) can be home signature user clothes
It is engaged in device (Home Subscriber Server, abbreviation HSS), can also be other network elements of core net, the present invention does not do this
It limits.
Wherein, core network element can establish connection, core network element by the modes such as wireless or wired and network management system
Quantity can be to be multiple, Fig. 1 only shows one, and the present invention is not limited with NE quantity shown in Fig. 1.Core network element can will be every
The data on flows at a moment is reported to network management system, and network management system can be by executing quality of service monitoring method provided by the invention
To determine the quality of service of core network element.
The process of quality of service monitoring method provided by the invention is executed to network management system below with reference to specific embodiment
Be illustrated, below these specific embodiments can be combined with each other, may be for the same or similar concept or process
It is repeated no more in some embodiments.Below in conjunction with attached drawing, the embodiment of the present invention is described.
Fig. 2 is the flow diagram of the embodiment one of quality of service monitoring method provided by the invention.As shown in Fig. 2, this
The quality of service monitoring method that embodiment provides, comprising:
S201, target network element is obtained in the data on flows at N number of moment in past.
Wherein, data on flows includes: to flow into the flow of the target network element and the flow of the outflow target network element
It is at least one, wherein N is the integer more than or equal to 1.
Wherein, target network element can be any network element in core net, and all network elements can detect each moment in core net
To data on flows be reported to network management system.
Ginseng is shown in Table 1, and the data on flows that target network element is reported to network management system can be as shown in table 1.
Table 1
Moment | Flow into flow | Flow out flow |
2018-8-12 20:00 | trunk_in1 | trunk_out1 |
2018-8-12 20:15 | trunk_in2 | trunk_out2 |
2018-8-12 20:30 | trunk_in3 | trunk_out3 |
2018-8-12 20:45 | trunk_in4 | trunk_out4 |
2018-8-12 21:00 | trunk_in5 | trunk_out5 |
2018-8-12 21:15 | trunk_in6 | trunk_out6 |
2018-8-12 21:30 | trunk_in7 | trunk_out7 |
2018-8-12 21:45 | trunk_in8 | trunk_out8 |
2018-8-12 22:00 | trunk_in9 | trunk_out9 |
2018-8-12 22:15 | trunk_in10 | trunk_out10 |
2018-8-12 22:30 | trunk_in11 | trunk_out11 |
2018-8-12 22:45 | trunk_in12 | trunk_out12 |
2018-8-12 23:00 | trunk_in13 | trunk_out13 |
2018-8-12 23:15 | trunk_in14 | trunk_out14 |
2018-8-12 23:30 | trunk_in15 | trunk_out15 |
2018-8-12 23:45 | trunk_in16 | trunk_out16 |
2018-8-13 00:00 | trunk_in17 | trunk_out17 |
2018-8-13 00:15 | trunk_in18 | trunk_out18 |
2018-8-13 00:30 | trunk_in19 | trunk_out19 |
2018-8-13 00:45 | trunk_in20 | trunk_out20 |
2018-8-13 01:00 | trunk_in21 | trunk_out21 |
2018-8-13 01:15 | trunk_in22 | trunk_out22 |
2018-8-13 01:30 | trunk_in23 | trunk_out23 |
2018-8-13 01:45 | trunk_in24 | trunk_out24 |
2018-8-13 02:00 | trunk_in25 | trunk_out25 |
2018-8-13 02:15 | trunk_in26 | trunk_out26 |
2018-8-13 02:30 | trunk_in27 | trunk_out27 |
Target network element shown in table 1 out of 2018-8-12 20:00 to 2018-8-13 02:30 this period, every
The data on flows that 15min is reported to network management system, table 1 by report interval time be 15min for illustrate, it should be noted that
Interval is reported to can also be 5min, 10min, 20min etc., which is not limited by the present invention.
Optionally, other network elements in core net can also report corresponding flow number to network management system through the above way
According to.After network management system receives the data on flows that each core network element reports, data on flows subnetting member can be stored to data
In library.
Wherein, the data on flows at N number of moment in past can be the data on flows at N number of moment in certain time period.Correspondingly,
Above-mentioned acquisition target network element can be in the implementation of the data on flows at N number of moment in past are as follows: target network is extracted from database
The data on flows at member moment N number of within a preset period of time.
