CN101031125A - Method for accounting base-station performance data - Google Patents
Method for accounting base-station performance data Download PDFInfo
- Publication number
- CN101031125A CN101031125A CNA2006100587069A CN200610058706A CN101031125A CN 101031125 A CN101031125 A CN 101031125A CN A2006100587069 A CNA2006100587069 A CN A2006100587069A CN 200610058706 A CN200610058706 A CN 200610058706A CN 101031125 A CN101031125 A CN 101031125A
- Authority
- CN
- China
- Prior art keywords
- performance
- data
- change
- performance data
- rate
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The method comprises: a) the performance statistics agent module running on the base station and switching device monitors and records the working status of system at certain time interval, and sends the recorded performance data to the performance management server; the performance management server receives the data and saves it in the database; b) according to the collected performance data of continuity attributes value, making calculation for its variance ratio in order to generate discretization data which can be processed by data mining technology; making data mining and relevant analysis for the discretization data to get the performance-based rule and knowledge; d) making visualization for said rule and knowledge to get a knowledge form which user can understand.
Description
Technical field
The present invention relates to a kind of method of accounting base-station performance data, be particularly useful for analyzing so that carry out trend analysis and prediction for performance statistic, performance statistics is the environmental data of acquisition system operation on base station and the switching equipment, then with these storage in database server side.
Background technology
At present, it is low that a lot of fields all exist data user rate, can't analytic trend, and defective and problem such as can't predict, and in performance statistics, this problem is particularly outstanding.To with the performance statistics actual conditions that example is set forth this problem below.
Performance management can be gathered, handle measurement data, according to measurement result, takes necessary network management control action, improves the overall performance level of network.The basis of performance statistics is the base station and the switching equipment of actual motion, the ruuning situation of these equipment is most important for the analysis of system, therefore, need with performance data from the operation the base station and switching equipment extract, be saved in the database server then, when need analyze some part of equipment, just the plurality of data that extracts wherein from database carries out statistics and analysis.So the function of performance management comprises performance data collection and performance management action two large divisions.Performance data collection is meant that long-term collection can reflect the performance data of aspects such as service quality, resource use, availability, and these data can reflect the operation conditions of system.The performance management action then is to carry out statistics and analysis to gathering the performance statistic that comes up.By to performance data long-time statistical analysis report, can be used for the network planning and construction, up-to-date measurement statistical report can be used for instructing takes suitable performance management action.
For performance statistics, the function of mainly finishing has the following aspects:
1, the collection of performance data: performance data is actually the true reflection of equipment operation situation, and the ruuning situation of awareness network if desired just must be carried out the collecting work of initial data on equipment.The collection point all is in the equipment operation flow process, point important and that enough cause user's interest and concern.
2, the storage of performance data: finish after the collection of data, the performance statistic that collects need be preserved.Data can be kept at the medium that can store data arbitrarily, and the purpose of preservation is and can analyzes the ruuning situation of system in the later stage.
3, performance data statistics and analysis: the purpose of performance data storage still is and can carries out statistics and analysis to performance data at any time, so that can understand and analyze for the history run situation of equipment.In fact, the work of this part can be divided into two parts, statistics and analysis.Statistical work mainly is to carry after data report and store, and the form of original performance data with performance index embodied, be with data directly or through showing indirectly after simple the calculating, embody the operation conditions of network.Analytical work mainly is to carry out on the basis of statistics, is after data statistics is come out, and according to the result who adds up the ruuning situation of equipment is analyzed, so that instruct current operation and long-range planning.
Above-mentioned all functions all be based upon conveniently, on reasonable, the reliable and stable basis, but because performance statistic module is finished the particularity and the importance of function, and the result of performance statistics and analysis will directly or indirectly be used for current optimization and long-range planning.Therefore, in existing performance statistic module, can there be the deficiency of following several respects:
1, the performance data utilance is low
The data volume that performance statistics is gathered is very big, all significant process and the environment of having contained the operation of base station and switching system, but these data often not all user all be concerned about, the end product that therefore can cause having expended the data of a lot of times and space acquisition is exactly because oversize and maintained program of holding time is deleted.The utilance of these data will be very low so, do not bring into play the real value of these data.
2, function is very weak
All be to analyze for the ruuning situation of system when the analytic function of performance statistics is most of in conjunction with the different index of certain time period and the knowwhy of base station and switching system.This analysis can combine the theoretical ruuning situation of system and the practical operation situation of system preferably, but this analysis more is by means of analyst's knowledge and analysis level, merely from performance statistic module, we can say does not have real analytic function, does not particularly have the function of trend analysis.Trend analysis is that the emphasis that will analyze concentrates on the data, the knowwhy of system just plays the effect (analyzing the trend of coming out sometimes also can't make an explanation by these knowledge) that a kind of trend is explained in trend analysis, at this moment often can access some trend and knowledge relevant with the traffic situation with practical operation situation, these knowledge parts can add in the knowwhy, and part can be used as a kind of reflection of practical operation situation.
