CN113890837B - Method and system for predicting index degradation based on sliding window cross algorithm - Google Patents
Method and system for predicting index degradation based on sliding window cross algorithm Download PDFInfo
- Publication number
- CN113890837B CN113890837B CN202111065844.0A CN202111065844A CN113890837B CN 113890837 B CN113890837 B CN 113890837B CN 202111065844 A CN202111065844 A CN 202111065844A CN 113890837 B CN113890837 B CN 113890837B
- Authority
- CN
- China
- Prior art keywords
- sliding window
- prediction model
- data
- prediction
- indexes
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000015556 catabolic process Effects 0.000 title claims abstract description 19
- 238000006731 degradation reaction Methods 0.000 title claims abstract description 19
- 238000004891 communication Methods 0.000 claims abstract description 35
- 230000001052 transient effect Effects 0.000 claims 4
- 238000013480 data collection Methods 0.000 claims 1
- 238000007405 data analysis Methods 0.000 abstract description 2
- 208000018910 keratinopathic ichthyosis Diseases 0.000 description 25
- 238000013178 mathematical model Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000006866 deterioration Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- 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
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- 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
Abstract
The invention discloses a method and a system for predicting index degradation based on a sliding window cross algorithm, relating to the technical field of data analysis; the method comprises the steps of taking the time point of finishing network operation as a demarcation point, taking historical data of the forward demarcation point as modeling data, establishing an instantaneous prediction model and a sliding window prediction model based on a sliding window crossing algorithm, respectively obtaining prediction results through the instantaneous prediction model and the sliding window prediction model, taking data of the backward section of the demarcation point as data to be compared, and judging whether indexes of the communication network KPI are degraded or not through comparing the prediction results with the data to be compared.
Description
Technical Field
The invention discloses a method and a system, relates to the technical field of data analysis, and particularly relates to a method and a system for predicting index degradation based on a sliding window cross algorithm.
Background
KPIs in communications, often referred to as key performance indicators. The contents are as follows: the call drop rate, the call completing rate, the data service downloading rate, the field test of the network quality, the leading degree of competitors, the completion condition of the network complaint total amount in the same ratio and the like.
After the network operation is completed, the communication network KPI is manually checked and analyzed, so that problems can not be found timely and accurately, and huge hidden dangers are brought to network operation and maintenance.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides the method and the system for predicting the index degradation based on the sliding window cross algorithm, and the method and the system have the characteristics of strong universality, simple and convenient implementation and the like, and have wide application prospects.
The specific scheme provided by the invention is as follows:
a method for predicting index degradation based on sliding window crossing algorithm includes using time point of network operation as boundary point, using historical data of boundary point forward as modeling data, establishing instantaneous prediction model and sliding window prediction model based on sliding window crossing algorithm, obtaining prediction results through instantaneous prediction model and sliding window prediction model respectively,
and taking the data of the demarcation point for a period of time later as the data to be compared, and judging whether the index of the communication network KPI is degraded or not by comparing the prediction result with the data to be compared.
Further, in the method for predicting index degradation based on the sliding window crossing algorithm, a time period before a demarcation point is divided, historical data is obtained according to the time period to serve as modeling data, a sliding window prediction model based on the sliding window crossing algorithm is established according to the modeling data obtained from the time period divided by days, and an instantaneous prediction model is established according to the modeling data obtained from the time period divided by hours.
Furthermore, in the method for predicting the index degradation based on the sliding window cross algorithm, the prediction result is determined by combining a sliding window prediction model and an instantaneous prediction model with a confidence interval algorithm.
Further, the method for predicting the index degradation based on the sliding window crossing algorithm includes:
if the indexes are larger and worse, the indexes of the communication network KPI behind the boundary point are greater than Max (the prediction result of the sliding window prediction model and the prediction result of the instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated;
if the indexes are smaller and worse, the indexes of the communication network KPI behind the boundary point are less than Max (prediction result of a sliding window prediction model and prediction result of an instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated.