Such as: ginseng is shown in Table 1, it is assumed that preset time period is 2018-8-12 20:00 to 2018-8-1222:00, then mentions
The data on flows at N number of moment is got as the data in table 2.
Table 2
Moment | Flow into flow | Flow out flow |
2018-8-12 20:00 | trunk_in1 | trunk_out1 |
2018-8-12 20:15 | trunk_in2 | trunk_out2 |
2018-8-12 20:30 | trunk_in3 | trunk_out3 |
2018-8-12 20:45 | trunk_in4 | trunk_out4 |
2018-8-12 21:00 | trunk_in5 | trunk_out5 |
2018-8-12 21:15 | trunk_in6 | trunk_out6 |
2018-8-12 21:30 | trunk_in7 | trunk_out7 |
2018-8-12 21:45 | trunk_in8 | trunk_out8 |
2018-8-12 22:00 | trunk_in9 | trunk_out9 |
S202, the data on flows according to N number of moment in the past and prediction model trained in advance, predict the target network
Data on flows of the member at following N number of moment.
Wherein, N number of moment in past and following N number of moment constitute continuous time series, for table 1, it is assumed that past N
A moment is 9 moment in 2018-8-12 20:00 to 2018-8-12 22:00, then predicting to obtain by prediction model
9 moment that are target network element after 2018-8-12 22:00 data on flows.
Optionally, before using prediction model predicted flow rate data, quality of service monitoring method provided by the invention is also
Can include: the prediction model is obtained, the prediction model is used to indicate the target network element in the flow number at N number of moment in past
According to the relationship with the data on flows at following N number of moment.
Optionally, the acquisition modes of above-mentioned prediction model can be with are as follows: firstly, extracting target network element from database in the past
The sample data at M moment, M are greater than 2N;Then, according to the sample data, training obtains the prediction model.Due to prediction
Model is obtained according to the training of the sample data of target network element, and the data on flows front and back that may learn target network element is N number of
The changing character at moment, therefore, the data on flows at the N number of moment in future predicted by the prediction model more close to
Actual flow data, the data on flows predicted according to prediction model are higher come the quality of service accuracy determined.
Specifically, S201 get over N number of moment data on flows and training obtain prediction model on the basis of,
The data on flows input prediction model at N number of moment in past, prediction model can be exported to the data on flows at following N number of moment.
S203, data on flows and the actual monitoring data at following N number of moment according to N number of moment in future, determine institute
State the quality of service of target network element.
Specifically, network management system receives the actual monitoring data at N number of moment in future that target network element reports, according to the reality
The data on flows at N number of moment in future that monitoring data and S202 are predicted, determines the link load state of the target network element,
And then according to the link load state of the target network element, the quality of service of the target network element is determined.
Specifically, the absolute difference of actual monitoring data and prediction data can be sought, if the absolute difference is greater than default threshold
Value, then it is assumed that the link load state of target network element is high load capacity, and the quality of service of target network element is not high;If above-mentioned absolute difference
Less than preset threshold, then it is assumed that the link load state of target network element is normal duty, and the quality of service of target network element is higher.
Optionally, link load state can also be refined, such as: it can be by link load state demarcation are as follows: high load capacity,
Normal duty and underload etc..At this moment need to be arranged multiple threshold values, the absolute difference of actual monitoring data and prediction data exists
When in different range, corresponding different link load state.
Quality of service monitoring method provided in this embodiment first obtains data on flows of the target network element at N number of moment in past,
Then by the data on flows input at N number of moment in past prediction model trained in advance, the flow number at following N number of moment is obtained
According to the actual monitoring data of the data on flows at obtained N number of moment in the future and following N number of moment being compared, thus really
Set the goal the quality of service of network element, compares with artificial setting alarm baseline in the prior art, the present embodiment is pre- by prediction model
The data on flows and business feature measured more match, therefore determines according to the data on flows that prediction model is predicted
Quality of service accuracy is higher.
The process that prediction model is obtained in above-described embodiment is described in detail below with reference to specific embodiment.Fig. 3
For the flow diagram of the embodiment two of quality of service monitoring method provided by the invention.As shown in figure 3, provided in this embodiment
Quality of service monitoring method, comprising:
S301, target network element is obtained in the data on flows at N number of moment in past.
