CN104468196B - Virtual network method for diagnosing faults and device based on evidence screening - Google Patents
Virtual network method for diagnosing faults and device based on evidence screening Download PDFInfo
- Publication number
- CN104468196B CN104468196B CN201410641874.5A CN201410641874A CN104468196B CN 104468196 B CN104468196 B CN 104468196B CN 201410641874 A CN201410641874 A CN 201410641874A CN 104468196 B CN104468196 B CN 104468196B
- Authority
- CN
- China
- Prior art keywords
- evidence
- virtual network
- matrix
- sub
- network components
- 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
Abstract
The present invention relates to Network Fault Diagnosis Technique field, and in particular to a kind of virtual network method for diagnosing faults and device based on evidence screening.A kind of virtual network method for diagnosing faults and device based on evidence screening provided by the invention, evidence matrix model is established by using the observation result to virtual network, the probability of malfunction of each virtual network components is solved using DS evidence theories, so that it is determined that faulty components, the dynamic, autgmentability and information uncertainty of virtual network are overcome.Meanwhile because the Screening Treatment that the technical solution adopted in the present invention is shifted to an earlier date to evidence so that fault location both maintains high accuracy, greatly improves time efficiency again so that overall efficiency maximizes.
Description
Technical field
The present invention relates to Network Fault Diagnosis Technique field, and in particular to a kind of virtual network failure based on evidence screening
Diagnostic method and device.
Background technology
In virtual network environment, multiple virtual networks are present on same bottom physical network simultaneously, traditional interconnection
Net service provider (Internet Service Provider, ISP) is divided into two parts：Infrastructure provider
(Infrastructure Providers, InPs) and network service operators (Service Providers, SPs), basis is set
Apply provider to be used for providing and managing physical infrastructure, network service operators utilize the resource that multiple InPs are provided, passed through
Abstract, distribution and isolation mech isolation test deployment virtual network, the end-to-end service of innovation and diversified business are provided for terminal user
Using.
Because the transparency of the bottom-up information for upper-level virtual network causes fault detection system can not in virtualized environment
Complete network knowledge is obtained, it is substantial amounts of uncertain so as to exist in virtual network fault diagnosis；In addition, virtual network
It is typical large-scale distributed network, wherein comprising substantial amounts of dummy node and virtual link, these components move with demand again
State changes, the network topology do not fixed.In addition influence of noise so that become in virtual environment to the fault diagnosis of virtual network
It is more difficult.
Existing technical scheme mainly using based on management level actively or passively detection failure localization method come to virtual net
Network carries out fault diagnosis.However, virtual network failure is diagnosed it should be understood that the Global Topological of network using the above method, it is impossible to compared with
The dynamic and autgmentability of virtual network are adapted to well.
The content of the invention
The defects of dynamic and autgmentability for can not preferably adapt to virtual network in the prior art, the present invention provide
A kind of virtual network method for diagnosing faults and device based on evidence screening.
On the one hand, a kind of virtual network method for diagnosing faults based on evidence screening provided by the invention, including：
Obtain the observation result whether each client breaks down to virtual network path corresponding to the client；
Evidence matrix is established, wherein the corresponding client of every a line of the evidence matrix, the of the evidence matrix
One row to should client observation result, remaining each corresponding virtual network components of row, the virtual network components bag
Include dummy node and virtual link；
The evidence matrix is split as more sub- evidence matrixes, columns and the card of each sub- evidence matrix
It is equal according to matrix column number；
For sub- evidence matrix each described, solved to obtain the hair of each virtual network components according to DS evidence theories
The probability of raw failure；
Choose the virtual network group for the maximum probability that breaks down successively according to the descending order of the probability to break down
Part, it is until the quantity for the virtual network path to break down that whole virtual network components of selection are covered reaches preset value
Only.
Further, the described the step of evidence matrix is split as more sub- evidence matrixes, including：
The evidence matrix is split as two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence
Matrix, the even number line of the evidence matrix is as the second sub- evidence matrix.