3, forecast function lacks
At present under the operating position, the major function of performance statistics is that the ruuning situation for base station and switching equipment history shows and analyzes, still can't realize forecast function to a certain extent by program.That is to say that always after system had moved, performance statistics just can reflect the ruuning situation of system, and can't predict for the ruuning situation of system in future according to the ruuning situation of device history.
Summary of the invention
The main technical problem to be solved in the present invention provides a kind of method of accounting base-station performance data, can improve the analytic function in the performance statistics.
In order to solve the problems of the technologies described above, the invention provides a kind of method of accounting base-station performance data, may further comprise the steps:
(a) the performance statistics proxy module that operates on base station and the switching equipment is regularly monitored and record the ruuning situation of system, and with the record performance data send to the performance management server, this performance management server will receive data and be kept in the database;
(b) performance data of the connection attribute value that collects is carried out asking for of rate of change, and the performance data rate of change that generates is processed into the data mode of the discretization that data mining can handle;
(c) data of discretization are carried out data mining and correlation analysis, obtaining with the performance data is based rule and knowledge;
(d) rule and the knowledge that obtains for excavation is carried out visualization processing.
Further, said method also can have following characteristics: when described step (a) is carried out asking for of rate of change, be to ask for the rate of change of the performance index of k+1 period with respect to k period performance index, obtain the performance data rate of change of one-dimensional sequence form.
Further, said method also can have following characteristics: in the described step (a), be by the performance data rate of change that generates is carried out preliminary treatment, and/or the performance data rate of change carried out discretization, thereby the performance data rate of change that generates is processed into the data mode that data mining can be handled.
Further, said method also can have following characteristics: described step (a) is carried out preliminary treatment to the performance data rate of change that generates and is comprised basic preliminary treatment and/or special pre-treatment; Wherein, basic preliminary treatment mainly is that the vacancy value and the noise data of performance data rate of change are handled, and special pre-treatment is that the data of performance data rate of change are carried out discretization.
Further, said method also can have following characteristics: described special pre-treatment is that the performance data rate of change with the one-dimensional sequence form is one group with continuous a plurality of periods, generates performance data variation rate two-dimensional sequence, so that carry out the discretization of data.
Further, said method also can have following characteristics: when described step (a) is carried out discretization with the performance data rate of change, be to turn to a plurality of closed intervals with the span of described performance data rate of change two-dimensional sequence is discrete, a numbering is given in each interval, generates sequence of intervals f
1, f
2, f
3..., f
m, then described performance data rate of change two-dimensional sequence is mapped in the interval tabulation that generates above, obtain discretization two-dimensional sequence afterwards.
Further, said method also can have following characteristics: described step (c) is excavated the rule obtain and knowledge and is carried out when visual, comprises visual and graphic form visual of textual form.
Further, said method also can have following characteristics: the knowledge form that described step (b) obtains is to be similar to: b
I-1∧ b
i→ b
I+1The rule of form promptly is illustrated in b
I-1And b
iUnder the prerequisite of Chu Xianing, can derive b simultaneously
I+1Appearance.
The beneficial effect of the inventive method is embodied in following several aspect:
Reduce the loss: among the process of existing system operation, after system went wrong, through performance statistical analysis problem place was so that deal with problems often.And after by data mining algorithm the original performance data being excavated, can analyze the operation trend of system, and can predict the ruuning situation in system future so that prevent trouble before it happens, can be before system goes wrong the appearance of proactive problem.
Save human cost: in the performance statistics of system, what finish more is the statistical function of performance data now, and the work of analysis is mainly by manually finishing.Data mining is applied to after the performance statistics, can partly finish automatic analytical work, but also may find that some are hidden in the data, be not the knowledge form known to the analyst, not only saved human cost, and operation knowledge that can expanding system.
Extensibility: now in the systematic function statistics, after having increased new statistics,, after must the detail knowledge new statistics, these data and analysis result thereof can be applied in the optimization of system for the analyst.And as data mining, as long as increased initial data, can initial data be analyzed, obtain being hidden in the knowledge in the data then, do not need to make any modification.