Wherein, the implementation of S301 can be found in above-described embodiment, and details are not described herein by the present invention.
S302-S304 is the process for obtaining prediction model below, specific:
S302, the target network element is extracted from database in the sample data at M moment of past.
S303, according to the sample data, establish the flux deepness learning model based on time series.
S304, the flux deepness learning model is trained, obtains the prediction model.
Specifically, the implementation being trained to the flux deepness learning model can be with are as follows:
Initialize the model parameter of the flux deepness learning model;According to the sample data, the model ginseng is adjusted
Number;When the precision of prediction of the flux deepness learning model reaches preset threshold, according to the determination of corresponding model parameter
Prediction model.
Wherein, model parameter includes: hidden layer quantity and learning rate etc..
Specifically, the above-mentioned implementation for adjusting model parameter according to sample data can be with are as follows:
According to the sample data, the extent of deviation of the flux deepness learning model calculated result is calculated;According to described
The extent of deviation of calculated result adjusts the model parameter.
Wherein, the sample data at M moment of past includes at least the data at 2N moment, time locating for the top n moment
Section is corresponding with the period locating in the past in S301 at N number of moment, period locating for rear N number of moment and the future to be predicted
Period locating for N number of moment is corresponding.
Specifically, in conjunction with table 1, if it is desired that when with 2018-8-12 20:00 to 2018-8-12 22:00 include 9
The data on flows at 9 moment, pre- in training in the data on flows prediction 2018-8-12 22:00 to 2018-8-12 24:00 at quarter
When surveying model, can using the data of 20:00 before 2018-8-12 to 22:00 and 22:00 to 24:00 as sample data, than
Such as: extremely by the data at 2018-8-11 20:00 to 2018-8-11 22:00 9 moment for including and 2018-8-11 20:00
The data at 9 moment that 2018-8-11 22:00 includes are as sample data.
It is illustrated below with reference to process of the target network element shown in table 1 to above-mentioned S302-S304:
Assuming that use target network element in the stream at 2018-8-12 20:00 to 2018-8-12 22:00 9 moment for including
The data on flows for measuring 9 moment in data prediction 2018-8-12 22:00 to 2018-8-12 24:00, need to train in advance and obtain
Corresponding prediction model extracts target network element at the past M specifically, the training process can be first, from database
The sample data at quarter, such as: the sample data extracted is the stream of 2018-01-01 to 2018-8-11 daily 20:00 to 22:00
Measure data and the data on flows of 2018-01-01 to 2018-8-11 daily 22:00 to 24:00.Then, establish and be based on the time
The flux deepness learning model of sequence, and initialize the model parameter of flow deep learning model.Then, extremely by certain day 20:00
In the data on flows input flow rate deep learning model of 22:00, the correspondence 22:00 that flux deepness learning model is predicted is calculated
Extent of deviation between the data on flows of the 22:00 to 24:00 arrived to the data on flows and said extracted of 24:00, and then basis
Model parameter is adjusted in the extent of deviation.Finally, working as the data on flows and propose that flux deepness learning model is predicted
When the extent of deviation between sample data got is less than preset threshold, the prediction mould is determined according to corresponding model parameter
Type.
S305, data on flows and the actual monitoring data at following N number of moment according to N number of moment in future, determine institute
State the link load state of target network element.
S306, the link load state according to the target network element, determine the quality of service of the target network element.
Wherein, the implementation of S305-S306 can be found in the S203 in above-described embodiment, and details are not described herein by the present invention.
Quality of service monitoring method provided in this embodiment, describes the training process of prediction model, due to prediction model
It is to be obtained according to the training of the sample data of target network element, may learn N number of moment before and after the data on flows of target network element
Changing character therefore pass through the data on flows more closing to reality at N number of moment in future that the prediction model is predicted
Data on flows, the data on flows predicted according to prediction model are higher come the quality of service accuracy determined.