Further, it is described to be solved to obtain the probability to break down of each virtual network components according to DS evidence theories
The step of, including：
For sub- evidence matrix each described, a m of each virtual network components is constructed according to DS evidence theories
Function；
For each virtual network components, according to the fusion rule of DS evidence theories by same virtual network components
All m functions are merged, and obtain the probability that the virtual network components break down.
Further, it is described each virtual network components is constructed according to DS evidence theories a m function the step of, bag
Include：
For i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N represents normal, A generations
Table failure；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiTherefore
The posterior probability of barrier.
Further, it is described to be merged all m functional values of same virtual network components according to DS evidence theories
The step of, including：
For
Wherein, X, B, C is burnt first, m1M functions corresponding to first sub- evidence matrix, m2For m corresponding to the second sub- evidence matrix
Function, K are normaliztion constant：
Further, the preset value is calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down；
Wherein antinoise coefficient is parameter preset.
Corresponding, the present invention also provides a kind of virtual network trouble-shooter based on evidence screening, including：
Acquisition module, for obtaining whether each client breaks down to virtual network path corresponding to the client
Observation result；
Module is established, for establishing evidence matrix, wherein the corresponding client of every a line of the evidence matrix, described
The first row of evidence matrix to should client observation result, remaining each corresponding virtual network components of row, the void
Intending networking component includes dummy node and virtual link；
Module is split, for the evidence matrix to be split as into more sub- evidence matrixes, each described sub- evidence matrix
Columns it is equal with the columns of the evidence matrix；
Module is solved, for for sub- evidence matrix each described, solving to obtain each void according to DS evidence theories
Intend the probability to break down of networking component；
Module is chosen, for choosing the maximum probability that breaks down successively according to the descending order of the probability to break down
Virtual network components, until the quantity of the virtual network path to break down that whole virtual network components of selection are covered
Untill reaching preset value.
Further, the fractionation module is specifically used for：
The evidence matrix is split as two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence
Matrix, the even number line of the evidence matrix is as the second sub- evidence matrix.
Further, the solution module is specifically used for：
For i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N represents normal, A generations
Table failure；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiTherefore
The posterior probability of barrier；
For
Wherein, X, B, C is burnt first, m1M functions corresponding to first sub- evidence matrix, m2For m corresponding to the second sub- evidence matrix
Function, K are normaliztion constant：
Further, the selection module is specifically used for：
The preset value is calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down；Wherein antinoise coefficient
For parameter preset.
A kind of virtual network method for diagnosing faults and device based on evidence screening provided by the invention, by using to void
The observation result for intending network establishes evidence matrix model, and the failure that each virtual network components are solved using DS evidence theories is general
Rate, so that it is determined that faulty components, overcome the dynamic, autgmentability and information uncertainty of virtual network.Meanwhile because this
The Screening Treatment that technical scheme is shifted to an earlier date to evidence used by invention so that fault location both maintains high accuracy,
Time efficiency is greatly improved again so that overall efficiency maximizes.