The system for predicting index degradation based on the sliding window cross algorithm comprises a data acquisition module, a model building module and a comparison module,
the data acquisition module takes the time point of completing the network operation as a demarcation point, takes the historical data before the demarcation point as modeling data, takes the data after the demarcation point for a period of time as data to be compared,
the model building module builds an instantaneous prediction model and a sliding window prediction model based on a sliding window crossing algorithm, obtains prediction results through the instantaneous prediction model and the sliding window prediction model respectively,
the comparison module judges whether the indicator of the communication network KPI is degraded or not by comparing the prediction result with the data to be compared.
Further, in the system for predicting index degradation based on the sliding window crossing algorithm, the data acquisition module divides a time period before a demarcation point, acquires historical data as modeling data according to the time period, the model establishment module establishes a sliding window prediction model based on the sliding window crossing algorithm according to the modeling data acquired by the time period divided by days, and establishes an instantaneous prediction model according to the modeling data acquired by the time period divided by hours.
Further, the model building module in the system for predicting index degradation based on the sliding window cross algorithm determines a prediction result through a sliding window prediction model and an instantaneous prediction model in combination with a confidence interval algorithm.
Further, the judging of degradation in the comparison module in the system for predicting index degradation based on the sliding window crossing algorithm includes:
if the indexes are larger and worse, the indexes of the communication network KPI behind the boundary point are greater than Max (the prediction result of the sliding window prediction model and the prediction result of the instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated;
if the indexes are smaller and worse, the indexes of the communication network KPI behind the boundary point are less than Max (prediction result of a sliding window prediction model and prediction result of an instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated.
The invention has the advantages that:
the invention provides a method for predicting index degradation based on a sliding window crossing algorithm, which is characterized in that an instantaneous prediction model and a sliding window prediction model based on the sliding window crossing algorithm are utilized to respectively obtain prediction results, whether the index of a communication network KPI is degraded or not is judged by comparing the prediction results with data to be compared, the communication network KPI can be further checked and analyzed, problems can be timely and accurately found, and the network operation can be efficiently, comprehensively and accurately monitored.
Drawings
FIG. 1 is a schematic diagram illustrating a sliding window prediction method according to the present invention.
FIG. 2 is a diagram showing the prediction of PRB utilization applied in the method of the present invention.
FIG. 3 is a schematic flow diagram of the process of the present invention.
Detailed Description
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
The invention provides a method for predicting index degradation based on a sliding window crossing algorithm, which takes the time point of completing network operation as a demarcation point, takes historical data of the forward demarcation point as modeling data, establishes an instantaneous prediction model and a sliding window prediction model based on the sliding window crossing algorithm, respectively obtains prediction results through the instantaneous prediction model and the sliding window prediction model,
and taking the data of the demarcation point for a period of time later as the data to be compared, and judging whether the index of the communication network KPI is degraded or not by comparing the prediction result with the data to be compared.
According to the method, the instantaneous prediction model and the sliding window prediction model based on the sliding window crossing algorithm are used for respectively obtaining the prediction results, whether the indexes of the communication network KPI are degraded or not is judged by comparing the prediction results with the data to be compared, the communication network KPI can be further checked and analyzed, the problems can be timely and accurately found, and the network operation can be efficiently, comprehensively and accurately monitored.
In a specific application, in some embodiments of the method of the present invention, the specific implementation process is as follows:
and acquiring data, wherein the time point of finishing network operation is a demarcation point, a time period before the demarcation point is divided, historical data of 14 days before the demarcation point and 24 hours after the demarcation point is used as modeling data, and data of 12 hours after the demarcation point is used as data to be compared.