Fig. 4 is the structural schematic diagram of quality of service monitoring device provided by the invention.As shown in figure 4, provided by the invention
Quality of service monitoring device, comprising:
Module 401 is obtained, for obtaining target network element in the data on flows at N number of moment in past, the data on flows includes:
It flows into the flow of the target network element and flows out at least one of the flow of the target network element, wherein N is to be greater than or wait
In 1 integer;
Prediction module 402, for prediction model trained according to the data on flows at N number of moment in the past and in advance, in advance
The target network element is surveyed in the data on flows at following N number of moment;
Determining module 403, for the data on flows and the actual monitoring at following N number of moment according to N number of moment in future
Data determine the quality of service of the target network element.
Optionally, the acquisition module 401, is also used to:
The target network element is extracted from database in the sample data at M moment of past;
According to the sample data, training obtains the prediction model.
Optionally, the acquisition module 401, is specifically used for:
According to the sample data, the flux deepness learning model based on time series is established;
The flux deepness learning model is trained, the prediction model is obtained.
Optionally, the acquisition module 401, is specifically used for:
Initialize the model parameter of the flux deepness learning model;
According to the sample data, the model parameter is adjusted;
When the precision of prediction of the flux deepness learning model reaches preset threshold, determined according to corresponding model parameter
The prediction model.
Optionally, the acquisition module 401, is specifically used for:
According to the sample data, the extent of deviation of the flux deepness learning model calculated result is calculated;
According to the extent of deviation of the calculated result, the model parameter is adjusted.
Optionally, the determining module 403, is specifically used for:
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target is determined
The link load state of network element;
According to the link load state of the target network element, the quality of service of the target network element is determined.
Quality of service monitoring device provided in this embodiment can be used for executing the quality of service of any of the above-described embodiment description
Monitoring method, it is similar that the realization principle and technical effect are similar, and details are not described herein.
Fig. 5 is the hardware structural diagram of network management system provided by the invention.As shown in figure 5, the network management system of the present embodiment
System may include:
Memory 501, for storing program instruction.
Processor 502, for being performed the quality of service for realizing any of the above-described embodiment description in described program instruction
Monitoring method, specific implementation principle can be found in above-described embodiment, and details are not described herein again for the present embodiment.
The present invention provides a kind of computer readable storage medium, is stored thereon with computer program, the computer program
The quality of service monitoring method of any of the above-described embodiment description is realized when being executed by processor.
The present invention also provides a kind of program product, described program product includes computer program, and the computer program is deposited
In readable storage medium storing program for executing, at least one processor can read the computer program, institute from the readable storage medium storing program for executing for storage
It states at least one processor and executes the business matter that the computer program makes network management system implement any of the above-described embodiment description
Monitoring method is measured,
In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit
Letter connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one
In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) or processor (English: processor) execute this hair
The part steps of bright each embodiment the method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory
(English: Read-Only Memory, abbreviation: ROM), random access memory (English: Random Access Memory, letter
Claim: RAM), the various media that can store program code such as magnetic or disk.
It should be understood that processor can be central processing unit (English: Central Processing Unit, referred to as:
CPU), it can also be other general processors, digital signal processor (English: Digital Signal Processor, letter
Claim: DSP), specific integrated circuit (English: Application Specific Integrated Circuit, referred to as: ASIC)
Deng.General processor can be microprocessor or the processor is also possible to any conventional processor etc..In conjunction with the application
The step of disclosed method, can be embodied directly in hardware processor and execute completion, or with the hardware and software in processor
Block combiner executes completion.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations;To the greatest extent
Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to
So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into
Row equivalent replacement;And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (10)
1. a kind of quality of service monitoring method characterized by comprising
Data on flows of the target network element at N number of moment in past is obtained, the data on flows includes: the stream for flowing into the target network element
Measure and flow out at least one of the flow of the target network element, wherein N is the integer more than or equal to 1;
Prediction model trained according to the data on flows at N number of moment in the past and in advance, predicts the target network element in the following N
The data on flows at a moment;
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target network element is determined
Quality of service.
2. the method according to claim 1, wherein the data on flows according to N number of moment in the past and
Trained prediction model in advance predicts the target network element before the data on flows at following N number of moment, further includes:
The prediction model is obtained, the prediction model is used to indicate the target network element in the data on flows at N number of moment in past
With the relationship of the data on flows at following N number of moment.