Brief description of the drawings
The features and advantages of the present invention can be more clearly understood by reference to accompanying drawing, accompanying drawing is schematically without that should manage
Solve to carry out any restrictions to the present invention, in the accompanying drawings：
Fig. 1 is that a kind of flow of the virtual network method for diagnosing faults based on evidence screening is shown in one embodiment of the invention
It is intended to；
Fig. 2 be in one embodiment of the invention using the time-consuming ratio of the present embodiment method and prior art fault detect institute with
The change schematic diagram of component count；
Fig. 3 be in one embodiment of the invention using the time-consuming ratio of the present embodiment method and prior art fault detect institute with
The change schematic diagram of evidence quantity；
Fig. 4 is to be carried out when fault rate is 0.5% in one embodiment of the invention using the present embodiment method and prior art
The accuracy rate comparison schematic diagram of fault detect；
Fig. 5 is to be carried out when fault rate is 0.6% in one embodiment of the invention using the present embodiment method and prior art
The accuracy rate comparison schematic diagram of fault detect；
Fig. 6 is to be carried out when fault rate is 0.8% in one embodiment of the invention using the present embodiment method and prior art
The accuracy rate comparison schematic diagram of fault detect；
Fig. 7 is to be carried out when fault rate is 1.0% in one embodiment of the invention using the present embodiment method and prior art
The accuracy rate comparison schematic diagram of fault detect；
Fig. 8 is to be carried out when fault rate is 0.5% in one embodiment of the invention using the present embodiment method and prior art
The misdiagnosis rate comparison schematic diagram of fault detect；
Fig. 9 is to be carried out when fault rate is 0.6% in one embodiment of the invention using the present embodiment method and prior art
The misdiagnosis rate comparison schematic diagram of fault detect；
Figure 10 is to be carried out when fault rate is 0.8% in one embodiment of the invention using the present embodiment method and prior art
The misdiagnosis rate comparison schematic diagram of fault detect；
Figure 11 is to be carried out when fault rate is 1.0% in one embodiment of the invention using the present embodiment method and prior art
The misdiagnosis rate comparison schematic diagram of fault detect；
Figure 12 is a kind of structure of the virtual network trouble-shooter based on evidence screening in one embodiment of the invention
Schematic diagram.
Embodiment
Technical solution of the present invention is further elaborated in conjunction with drawings and examples.
Fig. 1 shows a kind of flow signal of the virtual network method for diagnosing faults based on evidence screening in the present embodiment
Figure, as shown in figure 1, a kind of virtual network method for diagnosing faults based on evidence screening that the present embodiment provides, including：
S1, obtain the observation knot whether each client breaks down to virtual network path corresponding to the client
Fruit.
S2, evidence matrix is established, wherein the corresponding client of every a line of the evidence matrix, the evidence matrix
First row to should client observation result, remaining each corresponding virtual network components of row, the virtual network components
Including dummy node and virtual link.
Specifically, the evidence matrix includes：
The first of the evidence matrix is classified as the row that report an error, if the observation result of the client is failure, the client
The first of corresponding row is classified as 1, if the observation result of the client is normal, what the client corresponded to row first is classified as 0；
Since secondary series, each described virtual network components corresponds to a row of the evidence matrix, if the client
Virtual network path corresponding to end includes the virtual network components, then the client corresponds to the virtual network components of row correspondingly
Be classified as 1, remaining is classified as 0.
S3, the evidence matrix is split as more sub- evidence matrixes, the columns of each sub- evidence matrix and institute
The columns for stating evidence matrix is equal.
S4, for sub- evidence matrix each described, solved to obtain each virtual network components according to DS evidence theories
The probability to break down.
S5, choose the virtual network for the maximum probability that breaks down successively according to the descending order of the probability to break down
Component, until the quantity for the virtual network path to break down that whole virtual network components of selection are covered reaches preset value
Untill.
In order to avoid influence of the noise to testing result, the accurate of testing result is improved by introducing antinoise coefficient
Degree, i.e., described preset value are calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down.
Further, in order to obtain the m functions of more accurate virtual network components, odd even is used in the present embodiment
Row is split, and the evidence matrix is split as into two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence
Matrix, the even number line of the evidence matrix is as the second sub- evidence matrix.
Further, it is described to be solved to obtain the probability to break down of each virtual network components according to DS evidence theories
The step of, including：
For sub- evidence matrix each described, a m of each virtual network components is constructed according to DS evidence theories
Function；
For each virtual network components, according to the fusion rule of DS evidence theories by same virtual network components
All m functions are merged, and obtain the probability that the virtual network components break down.
Wherein, it is described each virtual network components is constructed according to DS evidence theories a m function the step of, including：
For i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N represents normal, A generations
Table failure；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiTherefore
The posterior probability of barrier.