The utilization of this method, the problem that occurs for the trend of analytical performance statistics, the situation of estimated performance statistics, the system that effectively prevents all has very strong booster action.And this method can also be generalized among the other field.Especially, on the base station operation maintenance terminal that the core technology and the algorithm of data mining is integrated into the performance statistic module place, be applied to base station system for analyze, prediction base station ruuning situation, improve base station running environment, adapt to configuration variation environment flexibly, whole system maintainability and power of test on the customer satisfaction with services, have important practical value.
Description of drawings
Fig. 1 is the performance statistics overall construction drawing;
Fig. 2 is a data-mining module operational flowchart in the performance statistics;
Embodiment
Because the ruuning situation of performance statistic and equipment is closely related, can be according to performance statistic situation analysis and prognoses system ruuning situation, present performance statistics is all very limited on the function of analyzing and predicting.Therefore, how to handle, to analyze, so that can find the operation trend of system and system be predicted the key that becomes optimization method for performance statistic.This scheme that the present invention proposes be with the notion of data mining and algorithm application among performance statistics, by to the processing of performance data with excavate the operation trend of the system that can obtain, ruuning situation to system future is predicted, time saving and energy saving on safeguarding and planning, clear on expression forms of information.
Data mining technology is applied to data excavates and analyze, initial data need possess following feature so:
Data volume is big: digging technology needs lot of data to carry out repeated calculation, so that obtain the knowledge of certain scale that is hidden in the data, rather than the very strong knowledge of the contingency of on small data quantity, obtaining.
Data grow with time: have only when data along with the time increases, can seek knowledge according to the data conditions in early stage, predict following data according to knowledge and current data then.
Performance data is exactly well to satisfy the data that can be applied to the data mining condition, will be example with the performance data below, carries out the description of detailed embodiment in conjunction with the accompanying drawings.
Fig. 1 is the module map of present embodiment, comprises performance statistic acquisition module, performance statistic reporting module, performance statistic storehouse, data-mining module, knowledge base, graph visualization module and text visualization model.When data mining is applied to performance statistics, at first carry out data acquisition, data are reported and store into the performance statistic storehouse then, then application data excavation module is excavated the performance statistic of storage, the result that will excavate is undertaken visual by text or figure at last, with the understandable form of knowledge performance adult, perhaps with knowledge store in knowledge base, so that use later on.
Fig. 2 then is the flow chart that data-mining module is applied to performance statistic, may further comprise the steps:
Step 100, the performance statistics proxy module that operates on base station and the switching equipment is regularly monitored the ruuning situation of system, and ruuning situation is noted, and sends to the performance management server;
Step 105, performance management server are saved in data in the database after receiving data, and these data also are the bases that performance statistics and additive method are analyzed;
Step 110 is carried out asking for of rate of change with the performance data of connection attribute value;
The initial data of performance statistics all is the data of connection attribute value, need analyze and predict for the rate of change of performance index in trend analysis and the prediction, therefore the performance data of connection attribute value need be carried out asking for of rate of change, so that continuous performance data can be reflected as continuous performance data rate of change.The concrete mode of calculating is to ask for the rate of change of the performance index of k+1 period with respect to k period performance index.
In an application example, on the base station operation maintenance terminal that the core technology and the algorithm of data mining is integrated into the performance statistic module place, be that example is illustrated with this terminal to the statistical analysis of call setup success rate below, suppose that original performance data sequence is:
a
1, a
2, a
3..., a
i..., with formation a
iExpression call setup success rate.
Then need be translated into:
(a
2-a
1)/a
1，(a
3-a
2)/a
2，(a
4-a
3)/a
3，......，(a
i+1-a
i)/a
i，......
Be abbreviated as b
1, b
2, b
3..., b
i..., the performance data rate of change sequence of the rate of change form of expression call setup success rate.
Step 120 is carried out preliminary treatment for the performance data rate of change that generates, and mainly is divided into two parts of basic preliminary treatment and special pre-treatment;
Wherein, basic preliminary treatment mainly is that vacancy value and noise data are handled, and special pre-treatment mainly is the important process for last trend analysis that generates and forecasting institute work.In trend analysis and prediction, the final knowledge form that forms is to be similar to: b
I-1∧ b
i→ b
I+1The rule of form, so that analyze the development trend of index a, and infer the number range that i+1 period index b according to the number range of i-1 period and i period index b.
Therefore, need be with top b
1, b
2, b
3..., b
i... the one-dimensional sequence preliminary treatment of form is (b
1, b
2, b
3), (b
2, b
3, b
4), (b
3, b
4, b
5) ..., (b
I-1, b
i, b
I+1) ....The rate of change that is about to the call setup success rate is one group of two-dimensional sequence that generates the two-dimensional sequence form with continuous three periods, so that carry out the discretization of data.
Step 130 will be carried out discretization through pretreated performance data rate of change, so that obtain the data mode that data mining can be handled;
In the above-mentioned example, will turn to m closed interval through the span of pretreated performance data rate of change two-dimensional sequence is discrete, a numbering is given in each interval, generates interval sequence f
1, f
2, f
3..., f
m, i.e. the interval of call setup success rate rate of change.And the two-dimensional sequence that step 120 generates is mapped in the interval tabulation that generates above, obtains discretization two-dimensional sequence afterwards, the rate of change two-dimensional array of the call setup success rate after the discretization is:
(f
a1，f
a2，f
a3)，(f
a2，f
a3，f
a4)，(f
a3，f
a4，f
a5)，......，(f
a(i-1)，f
ai，f
a(i+1))，......
For example: an interval [3,100] is arranged now, is 5 subintervals (m=5) with this interval division, be respectively [3,20], [20,40], [40,60], [60,80], [80,100] (numbering in these five intervals is exactly f1～f5), exist a two-dimensional array (21,4,98) (23,56,73) (3,12,67), this ordered series of numbers is mapped in the top interval, and show with the numbering in interval, just become (f2, f1, f5) (f2, f3, f4) (f1, f1, f4), in fact this be exactly that top this section described illustrated particular problem.
Step 140 is carried out data mining and correlation analysis to the data of discretization, and obtaining with the performance data is based rule and knowledge;
Two-dimensional sequence after the discretization that generates with step 130 is as the input data, carry out excavations such as correlation rule, cluster, classification and correlation analysis [referring to Agrawal R., Imielinski T., and SwamiA. " Mining association rules between sets of items in large databases " .In Proc.ACM SIGMOD Conference Management of Data, pages 207-216, WashingtonD.C., USA, May 1993.], obtaining with the performance data is based rule and knowledge.
Step 150, rule that obtains for excavation and knowledge are carried out visual, so that obtain the understandable knowledge form of user, as b
I-1∧ b
i→ b
I+1Form is perhaps undertaken visually by graphics mode, generate the graphical representation form of knowledge, so that the user analyzes and uses.
Owing to consider that the initial data for performance statistics is often to change, so operating process is designed to top form, adding new performance index so if desired excavates, only need be with the performance data of newly adding input as above-mentioned steps, step above carrying out successively then can be finished the excacation of newly-increased data.
Because the diversity of performance statistic and data mode, therefore the generalization mode of operation that only is aimed at the connection attribute value that proposes above.Be directed to the performance initial data of special shape, can save wherein step 120 or step 130.And, as long as the performance initial data is provided, can finishes by the step of front and analyze and forecast function.
Claims (8)
1, a kind of method of accounting base-station performance data may further comprise the steps:
(a) the performance statistics proxy module that operates on base station and the switching equipment is regularly monitored and record the ruuning situation of system, and with the record performance data send to the performance management server, this performance management server will receive data and be kept in the database;
(b) performance data of the connection attribute value that collects is carried out asking for of rate of change, and the performance data rate of change that generates is processed into the data mode of the discretization that data mining can handle;
(c) data of discretization are carried out data mining and correlation analysis, obtaining with the performance data is based rule and knowledge;
(d) rule and the knowledge that obtains for excavation is carried out visualization processing.
2, the method for claim 1, it is characterized in that, when described step (b) is carried out asking for of rate of change, be to ask for the rate of change of the performance index of k+1 period, obtain the performance data rate of change of one-dimensional sequence form with respect to k period performance index.
3, method as claimed in claim 1 or 2, it is characterized in that, in the described step (b), be by the performance data rate of change that generates is carried out preliminary treatment, and/or the performance data rate of change carried out discretization, thereby the performance data rate of change that generates is processed into the data mode that data mining can be handled.
4, method as claimed in claim 3 is characterized in that, described step (b) is carried out preliminary treatment to the performance data rate of change that generates and comprised basic preliminary treatment and/or special pre-treatment; Wherein, basic preliminary treatment mainly is that the vacancy value and the noise data of performance data rate of change are handled, and special pre-treatment is that the data of performance data rate of change are carried out discretization.
5, method as claimed in claim 4 is characterized in that, described special pre-treatment is that the performance data rate of change with the one-dimensional sequence form is one group with continuous a plurality of periods, generates performance data variation rate two-dimensional sequence, so that carry out the discretization of data.
6, method as claimed in claim 5, it is characterized in that, when described step (b) is carried out discretization with the performance data rate of change, be to turn to a plurality of closed intervals with the span of described performance data rate of change two-dimensional sequence is discrete, a numbering is given in each interval, generates sequence of intervals f
1, f
2, f
3..., f
m, then described performance data rate of change two-dimensional sequence is mapped in the interval tabulation that generates above, obtain discretization two-dimensional sequence afterwards.
7, the method for claim 1 is characterized in that, described step (d) is excavated the rule obtain and knowledge and carried out when visual, comprises visual and graphic form visual of textual form.
8, the method for claim 1 is characterized in that, the knowledge form that described step (c) obtains is b
I-1∧ b
i→ b
I+1Rule, promptly be illustrated in b
I-1And b
iUnder the prerequisite of Chu Xianing, can derive b simultaneously
I+1Appearance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006100587069A CN101031125A (en) | 2006-03-02 | 2006-03-02 | Method for accounting base-station performance data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006100587069A CN101031125A (en) | 2006-03-02 | 2006-03-02 | Method for accounting base-station performance data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101031125A true CN101031125A (en) | 2007-09-05 |
Family
ID=38716135
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006100587069A Pending CN101031125A (en) | 2006-03-02 | 2006-03-02 | Method for accounting base-station performance data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101031125A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103987069A (en) * | 2008-07-25 | 2014-08-13 | 高通股份有限公司 | System and method for network management |
| CN104615660A (en) * | 2015-01-05 | 2015-05-13 | 浪潮(北京)电子信息产业有限公司 | Method and system for monitoring database performance |
| CN104881436A (en) * | 2015-05-04 | 2015-09-02 | 中国南方电网有限责任公司 | Power communication equipment performance analyzing method and device based on big data |
| WO2022152027A1 (en) * | 2021-01-14 | 2022-07-21 | 华为技术有限公司 | Management and control method for data analysis apparatus, and communication apparatus |
-
2006
- 2006-03-02 CN CNA2006100587069A patent/CN101031125A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103987069A (en) * | 2008-07-25 | 2014-08-13 | 高通股份有限公司 | System and method for network management |
| CN104615660A (en) * | 2015-01-05 | 2015-05-13 | 浪潮(北京)电子信息产业有限公司 | Method and system for monitoring database performance |
| CN104881436A (en) * | 2015-05-04 | 2015-09-02 | 中国南方电网有限责任公司 | Power communication equipment performance analyzing method and device based on big data |
| CN104881436B (en) * | 2015-05-04 | 2019-04-05 | 中国南方电网有限责任公司 | A kind of electric power communication device method for analyzing performance and device based on big data |
| WO2022152027A1 (en) * | 2021-01-14 | 2022-07-21 | 华为技术有限公司 | Management and control method for data analysis apparatus, and communication apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108417033B (en) | Expressway traffic accident analysis and prediction method based on multi-dimensional factors | |
| CN101888309B (en) | Online log analysis method | |
| CN103176974B (en) | The method and apparatus of access path in optimization data storehouse | |
| CN102081625B (en) | Data query method and query server | |
| CN101344893B (en) | History data access method and apparatus | |
| CN104809188A (en) | Enterprise talent drainage data mining analysis method and device | |
| CN111858713B (en) | Object-based government informationized asset management method and system | |
| CN111259073A (en) | Intelligent business system running state studying and judging system based on logs, flow and business access | |
| CN106202566A (en) | A kind of magnanimity electricity consumption data mixing based on big data storage system and method | |
| CN109242170A (en) | A kind of City Road Management System and method based on data mining technology | |
| CN112527886A (en) | Data warehouse system based on urban brain | |
| CN101031125A (en) | Method for accounting base-station performance data | |
| CN110532122B (en) | Fault analysis method and system, electronic equipment and storage medium | |
| CN111125450A (en) | Management method of multilayer topology network resource object | |
| CN114428813A (en) | Data statistics method, device, equipment and storage medium based on report platform | |
| CN101075303A (en) | Data unearch model for pedicting service potential customers | |
| CN112561392A (en) | Evaluation system and method for green space service level of city park | |
| CN114756602A (en) | Real-time streaming process mining method and system and computer readable storage medium | |
| CN114819468A (en) | Complex network-based construction period index analysis method | |
| Trabelsi et al. | User-Centred Application for Modeling Journeys in Digital Libraries | |
| Kumar et al. | Jaccard similarity based mining for high utility webpage sets from weblog database. | |
| Lei et al. | The application of multidimensional data analysis in the EIA database of electric industry | |
| CN118132609B (en) | Streaming data service quick response method, computer program product and electronic equipment | |
| CN113138875B (en) | Fault detection method, terminal and computer storage medium | |
| Mai et al. | Enhanced Mining of High Utility Patterns from Streams of Dynamic Profit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20070905 |