3. according to the method described in claim 2, it is characterized in that, described obtain the prediction model, comprising:
The target network element is extracted from database in the sample data at M moment of past;
According to the sample data, training obtains the prediction model.
4. according to the method described in claim 3, training obtains described pre- it is characterized in that, described according to the sample data
Survey model, comprising:
According to the sample data, the flux deepness learning model based on time series is established;
The flux deepness learning model is trained, the prediction model is obtained.
5. according to the method described in claim 4, it is characterized in that, described be trained the flux deepness learning model,
Obtain the prediction model, comprising:
Initialize the model parameter of the flux deepness learning model;
According to the sample data, the model parameter is adjusted;
When the precision of prediction of the flux deepness learning model reaches preset threshold, according to the determination of corresponding model parameter
Prediction model.
6. according to the method described in claim 5, adjusting the model ginseng it is characterized in that, described according to the sample data
Number, comprising:
According to the sample data, the extent of deviation of the flux deepness learning model calculated result is calculated;
According to the extent of deviation of the calculated result, the model parameter is adjusted.
7. method according to claim 1-6, which is characterized in that the stream according to N number of moment in future
The actual monitoring data for measuring data and following N number of moment, determine the quality of service of the target network element, comprising:
According to the actual monitoring data of the data on flows at N number of moment in future and following N number of moment, the target network element is determined
Link load state;
According to the link load state of the target network element, the quality of service of the target network element is determined.
8. a kind of quality of service monitoring device characterized by comprising
Module is obtained, for obtaining target network element in the data on flows at N number of moment in past, the data on flows includes: to flow into institute
It states the flow of target network element and flows out at least one of the flow of the target network element, wherein N is more than or equal to 1
Integer;
Prediction module predicts the mesh for prediction model trained according to the data on flows at N number of moment in the past and in advance
Network element is marked in the data on flows at following N number of moment;
Determining module, for the data on flows and the actual monitoring data at following N number of moment according to N number of moment in future, really
The quality of service of the fixed target network element.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
Claim 1-7 described in any item methods are realized when processor executes.
10. a kind of network management system characterized by comprising
Processor;And
Memory, for storing the executable instruction of the processor;
Wherein, the processor is configured to realize that claim 1-7 is described in any item via the executable instruction is executed
Method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910280368.0A CN109995592A (en) | 2019-04-09 | 2019-04-09 | Quality of service monitoring method and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910280368.0A CN109995592A (en) | 2019-04-09 | 2019-04-09 | Quality of service monitoring method and equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109995592A true CN109995592A (en) | 2019-07-09 |
Family
ID=67132629
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910280368.0A Pending CN109995592A (en) | 2019-04-09 | 2019-04-09 | Quality of service monitoring method and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109995592A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111262728A (en) * | 2020-01-08 | 2020-06-09 | 国网福建省电力有限公司 | Flow load monitoring system based on log port flow |
CN111478981A (en) * | 2019-09-27 | 2020-07-31 | 腾讯科技(深圳)有限公司 | Network element determination method and device |
CN112769733A (en) * | 2019-11-05 | 2021-05-07 | 中国电信股份有限公司 | Network early warning method, device and computer readable storage medium |
CN113923096A (en) * | 2020-06-22 | 2022-01-11 | 中国联合网络通信集团有限公司 | Network element fault early warning method and device, electronic equipment and storage medium |
CN114266300A (en) * | 2021-12-16 | 2022-04-01 | 中国联合网络通信集团有限公司 | Feature prediction model training method and device and core network service anomaly detection method and device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120303413A1 (en) * | 2011-05-27 | 2012-11-29 | Vpisystems Inc. | Methods and systems for network traffic forecast and analysis |
CN107124320A (en) * | 2017-06-30 | 2017-09-01 | 北京金山安全软件有限公司 | Traffic data monitoring method and device and server |