For the posterior probability QiSolved using recursive function loop.For any virtual network components Ci.Note
Relatei is CiAssociated component set, i.e., comprising component CiThe path broken down in other assemblies set.If
The collection is combined into sky, then Qi=FaultRate, i.e. virtual network indigenous fault rate；If set is not sky, recursive function is utilized
Loop is proceeded as follows：
(1) all evidences for including Relatei [count] are deleted to atom evidence matrix M, obtains new matrix M1；
(2) Relatei [count] is removed from atom evidence matrix, obtains new matrix M2
1 is added to counter count, then recursive function loop returns to Prc*loop (count, M1)+(1-Prc) * loop
(count, M2), wherein, Prc is Relatei [count] prior probability.
Further, it is described to be merged all m functional values of same virtual network components according to DS evidence theories
The step of, including：
For
Wherein, X, B, C is burnt first, m1M functions corresponding to first sub- evidence matrix, m2For m corresponding to the second sub- evidence matrix
Function, K are normaliztion constant：Wherein antinoise coefficient is parameter preset.
In this example, the present embodiment detection method effect is illustrated using virtual scene experiment.Given birth to using INET instruments
Into the network topology of different nodes.Nodes choose therein 10% to 20% and are used as virtual component from 1000 to 20000,
Fault rate in virtual network is by 0.5% to 1.5%, and the evidence number that client is observed is from 1000 to 20000, wherein observing
As a result it is that the evidence to break down is referred to as negative evidence to be referred to as affirmative evidence and observation result for normal evidence, because evidence is
Generate at random, so most of evidence is affirmative evidence.Noise jamming rate in the system is 0.01%, and 70% route is jumped
Number is 1,20% for 2,7% for 3,3% for 4.
It is illustrated in figure 2 as package count increases under different faults rate, fault detect is carried out using the present embodiment method
The time-consuming change with carrying out the time-consuming ratio of fault detect using prior art, be illustrated in figure 3 under different faults rate with
Evidence number increases, and the time-consuming of fault detect is carried out with carrying out the time-consuming of fault detect using prior art using the present embodiment method
Ratio change.
Fig. 4 to Fig. 7 is respectively illustrated in the case of different faults rate, and fault detect is carried out with adopting using the present embodiment method
The accuracy rate that fault detect is carried out with prior art contrasts situation.Fig. 8 to Figure 11 is respectively illustrated in the case of different faults rate,
Fault detect is carried out using the present embodiment method and contrasts situation with carrying out the misdiagnosis rate of fault detect using prior art.
In summary, the time consumed using the present embodiment detection method, it is compared to what is consumed using prior art
Time averagely reduces 20~30%, but after being decoupled to evidence, accuracy rate and misdiagnosis rate almost do not have compared with prior art
Change.Therefore a kind of virtual network method for diagnosing faults based on evidence screening that the present embodiment provides largely improves
Time efficiency maintain high-accuracy.
A kind of virtual network method for diagnosing faults based on evidence screening that the present embodiment provides, by using to virtual net
The observation result of network establishes evidence matrix model, and the probability of malfunction of each virtual network components is solved using DS evidence theories, from
And faulty components are determined, overcome the dynamic, autgmentability and information uncertainty of virtual network.It is meanwhile because of the invention
The Screening Treatment that used technical scheme is shifted to an earlier date to evidence so that fault location both maintains high accuracy, and pole
Big improves time efficiency so that overall efficiency maximizes.
On the other hand, as shown in figure 12, the present embodiment additionally provides a kind of virtual network failure based on evidence screening and examined
Disconnected device, including：
Acquisition module 101, for obtaining whether virtual network path corresponding to the client occurs each client
The observation result of failure；
Module 102 is established, for establishing evidence matrix, wherein the corresponding client of every a line of the evidence matrix,
The first row of the evidence matrix to should client observation result, remaining each corresponding virtual network components of row, institute
Stating virtual network components includes dummy node and virtual link；
Module 103 is split, for the evidence matrix to be split as into more sub- evidence matrixes, each described sub- evidence square
The columns of battle array is equal with the columns of the evidence matrix；
Module 104 is solved, for for sub- evidence matrix each described, solving to obtain each according to DS evidence theories
The probability to break down of virtual network components；
Module 105 is chosen, for choosing the probability that breaks down successively according to the descending order of the probability to break down
Maximum virtual network components, until the virtual network path to break down that whole virtual network components of selection are covered
Untill quantity reaches preset value.