According to 14 days by 24 hours historical data divided by days, a sliding window prediction model based on a sliding window crossing algorithm is established, an instantaneous prediction model is established according to modeling data obtained 7 hours before a boundary point, wherein communication network KPI data takes one week time as a period, 24 rows by 8 columns of sliding windows shown in a thick frame in figure 1 are referred, one column is moved to the right each time, and finally (6 rows by 24) 8 columns of modeling data are generated. The obtained modeling data are 8 columns in total, the first 7 columns are input variables, and the last column is used for establishing a linear regression model for the target data. The obtained sliding window prediction model takes 7 data at the same time of the demarcation point of the past 7 days as input variables, outputs the prediction value 12 hours after the operation demarcation point,
the instantaneous mathematical model is realized by a gray prediction algorithm, the gray prediction algorithm is suitable for short-term and instantaneous prediction with small data quantity, and the method is a method for modeling by accumulating and generating original data to obtain an approximate exponential law. The instantaneous mathematical model takes the previous 7 hours of the demarcation point as an input variable and outputs a predicted value one hour after the demarcation point.
And determining a prediction result by combining a sliding window prediction model and an instantaneous prediction model with a confidence interval algorithm respectively. Wherein the confidence interval algorithm
The confidence interval algorithm is as follows:
wherein S is the standard deviation of the input parameters, n is the number of the input parameters, 1.96 is the coefficient corresponding to the confidence interval, b is the prediction bias,
the larger the service is considered to be, the more deteriorated the index is, the prediction result = prediction value + prediction bias;
the smaller the traffic hypothesis index is, the more deteriorated, the prediction result = prediction value-prediction bias.
The 12-hour prediction result and the 1-hour prediction result of the communication network KPI are obtained through the processes.
When the deterioration judgment is performed:
if the service considers that the index is larger and worse, and the KPI value at the next moment after the operation of the demarcation point is larger than Max (Y prediction result (sliding window model) and Y prediction result (instantaneous mathematical model)), the index is considered to be naturally deteriorated and is judged to be caused by non-operation;
if the service is considered that the index is deteriorated as the smaller the index is, the KPI value < Max (Y prediction result (sliding window model) or Y prediction result (instantaneous mathematical model)) at the next time after the operation of the demarcation point is considered to be naturally deteriorated, and it is determined that the index is not operated.
If the point of demarcation is not naturally deteriorated at the next moment, comparing the KPI value of the 12-hour communication network with the 12-hour prediction result predicted by the sliding window model, and if more than 8 hours in the 12 hours are index deterioration, judging that the operation is caused.
The method can be used for prediction display through the utilization rate of the PRB, for example, taking an operation log of a local 5G communication network as an example, a prediction display graph of the utilization rate of the upstream PRB of the example is shown in figure 2, the horizontal axis of a coordinate is the hours, namely time, the vertical axis of the coordinate is the utilization rate percentage, the index is more degraded as the index is smaller, wherein the short-point intercross line is a predicted value, the short-dotted line is a predicted result after a confidence interval algorithm, the solid line is data 12 hours after the operation, and the algorithm judges that the operation is degraded.
Meanwhile, the invention also provides a system for predicting index degradation based on the sliding window crossing algorithm, which comprises a data acquisition module, a model establishing module and a comparison module,
the data acquisition module takes the time point of completing the network operation as a demarcation point, takes the historical data before the demarcation point as modeling data, takes the data after the demarcation point for a period of time as data to be compared,
the model building module builds an instantaneous prediction model and a sliding window prediction model based on a sliding window crossing algorithm, obtains prediction results through the instantaneous prediction model and the sliding window prediction model respectively,
the comparison module judges whether the indicator of the communication network KPI is degraded or not by comparing the prediction result with the data to be compared.
The information interaction, execution process and other contents between the modules in the system are based on the same concept as the method embodiment of the present invention, and specific contents can be referred to the description in the method embodiment of the present invention, and are not described herein again.
Similarly, the system can respectively obtain the prediction results by utilizing the instantaneous prediction model and the sliding window prediction model based on the sliding window crossing algorithm, judge whether the indexes of the communication network KPI are degraded or not by comparing the prediction results with the data to be compared, further check and analyze the communication network KPI, find problems timely and accurately, and efficiently, comprehensively and accurately monitor the network operation.