CN108494746A (en) * | 2018-03-07 | 2018-09-04 | 长安通信科技有限责任公司 | A kind of network port Traffic anomaly detection method and system |
CN108880888A (en) * | 2018-06-20 | 2018-11-23 | 浙江工商大学 | A kind of SDN network method for predicting based on deep learning |
CN108985446A (en) * | 2018-07-24 | 2018-12-11 | 百度在线网络技术(北京)有限公司 | method and device for alarm |
-
2019
- 2019-04-09 CN CN201910280368.0A patent/CN109995592A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120303413A1 (en) * | 2011-05-27 | 2012-11-29 | Vpisystems Inc. | Methods and systems for network traffic forecast and analysis |
CN107124320A (en) * | 2017-06-30 | 2017-09-01 | 北京金山安全软件有限公司 | Traffic data monitoring method and device and server |
CN108494746A (en) * | 2018-03-07 | 2018-09-04 | 长安通信科技有限责任公司 | A kind of network port Traffic anomaly detection method and system |
CN108880888A (en) * | 2018-06-20 | 2018-11-23 | 浙江工商大学 | A kind of SDN network method for predicting based on deep learning |
CN108985446A (en) * | 2018-07-24 | 2018-12-11 | 百度在线网络技术(北京)有限公司 | method and device for alarm |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111478981A (en) * | 2019-09-27 | 2020-07-31 | 腾讯科技(深圳)有限公司 | Network element determination method and device |
CN112769733A (en) * | 2019-11-05 | 2021-05-07 | 中国电信股份有限公司 | Network early warning method, device and computer readable storage medium |
CN111262728A (en) * | 2020-01-08 | 2020-06-09 | 国网福建省电力有限公司 | Flow load monitoring system based on log port flow |
CN113923096A (en) * | 2020-06-22 | 2022-01-11 | 中国联合网络通信集团有限公司 | Network element fault early warning method and device, electronic equipment and storage medium |
CN113923096B (en) * | 2020-06-22 | 2023-05-30 | 中国联合网络通信集团有限公司 | Network element fault early warning method and device, electronic equipment and storage medium |
CN114266300A (en) * | 2021-12-16 | 2022-04-01 | 中国联合网络通信集团有限公司 | Feature prediction model training method and device and core network service anomaly detection method and device |
CN114266300B (en) * | 2021-12-16 | 2023-01-24 | 中国联合网络通信集团有限公司 | Feature prediction model training method and device and core network service anomaly detection method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109995592A (en) | Quality of service monitoring method and equipment | |
CN109508789A (en) | Predict method, storage medium, processor and the equipment of hands | |
CN110503207A (en) | Federation's study credit management method, device, equipment and readable storage medium storing program for executing | |
CN109583468A (en) | Training sample acquisition methods, sample predictions method and corresponding intrument | |
CN106651097A (en) | Data collection method, data collection device and data collection server based on crowd sourcing | |
CN106789844B (en) | Malicious user identification method and device | |
CN108960719A (en) | Selection method and apparatus and computer readable storage medium | |
CN108985954A (en) | A kind of method and relevant device of incidence relation that establishing each mark | |
CN104992347A (en) | Video matching advertisement method and device | |
CN109345200A (en) | Automatic auditing method and device, computer readable storage medium based on high-volume report | |
CN107767055A (en) | A kind of mass-rent result assemblage method and device based on collusion detection | |
CN112884289B (en) | Weapon equipment combination selection method and system based on system contribution rate | |
CN109615015A (en) | A kind of data preprocessing method based on block chain intelligence contract and machine learning | |
CN110019996A (en) | A kind of family relationship recognition methods and system | |
CN108259238A (en) | The method that network function service chaining maps in data center | |
CN107967204A (en) | Line pushes method, system and the terminal device surveyed | |
DE102016011527B4 (en) | Machine learning apparatus and method for learning an arrangement position of a magnet in a rotor, and rotor design apparatus including the machine learning apparatus | |
CN110991871A (en) | Risk monitoring method, device, equipment and computer readable storage medium | |
CN107689956A (en) | The intimidation estimating method and device of a kind of anomalous event | |
CN106910075B (en) | The intelligent processing system and method that client mobile communication is complained | |
Klement et al. | Using SOM in the performance monitoring of the emergency call-taking system | |
CN104092503A (en) | Artificial neural network spectrum sensing method based on wolf pack optimization | |
CN110138613A (en) | A kind of network test template construction method, electric power networks test method and relevant apparatus | |
CN107016460A (en) | User changes planes Forecasting Methodology and device | |
CN107832848A (en) | application management method, device, storage medium and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190709 |
|
RJ01 | Rejection of invention patent application after publication |