Further, the fractionation module 103 is specifically used for：
The evidence matrix is split as two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence
Matrix, the even number line of the evidence matrix is as the second sub- evidence matrix.
Further, the solution module 104 is specifically used for：
For sub- evidence matrix each described, a m of each virtual network components is constructed according to DS evidence theories
Function；
For each virtual network components, according to the fusion rule of DS evidence theories by same virtual network components
All m functions are merged, and obtain the probability that the virtual network components break down.
Specifically, it is directed to i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N is represented just
Often, A representing faults；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiTherefore
The posterior probability of barrier.
For
Wherein, X, B, C is burnt first, m1M functions corresponding to first sub- evidence matrix, m2For m corresponding to the second sub- evidence matrix
Function, K are normaliztion constant：
Further, the selection module 105 is specifically used for：
The preset value is calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down；Wherein antinoise coefficient
For parameter preset.
A kind of virtual network trouble-shooter based on evidence screening provided by the invention, by using to virtual network
Observation result establish evidence matrix model, the probability of malfunction of each virtual network components is solved using DS evidence theories, so as to
Faulty components are determined, overcome the dynamic, autgmentability and information uncertainty of virtual network.Meanwhile because institute of the present invention
The Screening Treatment that the technical scheme of use is shifted to an earlier date to evidence so that fault location both maintains high accuracy, and greatly
Improve time efficiency so that overall efficiency maximize.
Although being described in conjunction with the accompanying embodiments of the present invention, those skilled in the art can not depart from this hair
Various modifications and variations are made in the case of bright spirit and scope, such modifications and variations are each fallen within by appended claims
Within limited range.
Claims (10)
1. a kind of virtual network method for diagnosing faults based on evidence screening, it is characterised in that methods described includes：
Obtain the observation result whether each client breaks down to virtual network path corresponding to the client；
Evidence matrix is established, wherein the corresponding client of every a line of the evidence matrix, the first row of the evidence matrix
To should client the observation result, remaining each corresponding virtual network components of row, the virtual network components bag
Include dummy node and virtual link；
The evidence matrix is split as more sub- evidence matrixes, the columns of each sub- evidence matrix and the evidence square
The columns of battle array is equal；
For sub- evidence matrix each described, solved to obtain the generation event of each virtual network components according to DS evidence theories
The probability of barrier；
Choose the virtual network components for the maximum probability that breaks down successively according to the descending order of the probability to break down, directly
Untill the quantity for the virtual network path to break down that whole virtual network components of selection are covered reaches preset value.
2. according to the method for claim 1, it is characterised in that described that the evidence matrix is split as more sub- evidence squares
The step of battle array, including：
The evidence matrix is split as two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence square
Battle array, the even number line of the evidence matrix is as the second sub- evidence matrix.
3. method according to claim 1 or 2, it is characterised in that described to solve to obtain each according to DS evidence theories
The step of probability to break down of virtual network components, including：
For sub- evidence matrix each described, a m function of each virtual network components is constructed according to DS evidence theories；
For each virtual network components, according to the fusion rule of DS evidence theories owning same virtual network components
M functional values are merged, and obtain the probability that the virtual network components break down.
4. according to the method for claim 3, it is characterised in that described that each virtual network is constructed according to DS evidence theories
The step of one m function of component, including：
For i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N is represented normally, and A represents event
Barrier；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiAfter failure
Test probability.
5. according to the method for claim 4, it is characterised in that it is described according to DS evidence theories by same virtual network group
The step of all m functional values of part are merged, including：
For
Wherein, X, B, C is burnt first, m1For m functions, m corresponding to the first sub- evidence matrix2For m letters corresponding to the second sub- evidence matrix
Number, m1,2For m1Functional value and m2The m functions that functional value obtains after being merged, m1,2(X) it is burnt first X m1,2Functional value, m1(B)
For burnt first B m1Functional value, m2(C) it is burnt first C m2Functional value,Represent straight and computing, K are normaliztion constant：
6. according to the method for claim 1, it is characterised in that the preset value is calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down；
Wherein antinoise coefficient is parameter preset.
7. a kind of virtual network trouble-shooter based on evidence screening, it is characterised in that described device includes：
Acquisition module, the sight whether broken down to virtual network path corresponding to the client for obtaining each client
Examine result；
Module is established, for establishing evidence matrix, wherein the corresponding client of every a line of the evidence matrix, the evidence
The first row of matrix to should client the observation result, remaining each corresponding virtual network components of row, the void
Intending networking component includes dummy node and virtual link；
Module is split, for the evidence matrix to be split as into more sub- evidence matrixes, the row of each sub- evidence matrix
Number is equal with the columns of the evidence matrix；
Module is solved, for for sub- evidence matrix each described, solving to obtain each virtual net according to DS evidence theories
The probability to break down of network component；
Module is chosen, for choosing the void for the maximum probability that breaks down successively according to the descending order of the probability to break down
Intend networking component, until the quantity for the virtual network path to break down that whole virtual network components of selection are covered reaches
Untill preset value.
8. device according to claim 7, it is characterised in that the fractionation module is specifically used for：
The evidence matrix is split as two sub- evidence matrixes, the odd-numbered line of the evidence matrix is as the first sub- evidence square
Battle array, the even number line of the evidence matrix is as the second sub- evidence matrix.
9. device according to claim 8, it is characterised in that the solution module is specifically used for：
For i-th of virtual network components Ci, establish CiIdentification framework Θ={ Ni, Ai, wherein N is represented normally, and A represents event
Barrier；
Work as Qi>PiWhen, m (Ni)=min (1, log (Qi\Pi)), m ({ Ni, Ai)=1-m (Ni)；m(Ai)=0；
Work as Qi<=PiWhen, m (Ai)=min (1 ,-log (Qi\Pi))；m({Ni, Ai)=1-m (Ni), m (Ni)=0；
The QiFor the virtual network components CiNormal posterior probability, the PiFor the virtual network components CiAfter failure
Test probability；
For
Wherein, X, B, C is burnt first, m1For m functions, m corresponding to the first sub- evidence matrix2For m letters corresponding to the second sub- evidence matrix
Number, m1,2For m1Functional value and m2The m functions that functional value obtains after being merged, m1,2(X) it is burnt first X m1,2Functional value, m1(B)
For burnt first B m1Functional value, m2(C) it is burnt first C m2Functional value,Represent straight and computing, K are normaliztion constant：
10. device according to claim 7, it is characterised in that the selection module is specifically used for：
The preset value is calculated using below equation：
Preset value=all virtual network path quantity * (1- antinoises coefficient) to break down；
Wherein antinoise coefficient is parameter preset.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410641874.5A CN104468196B (en) | 2014-11-13 | 2014-11-13 | Virtual network method for diagnosing faults and device based on evidence screening |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410641874.5A CN104468196B (en) | 2014-11-13 | 2014-11-13 | Virtual network method for diagnosing faults and device based on evidence screening |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104468196A CN104468196A (en) | 2015-03-25 |
| CN104468196B true CN104468196B (en) | 2017-12-05 |
Family
ID=52913607
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410641874.5A Active CN104468196B (en) | 2014-11-13 | 2014-11-13 | Virtual network method for diagnosing faults and device based on evidence screening |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104468196B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105490836B (en) * | 2015-11-19 | 2019-04-19 | 湘南学院 | A kind of Monte Carlo appraisal procedure of complex network reliability |
| CN108520266B (en) * | 2018-03-01 | 2019-10-22 | 西北工业大学 | A kind of Time Domain Fusion method for diagnosing faults based on DS evidence theory |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101350739A (en) * | 2008-09-05 | 2009-01-21 | 北京邮电大学 | Method for locating fault in IP network |
| CN103763132A (en) * | 2014-01-02 | 2014-04-30 | 北京邮电大学 | Network virtualization environment fault diagnosis method based on relevance of symptoms and faults |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8868344B2 (en) * | 2011-09-22 | 2014-10-21 | Honeywell International Inc. | Systems and methods for combining a priori data with sensor data |
-
2014
- 2014-11-13 CN CN201410641874.5A patent/CN104468196B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101350739A (en) * | 2008-09-05 | 2009-01-21 | 北京邮电大学 | Method for locating fault in IP network |
| CN103763132A (en) * | 2014-01-02 | 2014-04-30 | 北京邮电大学 | Network virtualization environment fault diagnosis method based on relevance of symptoms and faults |
Non-Patent Citations (4)
| Title |
|---|
| Evidence Relationship Matrix and Its Application to D-S Evidence Theory for Information Fusion;Xianfeng Fan, Hong-Zhong Huang, and Qiang Miao;《IDEAL》;20060923;全文 * |
| Network Security Situation Evaluation Based on Modified D-S Evidence Theory;WANG Chundong, ZHANG Yu;《Wuhan University Journal of Natural Sciences》;20141031;第19卷(第5期);全文 * |
| 一种改进的D-S证据理论合成方法;杨靖，林益，洪露，李泽滔;《计算机工程与应用》;20120425;第48卷(第20期);全文 * |
| 基于D-S 证据理论的网络异常检测方法;诸葛建伟，王大为，陈 昱，叶志远，邹维;《软件学报》;20060331;第17卷(第3期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104468196A (en) | 2015-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106874187B (en) | Code coverage rate collection method and device | |
| CN105786678B (en) | A kind of testability prediction method based on correlation models | |
| CN107690778A (en) | Use the performance of mirror image detection grouping mea-sure network | |
| CN105629156B (en) | Analog circuit fault based on genetic planning tests optimal sequential search method | |
| CN104597892A (en) | Hierarchical fault diagnosis method for electronic information equipment | |
| CN104796240B (en) | A kind of fuzz testing system of stateful procotol | |
| CN105512011B (en) | A kind of electronics testability modeling appraisal procedure | |
| CN103713989A (en) | Test case generating method and test case generating device for user terminal | |
| CN105357076B (en) | Network connectivty detection method and device between a kind of node | |
| CN103064008A (en) | Nolinear analog circuit soft fault diagnostic method based on Hilbert-huang transform | |
| CN105701097A (en) | Social-network-platform-based public opinion analysis method and system | |
| CN104506229B (en) | A kind of satellite network failure diagnostic method based on sub-clustering | |
| CN106411629A (en) | Method used for monitoring state of CDN node and equipment thereof | |
| CN103763132B (en) | Network virtualization environment fault diagnosis method based on relevance of symptoms and faults | |
| CN103957132B (en) | Reflecting internal storage network node card shared storage zone data mapping error rate testing method | |
| CN107966648B (en) | A kind of embedded failure diagnosis method based on correlation matrix | |
| CN104468196B (en) | Virtual network method for diagnosing faults and device based on evidence screening | |
| CN106845137A (en) | A kind of sacred disease analysis method based on brain network structure | |
| CN103995901B (en) | A kind of method for determining back end failure | |
| CN106021642A (en) | Reachable identification based process model repairing method | |
| CN103684890B (en) | Server stress method of testing and system | |
| CN105228185A (en) | A kind of method for Fuzzy Redundancy node identities in identification communication network | |
| Zhang et al. | Service failure diagnosis in service function chain | |
| CN104486337B (en) | Validation of Data method and device | |
| CN110096753A (en) | The Bayes appraisal procedure of lower pair of exponential distribution of two kinds of stochastic variables |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |