CN102025733B - Health degree evaluation method based on cognitive network - Google Patents
Health degree evaluation method based on cognitive network Download PDFInfo
- Publication number
- CN102025733B CN102025733B CN 201010576217 CN201010576217A CN102025733B CN 102025733 B CN102025733 B CN 102025733B CN 201010576217 CN201010576217 CN 201010576217 CN 201010576217 A CN201010576217 A CN 201010576217A CN 102025733 B CN102025733 B CN 102025733B
- Authority
- CN
- China
- Prior art keywords
- network
- business
- router
- evaluation
- health degree
- 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
Images
Abstract
The invention provides a health degree evaluation method based on a cognitive network, belonging to the technical field of network health assessment. In the method, by establishing a comprehensive fuzzy evaluation system which comprises a service evaluation module, a router evaluation module and a link evaluation module, the health degree value of a network is comprehensively obtained through the service evaluation module, the router evaluation module and the link evaluation module, and then the result is input into a strategy database for adjustment and optimization of the network. From the service perspective, the QoS (quality of service) performance of the cognitive network is comprehensively evaluated by combining a network element with a link so as to adjust the QoS of the network according to feedback, thus improving the self-adaptivity, self-management performance and end-to-end QoS performance of the cognitive network.
Description
Technical field:
The present invention relates to a kind of health degree evaluation method based on cognition network, belong to network health assessment technique field.
Background technology:
Present internet is a passive network that depends on to a great extent manual intervention: data exchange having between the fringe node of network intelligence.Backbone network also is indifferent to the particular content of the data that will transmit.In case transfer of data breaks down, fringe node can be distinguished and go wrong, but what backbone network can not decision problem be, more need not carry and having dealt with problems.In this case, the network manager need to intervene network, determines failure cause.In most of the cases, the self-configuring solution of describing network failure with high-level language (as natural language) is impossible to the keeper, must recover network function by backbone network is carried out the equipment concrete configuration.The scholar of Virginia, US engineering college in 2005 clearly proposes the cognition network definition first: cognition network is to have cognitive process, can perception current network condition, and the network of then making planning, decision-making and taking to move according to these conditions.Cognition network must be from perception: should be able to know what has occured in inside, and what must be done; Must determine that suitably action goes to reach target and study is done all these.It should be with the cognitive style self-configuration, self-optimization, self-regeneration and conduct protection.Therefore need to need a complete appraisement system to cognition network, help himself improve and optimize.
At present the integrated evaluating method of network performance mainly contained 3 kinds: 1) real network system operation conditions is observed, carried out analysis-by-synthesis and evaluation by the various parameters of collecting with corresponding relative index; 2) to the computer program description of real network system, move by program the simulation that the result that obtains is come the phase-split network performance; 3) real network system is set up Mathematical Modeling, simulate and the evaluating network performance by a series of mathematical formulae.But these evaluation methods to network performance are more unilateral, mainly for certain a bit, or certain one side, can not be thorough deeply reflect the network real conditions, also be unsuitable for using cognition network.
In network performance evaluation, mainly use the many indexs such as delay, packet loss, bandwidth and throughput to carry out performance evaluation to measured network.Though but this index system method can reflect the state of development of some things comprehensively, run into again difficulty when comparing between different things.Because use in the time of each index, the situation of comparison often can occur can't unify between different indexs, thereby can not make overall contrast on time and space to being evaluated object.Thereby cognition network is estimated more accurately.
Summary of the invention
Goal of the invention:
Technical problem to be solved of the present invention is the defective for the above-mentioned background technology, and a kind of health degree evaluation method based on cognition network is provided.From operational angle, the QoS performance of comprehensive network element and link thoroughly evaluating cognition network makes network service quality adjust by feedback, thereby improves the adaptivity of cognition network, from managerial and End-to-end QoS performance.