It should be noted that not all steps and modules in the processes and system structures in the preferred embodiments are necessary, and some steps or modules may be omitted according to actual needs. The execution order of the steps is not fixed and can be adjusted as required. The system structure described in the above embodiments may be a physical structure or a logical structure, that is, some modules may be implemented by the same physical entity, or some modules may be implemented by a plurality of physical entities, or some components in a plurality of independent devices may be implemented together.
The above-mentioned embodiments are merely preferred embodiments for fully illustrating the present invention, and the scope of the present invention is not limited thereto. The equivalent substitutions or changes made by the person skilled in the art on the basis of the present invention are all within the protection scope of the present invention. The protection scope of the invention is subject to the claims.
Claims (6)
1. A method for predicting index degradation based on sliding window crossing algorithm is characterized in that a time point for completing network operation is taken as a demarcation point, historical data of the demarcation point forward is taken as modeling data, an instantaneous prediction model and a sliding window prediction model based on the sliding window crossing algorithm are established, prediction results are respectively obtained through the instantaneous prediction model and the sliding window prediction model,
taking data of a later period of time of the demarcation point as data to be compared, and judging whether the index of the KPI of the communication network is degraded or not by comparing the prediction result with the data to be compared, wherein the judging step comprises the following steps:
if the indexes are larger and worse, the indexes of the communication network KPI behind the boundary point are greater than Max (the prediction result of the sliding window prediction model and the prediction result of the instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated;
if the indexes are smaller and worse, the indexes of the communication network KPI behind the boundary point are less than Max (prediction result of a sliding window prediction model and prediction result of an instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated.
2. The method of claim 1, wherein a time period before a demarcation point is divided, and historical data is acquired as modeling data according to the time period, wherein a sliding window prediction model based on a sliding window crossing algorithm is established according to the modeling data acquired in the time period divided by days, and a transient prediction model is established according to the modeling data acquired in the time period divided by hours.
3. The method of claim 1 or 2, wherein the prediction result is determined by combining a sliding window prediction model and a transient prediction model with a confidence interval algorithm, respectively.
4. A system for predicting index degradation based on a sliding window cross algorithm is characterized by comprising a data acquisition module, a model building module and a comparison module,
the data acquisition module takes the time point of completing the network operation as a demarcation point, takes the historical data before the demarcation point as modeling data, takes the data after the demarcation point for a period of time as data to be compared,
the model building module builds an instantaneous prediction model and a sliding window prediction model based on a sliding window crossing algorithm, obtains prediction results through the instantaneous prediction model and the sliding window prediction model respectively,
the comparison module judges whether the indicator of the communication network KPI is degraded or not by comparing the prediction result with the data to be compared, and comprises the following steps:
if the indexes are larger and worse, the indexes of the communication network KPI behind the boundary point are greater than Max (the prediction result of the sliding window prediction model and the prediction result of the instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated;
if the indexes are smaller and worse, the indexes of the communication network KPI behind the boundary point are less than Max (prediction result of a sliding window prediction model and prediction result of an instantaneous prediction model), the communication network KPI indexes are considered to be naturally deteriorated.
5. The system according to claim 4, wherein the data collection module divides a time period before the demarcation point, and acquires historical data as modeling data according to the time period, and wherein the model creation module creates a sliding window prediction model based on the sliding window crossing algorithm based on the modeling data acquired from the time period divided by day, and creates a transient prediction model based on the modeling data acquired from the time period divided by hour.