Technical scheme:
The present invention adopts following technical scheme for achieving the above object:
The present invention proposes a kind of health degree evaluation method based on cognition network, steps of the method are:
Step 1) set up the Comprehensive Fuzzy Evaluation system, described Fuzzy Evaluation System comprises business evaluation module, router evaluation module, link evaluating module;
Step 2) gather active user's qos parameter, described qos parameter comprises packet loss, shake, time delay, and the qos parameter that collects is stored in database;
Step 3) obtain the SLA of corresponding service from database, SLA value and step 2 with the business of needs evaluations) the qos parameter value that collects incoming traffic evaluation module together, the business evaluation module carries out normalized parameter to input numerical value to be processed, and calculates the health degree value of current business or the health degree value of the business of some periods;
Step 4) according to step 3) the health degree value of the business that draws calculates the irrelevance of itself and standard value, and need to judge whether evaluation or change the QoS that network state is improved business by irrelevance;
Step 5) not up to standard when the health degree of the business of discovery, adopt the router evaluation module that router is estimated, evaluation procedure is:
A, choose 5 key parameters of router, be respectively time delay, shake, packet loss, throughput and buffer memory;
B, with method for normalizing, above-mentioned parameter to be carried out unit unified;
C, the corresponding business of basis increase weighted value to the These parameters parameter, compare the health degree value of router;
Whether the health degree value of D, the router that draws according to C step judges this router health, adjusts and optimizes;
Step 6) when the router evaluation module is estimated, adopt the parameter of link evaluating module collection to make evaluation to existing link circuit condition, estimate the health status of link by carrying degree and stability, and draw irrelevance;
Step 7) draw the health degree value of network by business evaluation module, router evaluation module, link evaluating module synthesis, and with input policing storehouse as a result, network is adjusted and optimized.
Further, the step 6 of above-mentioned health degree evaluation method based on cognition network) described in, the carrying degree draws by the ratio of calculating link utilization, and described stability draws by the changing value of packet loss and time delay.
QoS (Quality of Service) is service quality, is a kind of security mechanism of network, is with a kind of technology that solves the problems such as network delay and obstruction.Under normal circumstances, if network only is used for specific timeless application system, do not need QoS, such as Web uses, or E-mail arranges etc.But just very necessary to key application and multimedia application.When network over loading or when congested, QoS can guarantee that the important service amount is not postponed or abandons, and guarantees simultaneously the efficient operation of network.
The abbreviation of SLA:Service-Level Agreement, the meaning are service-level agreements, are for ensureing the Performance And Reliability of service, the agreement of a kind of mutual concession that defines between service provider and user under certain expense.
Beneficial effect:
1, the characteristics (weight of each qos parameter) according to business self have drawn the health degree value.
2, take business as basic evaluation the health degree value of router and link, in order in time adjust network strategy.
3, by this complete appraisement system, improve cognition network from managerial and from controlled, more guaranteed the QoS of different user without business.
Description of drawings:
Fig. 1 is the flow chart of health degree appraisement system.
Fig. 2 is business health degree flow chart.
Fig. 3 is network element health degree flow chart.
Fig. 4 is link health degree flow chart.
Specific embodiments:
Be described in further detail below in conjunction with the enforcement of accompanying drawing to technical scheme:
Cognition network is a kind of network schemer of rising in recent years, and he is different from legacy network, and its key character is self-perception, self-decision-making, and the oneself controls.The QoS that the present invention is conceived to cognition network uses, service-oriented.Because the enforcement of QoS of survice relates to each management domain and the apparatus for network node that business is flowed through, therefore can pass through cognition module, behavior model will obtain the QoS data of each network element device, realize the flexible active management to the different isomerization network QoS, and the data of obtaining are estimated.The present invention combines network element centered by business, link and performance are end to end chosen representational parameter, set up the overall evaluation system of cognition network health degree, and reaction network provides the ability of service for Business Stream, and overall performance of network.
as shown in Figure 1, business module is by the monitoring to business, collect time delay, shake, the parameters such as packet loss, calculate health degree value and degree of balance value, and result of calculation is stored in database, database root carries out some anticipations according to historical results, infer that the problem of network may be because what causes this moment, if the reason of router, just start router-module, calculate the health degree value of route, if instead be the just health degree value of the company's of calculating trousers of link module, say that its result of calculation is stored in database, and send into simultaneously policy library, network can find corresponding strategy to adjust network according to result of calculation.
1, set up overall evaluation system
A: network element set
B: collection of services
C: end-to-end link set
H: health degree
All index system Ω, Ω=A ∩ B ∩ C,
Efficiency index system complete or collected works G in numerous system indexs, is not that each index is useful for different business, and we choose effective index for concrete condition so, remove those insignificant Index elements
Represent that various performance index collection represent respectively time delay, shake, packet loss, CPU usage, bandwidth availability ratio;
A={A
1, A
2, A
3... A
n, wherein 1,2,3...n refers to n router in certain network domains;
B={B
1, B
2, B
3Video traffic, speech business, data service have been represented respectively.These three kinds of business are more representational business.
C=(C
1, C
2, C3...), represent the bar of the N end to end link in certain network domains;
H=represents health degree, and the value of H is more large represents that more the health degree of business is better.
2, three concrete module evaluations
1) business evaluation
The traffic affecting parameter has a lot, and the present invention has chosen according to voice, data, these three kinds of sorting techniques of video the index that time delay, shake, packet loss are estimated as business.By to the actual value of these three dynamic indicators and normal value relatively, comprehensively draw not health degree in the same time.
As shown in Figure 2, carry out comparing with the SLA value for each parameter of certain business, because comparative result may exceed positive integer, at first process making data in tolerance interval by parameter, avoid occurring negative value and zero to its normalization.All qos parameter weighted of secondly comprehensive this business draw the health degree value of business this moment.Add integration according to actual conditions again, draw the health degree value of period.The business of considering is taking resource, therefore in the situation that to satisfy the not all parameter of SLA threshold values more high better, need the equilibrium index of computing service parameters, exponential quantity is less, illustrate that the degree of balance between every parameter that has satisfied the health degree value is less, waste resource degree is larger.
2) network element evaluation
Select local area network (LAN) a: B={B1, B2, B3...Bn}, 1,2,3...n refer to n router in this local area network (LAN).Each router has again evaluation index separately, we to 5 indexs of n router
Time delay,
Shake,
Packet loss,
Throughput,
Buffer memory,
Do the matrix of a n*5;
As shown in Figure 3, each row represents in network that (1＜n), every delegation represents the value of same router different parameters for the value Bn of a different router identical parameters.Bjj in matrix can be the value in some moment, also can according to the frequency of measuring time of heartbeat mechanism for example, get the mean value of a minor time slice.(1) formula is that (2) formula is the normalization that aligns property parameters to the normalization of negative property parameters.After parameter normalization, parameter weighting is heavily weighed.
Wherein
The maximum of certain index of expression router,
Represent corresponding minimum value, Aij refers to the target nominal value.
Therefore evaluation method of the present invention is centered by business, for different business, the various parameters of route is increased weights, thereby compares the health degree of certain business of running on which router.Make H (Ak) be router health nominal value at this moment,
Be the tolerance interval of this router index parameter, can set up on their own according to diverse network state and requirement.By to the determining of the deviate of each parameter of each router, add the weight of each straggling parameter, draw the router health degree of this moment.Draw the health degree value by each the service feature parameter weighting to router, for the discontented toe mark of single parameter, carry out irrelevance and calculate, adjust the concrete situation of router according to the non-health degree that calculates.
3) link evaluating
As shown in Figure 4, after the health degree of having described router and business, a comprehensive comprehensively evaluation end to end be arranged to network, just need to also make assessment to the health degree of link; It is the bearing capacity of link that the present invention is divided into two first parts of part to link, and second portion is the stability of link.
The bearing capacity α of link, α (t)=1 represent that t moment link is busy, and α (t)=0 represents t link idle constantly, and α (x) is the x instantaneous utilance of link constantly, and in the time [t, t+ τ], the ratio that link is in busy condition is link utilization.
The stability of link definition β, what stability should be by packet loss and time delay overall merit.
A parameter can not illustrate the stability of link, and we need to know the situation of a time period [t, t+ τ], and in to [t, t+ τ], we say the value that at every turn measures
Be defined as
At the Measuring Time internal variance, the minimal time delay of every link is all different, increases minimal time delay with the characteristic of reflection link.Get again the stability of outgoing link by the changing value to packet loss and time delay, these two parameters have certain restriction relation each other, directly addition, therefore we are to it with add constraint, and guarantee that its value falls between [0,1], under ideal state, when link did not have the variation of time delay there is no packet loss yet, the value of β was 1.
The stability of link and link bearer degree are united to get the health degree index of outgoing link.
Claims (2)
1. the health degree evaluation method based on cognition network, is characterized in that: comprise the steps:
Step 1) is set up the Comprehensive Fuzzy Evaluation system, and described Fuzzy Evaluation System comprises business evaluation module, router evaluation module, link evaluating module;
Step 2) gather active user's qos parameter, described qos parameter comprises packet loss, shake, time delay, and the qos parameter that collects is stored in database;
Step 3) is obtained the service-level agreement SLA of corresponding service from database, service-level agreement SLA value and step 2 with the business of needs evaluations) the qos parameter value that collects incoming traffic evaluation module together, the business evaluation module carries out normalized parameter to input numerical value to be processed, and calculates the health degree value of current business or the health degree value of the business of some periods;
The health degree value of the business that step 4) draws according to step 3) is calculated the irrelevance of itself and standard value, need to judge whether evaluation or changes the QoS that network state is improved business by irrelevance;
Step 5) is not up to standard when the health degree of discovery business, adopts the router evaluation module that router is estimated, and evaluation procedure is:
A, choose 5 key parameters of router, be respectively time delay, shake, packet loss, throughput and buffer memory;
B, with method for normalizing, above-mentioned parameter to be carried out unit unified;
C, the corresponding business of basis increase weighted value to the These parameters parameter, compare the health degree value of router;
Whether the health degree value of D, the router that draws according to C step judges this router health, adjusts and optimizes;
Step 6) is when the router evaluation module is estimated, and the parameter that adopts the link evaluating module to collect is made evaluation to existing link circuit condition, estimates the health status of link by carrying degree and stability, and draws irrelevance;
Step 7) draws the health degree value of network by business evaluation module, router evaluation module, link evaluating module synthesis, and with input policing storehouse as a result, network is adjusted and optimized.
2. according to claim 1 based on the health degree evaluation method of cognition network, it is characterized in that: the degree of carrying described in step 6) draws by the ratio of calculating link utilization, and described stability draws by the changing value of packet loss and time delay.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010576217 CN102025733B (en) | 2010-12-07 | 2010-12-07 | Health degree evaluation method based on cognitive network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010576217 CN102025733B (en) | 2010-12-07 | 2010-12-07 | Health degree evaluation method based on cognitive network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102025733A CN102025733A (en) | 2011-04-20 |
| CN102025733B true CN102025733B (en) | 2013-05-08 |
Family
ID=43866587
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010576217 Active CN102025733B (en) | 2010-12-07 | 2010-12-07 | Health degree evaluation method based on cognitive network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102025733B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102724193B (en) * | 2012-06-14 | 2015-01-21 | 西安交通大学 | Control method aiming at streaming service survivability in IP (Internet protocol) network environment |
| CN103117888B (en) * | 2013-01-17 | 2017-02-08 | 深信服网络科技（深圳）有限公司 | Method and device and system for carrying out application performance evaluation through network event |
| CN103152218B (en) * | 2013-01-30 | 2016-12-28 | 北京奇虎科技有限公司 | Computer network state-detection and restorative procedure and device |
| CN104580090B (en) * | 2013-10-18 | 2018-03-13 | 华为技术有限公司 | The method and device that security strategy O＆M is assessed |
| CN103561085B (en) * | 2013-10-30 | 2016-08-31 | 南京邮电大学 | A kind of service cloud evaluation method based on service level agreement constraint |
| CN103763123A (en) * | 2013-12-26 | 2014-04-30 | 华为技术有限公司 | Method and device for evaluating health condition of network |
| CN103728974B (en) * | 2014-01-20 | 2016-11-23 | 北京航空航天大学 | The dynamic network scheduling evaluated based on QoS and control method, system and device |
| CN104468200B (en) * | 2014-11-25 | 2018-06-15 | 中国人民解放军国防科学技术大学 | The Adaptive critic method of data center network apparatus health degree |
| CN104579843B (en) * | 2015-01-14 | 2018-09-28 | 浪潮天元通信信息系统有限公司 | A kind of network element health degree analysis method and device based on listed mechanism |
| CN110401551B (en) * | 2018-04-24 | 2022-05-13 | 中国移动通信集团广东有限公司 | Internet health degree evaluation method and system based on S1 interface |
| CN108768710B (en) * | 2018-05-18 | 2021-12-24 | 国家电网公司信息通信分公司 | Dynamic weight evaluation method, model and device for optical transmission network health |
| CN111277542A (en) * | 2018-12-04 | 2020-06-12 | 国网电动汽车服务有限公司 | Method and device for determining information security state of electric vehicle charging network |
| CN113541983B (en) * | 2020-04-14 | 2023-04-25 | 中国移动通信集团浙江有限公司 | Method, device, computing equipment and computer storage medium for optimizing network quality |
| CN115701057A (en) * | 2021-07-26 | 2023-02-07 | 华为技术有限公司 | Path determination method, device, equipment, system and computer readable storage medium |
| CN114095345A (en) * | 2021-10-22 | 2022-02-25 | 深信服科技股份有限公司 | Method, device, equipment and storage medium for evaluating health condition of host network |
| CN116633434B (en) * | 2023-07-24 | 2023-09-19 | 北京翌特视讯科技有限公司 | Transmission monitoring method and system of multifunctional integrated service optical transceiver |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001093037A2 (en) * | 2000-06-01 | 2001-12-06 | Aerocast.Com, Inc. | Client side holistic health check |
| CN1588881A (en) * | 2004-07-01 | 2005-03-02 | 北京邮电大学 | Method and device for controlling close ring feedback in IP network service quality management system |
| CN101534523A (en) * | 2009-04-08 | 2009-09-16 | 西安电子科技大学 | Cognitive network route method with service sensing ability |
-
2010
- 2010-12-07 CN CN 201010576217 patent/CN102025733B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001093037A2 (en) * | 2000-06-01 | 2001-12-06 | Aerocast.Com, Inc. | Client side holistic health check |
| CN1588881A (en) * | 2004-07-01 | 2005-03-02 | 北京邮电大学 | Method and device for controlling close ring feedback in IP network service quality management system |
| CN101534523A (en) * | 2009-04-08 | 2009-09-16 | 西安电子科技大学 | Cognitive network route method with service sensing ability |
Non-Patent Citations (2)
| Title |
|---|
| 基于认知层的认知网络结构及其认知方法;邵飞等;《北京工业大学学报》;20090930;第35卷(第9期);1181-1187 * |
| 邵飞等.基于认知层的认知网络结构及其认知方法.《北京工业大学学报》.2009,第35卷(第9期),1181-1187. |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102025733A (en) | 2011-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102025733B (en) | Health degree evaluation method based on cognitive network | |
| US10938664B2 (en) | Detecting network entity groups with abnormal time evolving behavior | |
| US11616682B2 (en) | Threshold selection for KPI candidacy in root cause analysis of network issues | |
| US8773992B2 (en) | Methods and apparatus for hierarchical routing in communication networks | |
| US8073945B2 (en) | Method and apparatus for providing a measurement of performance for a network | |
| US20190028909A1 (en) | Adaptive health status scoring for network assurance | |
| US20180365581A1 (en) | Resource-aware call quality evaluation and prediction | |
| US20190372827A1 (en) | Anomaly severity scoring in a network assurance service | |
| US10735274B2 (en) | Predicting and forecasting roaming issues in a wireless network | |
| CN106302012A (en) | A kind of PTN network simulation-optimization method and system | |
| Aimtongkham et al. | An enhanced CoAP scheme using fuzzy logic with adaptive timeout for IoT congestion control | |
| Umoh et al. | Fuzzy logic-based quality of service evaluation for multimedia transmission over wireless ad hoc networks | |
| Rezaul et al. | An overview of long-range dependent network traffic engineering and analysis: characteristics, simulation, modelling and control | |
| Na et al. | A novel adaptive traffic prediction AQM algorithm | |
| Farhoudi et al. | Server load balancing in software-defined networks | |
| Ganesh et al. | Congestion notification and probing mechanisms for endpoint admission control | |
| Jung et al. | A prediction method of network traffic using time series models | |
| CN111327494B (en) | Multi-domain SDN network flow situation assessment method and system | |
| WO2017062795A1 (en) | Automatic customer bandwidth utilization analysis for promoting dynamic capacity | |
| Nie et al. | A reconstructing approach to end‐to‐end network traffic based on multifractal wavelet model | |
| Tang et al. | Network availability evaluation based on markov chain of qos-aware | |
| CN109446020A (en) | Dynamic evaluation method and device of cloud storage system | |
| Meshinchi | QoS-Aware and Status-Aware Adaptive Resource Allocation Framework in SDN-Based IoT Middleware | |
| Arzo et al. | Intelligent qos agent design for qos monitoring and provisioning in 6g network | |
| CN101820389B (en) | Network path situation assessment method based on intelligent computation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20110420 Assignee: Nanjing Axon Technology Co., Ltd. Assignor: Nanjing Post & Telecommunication Univ. Contract record no.: 2017320000034 Denomination of invention: Health degree evaluation method based on cognitive network Granted publication date: 20130508 License type: Exclusive License Record date: 20170306 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EC01 | Cancellation of recordation of patent licensing contract |
Assignee: NANJING AXON SCIENCE & TECHNOLOGY Co.,Ltd. Assignor: NANJING University OF POSTS AND TELECOMMUNICATIONS Contract record no.: 2017320000034 Date of cancellation: 20210604 |
|
| EC01 | Cancellation of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20110420 Assignee: Jiangsu Tuoyou Information Intelligent Technology Research Institute Co.,Ltd. Assignor: NANJING University OF POSTS AND TELECOMMUNICATIONS Contract record no.: X2021320000043 Denomination of invention: Health evaluation method based on cognitive network Granted publication date: 20130508 License type: Common License Record date: 20210616 |
|
| EE01 | Entry into force of recordation of patent licensing contract |