6. The system according to claim 4 or 5, wherein the model building module determines the prediction result by combining the sliding window prediction model and the transient prediction model with the confidence interval algorithm, respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111065844.0A CN113890837B (en) | 2021-09-13 | 2021-09-13 | Method and system for predicting index degradation based on sliding window cross algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111065844.0A CN113890837B (en) | 2021-09-13 | 2021-09-13 | Method and system for predicting index degradation based on sliding window cross algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113890837A CN113890837A (en) | 2022-01-04 |
| CN113890837B true CN113890837B (en) | 2023-03-24 |
Family
ID=79008997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111065844.0A Active CN113890837B (en) | 2021-09-13 | 2021-09-13 | Method and system for predicting index degradation based on sliding window cross algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113890837B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112257901A (en) * | 2020-09-24 | 2021-01-22 | 北京航天测控技术有限公司 | Abnormity early warning method and device for spacecraft |
| US11063842B1 (en) * | 2020-01-10 | 2021-07-13 | Cisco Technology, Inc. | Forecasting network KPIs |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8676964B2 (en) * | 2008-07-31 | 2014-03-18 | Riverbed Technology, Inc. | Detecting outliers in network traffic time series |
| CN103178990A (en) * | 2011-12-20 | 2013-06-26 | 中国移动通信集团青海有限公司 | Network device performance monitoring method and network management system |
| CN105512741A (en) * | 2014-09-26 | 2016-04-20 | 山西云智慧科技股份有限公司 | Bus passenger traffic combined prediction method |
| CN110086649B (en) * | 2019-03-19 | 2023-06-16 | 深圳壹账通智能科技有限公司 | Abnormal flow detection method, device, computer equipment and storage medium |
| CN113065678A (en) * | 2019-12-13 | 2021-07-02 | 中兴通讯股份有限公司 | Performance index early warning method, device, equipment and storage medium |
| CN113099476B (en) * | 2021-05-13 | 2022-12-06 | 中国联合网络通信集团有限公司 | Network quality detection method, device, equipment and storage medium |
-
2021
- 2021-09-13 CN CN202111065844.0A patent/CN113890837B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11063842B1 (en) * | 2020-01-10 | 2021-07-13 | Cisco Technology, Inc. | Forecasting network KPIs |
| CN112257901A (en) * | 2020-09-24 | 2021-01-22 | 北京航天测控技术有限公司 | Abnormity early warning method and device for spacecraft |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113890837A (en) | 2022-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537544B (en) | Real-time monitoring method and monitoring system for transaction system | |
| CN107204894B (en) | Method and device for monitoring network service quality | |
| CN109656793A (en) | A kind of information system performance stereoscopic monitoring method based on multi-source heterogeneous data fusion | |
| CN101808351B (en) | Method and system for business impact analysis | |
| EP3349126B1 (en) | Method, device, storage medium, and apparatus for automatically discovering fuel station poi | |
| CN102082703A (en) | Method and device for monitoring equipment performance of service supporting system | |
| CN102487523B (en) | User compliant analysis method and device | |
| CN101442762A (en) | Method and apparatus for analyzing network performance and locating network fault | |
| CN111915143B (en) | Complex product assembly management and control system based on intelligent contract | |
| EP3591894B1 (en) | Tariff data determination method and device | |
| CN103730118A (en) | Voice signal collecting method and mobile terminal | |
| CN106372134A (en) | Internet of vehicles real-time data processing method and system | |
| CN110134040B (en) | Method and system for processing operation data of industrial equipment | |
| CN110730100B (en) | Alarm information processing method and device and server | |
| CN113890837B (en) | Method and system for predicting index degradation based on sliding window cross algorithm | |
| CN112787878A (en) | Network index prediction method and electronic equipment | |
| CN114841843A (en) | Method and system for analyzing suspicious green traffic | |
| CN109995549B (en) | Method and device for evaluating flow value | |
| CN114283590A (en) | Traffic flow peak prediction method and device and electronic equipment | |
| CN111614520B (en) | IDC flow data prediction method and device based on machine learning algorithm | |
| CN116109007B (en) | Power generation power determination method, server and storage medium | |
| CN102547789B (en) | Early warning method, device and system for quality of peer-to-peer service | |
| CN114661700A (en) | Artificial influence weather operation effect inspection method based on AI | |
| CN113743025A (en) | Meteorological forecast data correction method, device, electronic equipment and storage medium | |
| EP2947910A2 (en) | Performance optimizations for wireless access points |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |