CN106506192A - A kind of method and apparatus of identification network key node - Google Patents
A kind of method and apparatus of identification network key node Download PDFInfo
- Publication number
- CN106506192A CN106506192A CN201610880461.1A CN201610880461A CN106506192A CN 106506192 A CN106506192 A CN 106506192A CN 201610880461 A CN201610880461 A CN 201610880461A CN 106506192 A CN106506192 A CN 106506192A
- Authority
- CN
- China
- Prior art keywords
- index
- network node
- measurement
- node
- measurement index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/091—Measuring contribution of individual network components to actual service level
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/142—Network analysis or design using statistical or mathematical methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Physics & Mathematics (AREA)
- Algebra (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Pure & Applied Mathematics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a kind of method and apparatus of identification network key node, the method includes:Multiple measurement indexs are selected from degree index, subgraph index, betweenness index, characteristic vector index, approximate characteristic vector index, nearness index, knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index;According to the multiple measurement indexs for selecting, comprehensive evaluation value is calculated to each network node in all-network node;One threshold value is set, if the comprehensive evaluation value of some network node is more than the threshold value, using the network node as key node;Or the ratio that key node number accounts for network node sum is set, all-network node is sorted from big to small according to comprehensive evaluation value, network node within the ratio will be come as key node.This programme is chosen multiple measurement indexs and is fused into single comprehensive evaluation value, as the foundation for weighing network node importance, improves the accuracy of identification network key node.
Description
Technical field
The present invention relates to communication technical field, and in particular to a kind of method and apparatus of identification network key node.
Background technology
From in terms of network structure feature, most battlefield cordless communication networks belong to Scale-free Network.Its typical structure
It is characterized in that:In network, most of node only has a few connection, and only a small amount of node has a large amount of connections with other nodes.
This network is very strong to the resistivity of random disturbances.But during in the face of deliberating the interference for key node, it may but can't bear
A blow.
The analysis of network key node is the basis that " war of attacking a vital point " is implemented, and is the premise of network selectivity interference.By interference
Network key node, can not only play the interference effect for getting twice the result with half the effort, and can give full play to the work of our network antagonism equipment
War efficiency.Therefore, become the important content of analysis of network to the analysis of battlefield cordless communication network Key of Implementation node.
Network key node refers to the node played an important role in network-in-dialing, for portraying network node in a network
Importance, portraying, in known measurement index, betweenness index is to portray battlefield channel radio for the centrad of conventional network node
The maximally effective index of letter net node center degree.Measurement index identification key node of single be usually used at present, but network section
The importance of point is affected by multiple measurement indexs simultaneously, only recognizes using certain single measurement index that key node is accurate
Really property is not enough.
Content of the invention
The invention provides a kind of method and apparatus of identification network key node, to solve single to estimate using certain
The index problem not enough to recognize key node accuracy.
According to one aspect of the present invention, the invention provides a kind of method of identification network key node, including:
From degree index, subgraph index, betweenness index, characteristic vector index, approximate characteristic vector index, nearness index,
Multiple measurement indexs are selected in knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index;
According to the multiple measurement indexs for selecting, overall merit is calculated to each network node in all-network node
Value;
One threshold value is set, if the comprehensive evaluation value of some network node is more than the threshold value, the network node is made
For key node;Or the ratio that key node number accounts for network node sum is set, by all-network node according to overall merit
Value is sorted from big to small, will come network node within the ratio as key node.
According to another aspect of the present invention, the invention provides a kind of device of identification network key node, including:
Measurement index select unit, for receiving from degree index, subgraph index, betweenness index, characteristic vector index, approximate
Characteristic vector index, nearness index, knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination refer to
The multiple measurement indexs selected in mark;
Comprehensive evaluation value computing unit, for the multiple measurement indexs received according to the measurement index select unit, right
Each network node in all-network node calculates comprehensive evaluation value;
Key node judging unit, for arranging a threshold value, if the comprehensive evaluation value of some network node is more than be somebody's turn to do
Threshold value, then using the network node as key node;Or the total ratio of network node is accounted for for arranging key node number, will
All-network node is sorted from big to small according to comprehensive evaluation value, and the network node come within the ratio is saved as crucial
Point.
The invention has the beneficial effects as follows:The embodiment of the present invention chooses multiple measurement indexs, to each network node, will choosing
The multiple measurement indexs for taking are fused into single comprehensive evaluation value, as weigh the network node importance foundation, compared to
Only recognize using a certain measurement index that network key point, this programme have higher accuracy.
Description of the drawings
Fig. 1 is a kind of flow chart of the method for identification network key node that one embodiment of the invention is provided;
Fig. 2 is a kind of functional block diagram of the device of identification network key node that one embodiment of the invention is provided;
Fig. 3 is OPNET scenes deployment scheme schematic diagram in one embodiment of the invention;
Fig. 4 is the parameter setting interface schematic diagram of the device for recognizing network key node in one embodiment of the invention;
Fig. 5 is the result figure for recognizing network key node in one embodiment of the invention using manual weighted comprehensive method;
Fig. 6 is the result for recognizing network key node in one embodiment of the invention using entropy weight gray relative analysis method
Figure.
Specific embodiment
The present invention design concept be:The importance of network node is affected by multiple measurement indexs simultaneously, is only used
Certain single measurement index is difficult to accurately identify key node.The present invention using analytic hierarchy process (AHP) select multiple measurement indexs come
The importance of network node is evaluated, due to various measurement indexs for network node is not no less important, therefore can be with
In conjunction with Fuzzy Comprehensive Evaluation, reasonably determine the weight of evaluation index, will obtain based on multinomial after multiple measurement index synthesis
The comprehensive single evaluation desired value of measurement index, by the ambiguity quantification of evaluation result, so as to identify network key node,
The accuracy of identification can be improved.
Embodiment one
Fig. 1 is a kind of flow chart of the method for identification network key node that one embodiment of the invention is provided, such as Fig. 1 institutes
Show, the method for the identification network key node that the present embodiment is provided includes:
Step S110:From degree index, subgraph index, betweenness index, characteristic vector index, approximate characteristic vector index, connect
Multiple surveys are selected in recency index, knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index
Degree index.
Step S120:According to the multiple measurement indexs for selecting, each network node in all-network node is calculated
Comprehensive evaluation value.
Step S130:One threshold value is set, if the comprehensive evaluation value of some network node is more than the threshold value, by the net
Network node is used as key node;Or arrange key node number account for network node sum ratio, by all-network node according to
Comprehensive evaluation value is sorted from big to small, will come network node within the ratio as key node.
The present embodiment chooses multiple measurement indexs, to each network node, the multiple measurement indexs that chooses is fused into
Single comprehensive evaluation value, as the foundation for weighing the network node importance, next compared to only using a certain measurement index
Identification network key point, the recognition methodss that the present embodiment is provided have higher accuracy.
Embodiment two
The present embodiment calculates the comprehensive evaluation value of each network node using manual weighted comprehensive method, i.e., according to defeated manually
The weighting parameter for entering is each measurement index weighting, then seeks their weighted sum as final comprehensive evaluation value.
Step i):For with n network node, the situation of m measurement index, index matrix is set up:
Wherein, xijJ-th measurement index for i-th network node.
Step ii):The matrix is normalized, normalization matrix is obtained:
Wherein,xmax(j) and xminJ () is respectively j-th of all-network node and estimates
The optimal value of index and worst-case value, j=1,2 ..., m.
Step iii):Weights are set for m measurement index, obtain weight vector W=(w1,w2,...,wm), according to formulaThe comprehensive evaluation value F of each network node can be calculatedi.
Embodiment three
Manual weighted comprehensive method due to adopting in embodiment two has certain subjectivity random, in some cases without
Method accurately reflects the characteristic distributions of each measurement index, in order to obtain automatically the weights shared by each index, the present embodiment
The comprehensive evaluation value that each network node is calculated using entropy weight weighted comprehensive method.
Entropy is the amount for representing material system mode, represents the degree that the state is likely to occur.Certain comentropy that estimates is got over
Little, illustrate that its degree of variation is bigger, there is provided quantity of information also bigger, then weight is also bigger;If on the contrary, certain measurement index
Comentropy bigger, there is provided quantity of information less, its weight is also correspondingly less.It can be considered to each measurement index
Degree of variation, determines the weight of each measurement index according to calculated entropy, then to each measurement index weighted,
Finally obtain more comprehensive more objective appraisal result.
Step i):For one with n network node, the problem of m measurement index sets up initial matrix:
Wherein, xijJ-th measurement index for i-th network node.
Step ii):Initial matrix is normalized, normalization matrix is obtained:
Wherein,xmax(j) and xminJ () is respectively j-th of all-network node and estimates
The optimal value of index and worst-case value, j=1,2,3 ..., m.
Step iii):According to formula:
Calculate the entropy S of j-th measurement indexj, wherein,
If Pij=1, then PijlnPij=0, thus the information disordering with entropy in theory of information disagree, therefore need right
PijDefinition correction be:
Step iv) according to formula:
Can determine the entropy weight w of j-th measurement indexj, so as to further obtain the entropy vector W=of m measurement index
(w1,w2,...,wm).
Step v) is according to formula:
The comprehensive evaluation value F of each network node can be calculatedi.
Example IV
The present embodiment calculates the comprehensive evaluation value of each network node, grey correlation analysis using gray relative analysis method
Method is expected that by special method according to the different or similarity degree of development trend between each factor, finds the number between them
Value relation, is a kind of effective ways for weighing correlation degree.The method quantified system development change situation, be very suitable for into
Mobile state course is analyzed, and what this was also changed with the complex network moment is actually consistent.
Step i):For one with n network node, the problem of m measurement index the following is i-th network node
Key measure vector:
Xi=(xi1,xi2,...,xim)
Wherein, xijFor j-th measurement index of i-th network node, 1≤i≤n.
Step ii):The measurement index of Integrated comparative all-network node, obtains the reference vector of measurement index:
Yi=(y1,y2,...,ym)
Wherein, ykIt is the optimal value of k-th measurement index of all-network node.
Step iii):As the dimension of network node each measurement index might not be identical, and the numerical value quantity having
Level differs greatly, and therefore will carry out nondimensionalization process to them, and the present embodiment adopts " averaging method ", to key measure vector
XiWith reference vector YiThe key measure vector after nondimensionalization is obtained after carrying out nondimensionalization process:
And the reference vector after nondimensionalization:
Wherein,Represent the meansigma methodss of k-th measurement index, 1≤k≤m.
Step iv):Set up matrix of differences:
Find out the maximum D in matrixmaxWith minima Dmin.
Step v):According to formula:
Calculate the coefficient of association r of each measurement index of each network nodeik, wherein, rikFor i-th network node kth
The coefficient of association of individual measurement index, ρ is resolution ratio, 0 < ρ < 1, it is preferable that ρ values 0.5.
Step vi):According to formula
Calculate the comprehensive evaluation value R of each network nodei, compare the comprehensive evaluation value R of each network nodeiSize,
So as to draw the key sequence of network node, comprehensive evaluation value RiBigger, node is more important.One threshold value can be set, comprehensive
Close evaluation of estimate RiThe network node for being more than the threshold value is exactly key node.
Embodiment five
The present embodiment is commented with reference to the synthesis that entropy weight weighted comprehensive method and gray relative analysis method calculate each network node
It is worth, further to improve accuracy.
Step i):For one with n network node, the problem of m measurement index the following is i-th network node
Key measure vector:
Xi=(xi1,xi2,...,xim)
Wherein, xijFor j-th measurement index of i-th network node, 1≤i≤n.
Step ii):The measurement index of Integrated comparative all-network node, obtains the reference vector of measurement index:
Yi=(y1,y2,...,ym)
Wherein, ykIt is the optimal value of k-th measurement index of all-network node.
Step iii):To key measure vector XiWith reference vector YiNondimensionalization is obtained after carrying out nondimensionalization process
Key measure vector afterwards:
And the reference vector after nondimensionalization:
Wherein,Represent the meansigma methodss of k-th measurement index, 1≤k≤m.
Step iv):Set up matrix of differences:
Find out the maximum D in matrixmaxWith minima Dmin.
Step v):According to formula:
Calculate the coefficient of association r of each measurement index of each network nodeik, wherein, rikFor i-th network node kth
The coefficient of association of individual measurement index, ρ is resolution ratio, 0 < ρ < 1, it is preferable that ρ values 0.5.
Step vi):Set up index matrix:And be normalized, obtain normalization
Matrix:Wherein,xmax(j) and xminJ () is respectively all-network
The optimal value of j-th measurement index of node and worst-case value.
Step vii):According toDetermine the entropy S of j-th measurement indexj, wherein,And according toDetermine the entropy weight w of j-th measurement indexj, obtain m measurement index
Entropy vector W=(w1,w2,...,wm).
Step viii):According to formula:
The comprehensive evaluation value R of each network node can be calculatedi.Again may be by comparing each network node
Comprehensive evaluation value RiSize, draw the key sequence of network node, comprehensive evaluation value RiBigger, node is more important.Or
A threshold value is may also set up, by comprehensive evaluation value RiIt is more than the network node of the threshold value as key node.
Embodiment six
Fig. 2 be one embodiment of the invention provide a kind of identification network key node device functional block diagram, such as Fig. 2
Shown, the device of the identification network key node that the present embodiment is provided includes:Measurement index select unit 210, comprehensive evaluation value
Computing unit 220, key node judging unit 230.
Measurement index select unit 210 is received from degree index, subgraph index, betweenness index, characteristic vector index, approximate spy
Levy to figureofmerit, nearness index, knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index
Multiple measurement indexs of middle selection.
Multiple measurement indexs that comprehensive evaluation value computing unit 220 is received according to measurement index select unit 210, to all
Each network node in network node calculates comprehensive evaluation value.
Key node judging unit 230 arranges a threshold value, if the comprehensive evaluation value of some network node is more than the threshold
Value, then using the network node as key node;Or key node judging unit 230 arranges key node number and accounts for network node
The ratio of sum, all-network node is sorted from big to small according to comprehensive evaluation value, the network section within the ratio will be come
Point is used as key node.
In a preferred embodiment, comprehensive evaluation value computing unit 220 specifically for:
For with n network node, the situation of m measurement index, index matrix is set up:
Wherein, xijJ-th measurement index for i-th network node.
The matrix is normalized, normalization matrix is obtained:
Wherein,xmax(j) and xminJ () is respectively j-th of all-network node and estimates
The optimal value of index and worst-case value, j=1,2 ..., m.
Weights are set for m measurement index, obtain weight vector W=(w1,w2,...,wm), according to formulaCalculate the comprehensive evaluation value F of each network nodei.
In another preferred embodiment, comprehensive evaluation value computing unit 220 specifically for:
For one with n network node, the problem of m measurement index sets up initial matrix:
Wherein, xijJ-th measurement index for i-th network node.
Initial matrix is normalized, normalization matrix is obtained:
Wherein,xmax(j) and xminJ () is respectively j-th of all-network node and estimates
The optimal value of index and worst-case value, j=1,2,3 ..., m.
According to formula:
Calculate the entropy S of j-th measurement indexj, wherein,
According to formula:
Can determine the entropy weight w of j-th measurement indexj, so as to further obtain the entropy vector W=of m measurement index
(w1,w2,...,wm).
According to formula:
Calculate the comprehensive evaluation value F of each network nodei.
In a further preferred embodiment, comprehensive evaluation value computing unit 220 specifically for:
For one with n network node, the problem of m measurement index the following is the key of i-th network node
Measure vector:
Xi=(xi1,xi2,...,xim)
Wherein, xijFor j-th measurement index of i-th network node, 1≤i≤n.
The measurement index of Integrated comparative all-network node, obtains the reference vector of measurement index:
Yi=(y1,y2,...,ym)
Wherein, ykIt is the optimal value of k-th measurement index of all-network node.
To key measure vector XiWith reference vector YiObtain after carrying out nondimensionalization process key after nondimensionalization
Measure vector:
And the reference vector after nondimensionalization:
Wherein,Represent the meansigma methodss of k-th measurement index, 1≤k≤m.
Set up matrix of differences:
Find out the maximum D in matrixmaxWith minima Dmin.
According to formula:
Calculate the coefficient of association r of each measurement index of each network nodeik, wherein, rikFor i-th network node kth
The coefficient of association of individual measurement index, ρ is resolution ratio, 0 < ρ < 1, it is preferable that ρ values 0.5.
According to formula
Calculate the comprehensive evaluation value R of each network nodei.
In being further preferable to carry out, comprehensive evaluation value computing unit 220 specifically for:
For one with n network node, the problem of m measurement index the following is the key of i-th network node
Measure vector:
Xi=(xi1,xi2,...,xim)
Wherein, xijFor j-th measurement index of i-th network node, 1≤i≤n.
The measurement index of Integrated comparative all-network node, obtains the reference vector of measurement index:
Yi=(y1,y2,...,ym)
Wherein, ykIt is the optimal value of k-th measurement index of all-network node.
To key measure vector XiWith reference vector YiObtain after carrying out nondimensionalization process key after nondimensionalization
Measure vector:
And the reference vector after nondimensionalization:
Wherein,Represent the meansigma methodss of k-th measurement index, 1≤k≤m.
Set up matrix of differences:
Find out the maximum D in matrixmaxWith minima Dmin.
According to formula:
Calculate the coefficient of association r of each measurement index of each network nodeik, wherein, rikFor i-th network node kth
The coefficient of association of individual measurement index, ρ is resolution ratio, 0 < ρ < 1, it is preferable that ρ values 0.5.
Set up index matrix:And be normalized, obtain normalization matrix:Wherein,xmax(j) and xminJ () is respectively all-network node
J-th measurement index optimal value and worst-case value.
According toDetermine the entropy S of j-th measurement indexj, wherein,And root
According toDetermine the entropy weight w of j-th measurement indexj, obtain the entropy vector W=(w of m measurement index1,
w2,...,wm).
According to formula:
Calculate the comprehensive evaluation value R of each network nodei.
Fig. 3 is OPNET scenes deployment scheme schematic diagram in one embodiment of the invention, as shown in figure 3, the present embodiment design
One wireless communication scene for having 34 nodes, scene are deployed in the range of 200km*100km, comprising three before
Post adds a NCS node, four sensor nodes, 20 common forward node, command center node, five attacks
Platform nodes.
Before network key node identification is carried out, need to carry out necessary setting, such as from the ten kinds of measurement indexs for being given
The middle measurement index for selecting multiple needs, selection calculate the method for comprehensive evaluation value, arrange threshold value or the ratio for judging key node
If the weights that example is also needed to arrange each measurement index using manual weighted comprehensive method, as shown in Figure 4.Fig. 5 is a reality of the invention
Apply the result figure that network key node is recognized in example using manual weighted comprehensive method, Fig. 6 is adopted in one embodiment of the invention
Entropy weight gray relative analysis method recognizes the result figure of network key node, contrasts Fig. 5 and Fig. 6, and the measurement index of selection is 5
Individual:Degree index, nearness index, knot removal lose index, betweenness index and accumulative nomination, and the ratio setting of key node is
10%, the weights arranged for each measurement index when adopting manual weighted comprehensive method are 0.2.As shown in figure 5, being added using manual
During power synthetic method identification network key node, the network key node for identifying is 14,10,24,26;As shown in fig. 6, adopting entropy
Power gray relative analysis method calculate automatically weighter factor identification network key point when, the network key node for identifying be 14,10,
9、24.Entropy weight gray relative analysis method is more accurate.
The above, specific embodiment only of the invention, under the above-mentioned teaching of the present invention, those skilled in the art
Other improvement or deformation can be carried out on the basis of above-described embodiment.It will be understood by those skilled in the art that above-mentioned tool
Body description simply preferably explains that the purpose of the present invention, protection scope of the present invention should be defined by scope of the claims.
Claims (10)
1. a kind of identification network key node method, it is characterised in that include:
From degree index, subgraph index, betweenness index, characteristic vector index, approximate characteristic vector index, nearness index, node
Delete in loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index and select multiple measurement indexs;
According to the multiple measurement indexs for selecting, comprehensive evaluation value is calculated to each network node in all-network node;
One threshold value is set, if the comprehensive evaluation value of some network node is more than the threshold value, using the network node as pass
Key node;Or arrange key node number account for network node sum ratio, by all-network node according to comprehensive evaluation value from
Little sequence is arrived greatly, network node within the ratio will be come as key node.
2. the method for claim 1, it is characterised in that described according to the multiple measurement indexs for selecting, to all-network
Each network node in node calculates comprehensive evaluation value, specifically includes:
If the sum of all-network node is n, the measurement index number of selection is m, then estimated according to m of n network node
Index Establishment index matrix
It is normalized, obtains normalization matrixWherein,
xmax(j) and xminJ () is respectively the optimal value of j-th measurement index of all-network node and worst-case value;
Weight vector W=(the w of m measurement index are set1,w2,...,wm)
According toCalculate the comprehensive evaluation value F of each network nodei.
3. the method for claim 1, it is characterised in that described according to the multiple measurement indexs for selecting, to all-network
Each network node in node calculates comprehensive evaluation value, specifically includes:
If the sum of all-network node is n, the measurement index number of selection is m, then estimated according to m of n network node
Index Establishment index matrix:
It is normalized, obtains normalization matrix:Wherein,
xmax(j) and xminJ () is respectively the optimal value of j-th measurement index of all-network node and worst-case value;
According toDetermine the entropy S of j-th measurement indexj, wherein,
According toDetermine the entropy weight w of j-th measurement indexj, obtain the entropy vector W=(w of m measurement index1,
w2,...,wm),
According toCalculate the comprehensive evaluation value F of each network nodei.
4. the method for claim 1, it is characterised in that described according to the multiple measurement indexs for selecting, to all-network
Each network node in node calculates comprehensive evaluation value, specifically includes:
If the sum of all-network node is n, the measurement index number of selection is m, then estimated according to m of n network node
Index, determines the key measure vector X of each network nodei=(xi1,xi2,...,xim);
The measurement index of Integrated comparative all-network node, obtains reference vector Y of measurement indexi=(y1,y2,...,ym), its
In, ykIt is the optimal value of k-th measurement index of all-network node;
To the key measure vector XiWith reference vector YiNondimensionalization process is carried out, the pass after nondimensionalization is obtained
Key measure vectorAnd the reference vector after nondimensionalization
Wherein,
Set up matrix of differences
According toCalculate the coefficient of association r of each measurement index of each network nodeik, wherein,
DmaxAnd DminFor the maximum in the matrix of differences Δ and minima, ρ is resolution ratio, 0 < ρ < 1;
According toCalculate the comprehensive evaluation value R of each network nodei.
5. the method for claim 1, it is characterised in that described according to the multiple measurement indexs for selecting, to all-network
Each network node in node calculates comprehensive evaluation value, specifically includes:
If the sum of all-network node is n, the measurement index number of selection is m, then estimated according to m of n network node
Index, determines the key measure vector X of each network nodei=(xi1,xi2,...,xim);
The measurement index of Integrated comparative all-network node, obtains reference vector Y of measurement indexi=(y1,y2,...,ym), its
In, ykIt is the optimal value of k-th measurement index of all-network node;
To the key measure vector XiWith reference vector YiNondimensionalization process is carried out, the key after nondimensionalization is obtained
Property measure vectorAnd the reference vector after nondimensionalization
Wherein,
Set up matrix of differences
According toCalculate the coefficient of association r of each measurement index of each network nodeik, wherein,
DmaxAnd DminFor the maximum in the matrix of differences Δ and minima, ρ is resolution ratio, 0 < ρ < 1;
Set up index matrix:And be normalized, obtain normalization matrix:Wherein,xmax(j) and xminJ () is respectively all-network node
J-th measurement index optimal value and worst-case value;
According toDetermine the entropy S of j-th measurement indexj, wherein,According toDetermine the entropy weight w of j-th measurement indexj, obtain the entropy vector W=(w of m measurement index1,w2,...,
wm);
According toCalculate the comprehensive evaluation value R of each network nodei.
6. a kind of identification network key node device, it is characterised in that include:
Measurement index select unit, for receiving from degree index, subgraph index, betweenness index, characteristic vector index, approximation characteristic
To in figureofmerit, nearness index, knot removal loss index, stream betweenness index, approximate stream betweenness index, accumulative nomination index
The multiple measurement indexs for selecting;
Comprehensive evaluation value computing unit, for the multiple measurement indexs received according to the measurement index select unit, to all
Each network node in network node calculates comprehensive evaluation value;
Key node judging unit, for arranging a threshold value, if the comprehensive evaluation value of some network node is more than the threshold value,
Then using the network node as key node;Or the total ratio of network node is accounted for for arranging key node number, will be all
Network node is sorted from big to small according to comprehensive evaluation value, will come network node within the ratio as key node.
7. device as claimed in claim 6, it is characterised in that the comprehensive evaluation value computing unit specifically for:
The m measurement index according to n network node sets up index matrix:And at normalization
Reason, obtains normalization matrixWherein,xmax(j) and xmin(j)
The respectively optimal value of j-th measurement index of all-network node and worst-case value;
Weight vector W=(the w of m measurement index are set1,w2,...,wm), according toCalculate each network section
The comprehensive evaluation value F of pointi.
8. device as claimed in claim 6, it is characterised in that the comprehensive evaluation value computing unit specifically for:
The m measurement index according to n network node sets up index matrixAnd be normalized
Process, obtain normalization matrix:Wherein,xmax(j) and xmin
J () is respectively the optimal value of j-th measurement index of all-network node and worst-case value;
According toDetermine the entropy S of j-th measurement indexj, wherein,
According toDetermine the entropy weight w of j-th measurement indexj, obtain the entropy vector W=of m measurement index
(w1,w2,...,wm), according toCalculate the comprehensive evaluation value F of each network nodei.
9. device as claimed in claim 6, it is characterised in that the comprehensive evaluation value computing unit specifically for:
According to m measurement index of n network node, the key measure vector X of each network node is determinedi=(xi1,
xi2,...,xim);
The measurement index of Integrated comparative all-network node, obtains reference vector Y of measurement indexi=(y1,y2,...,ym), its
In, ykIt is the optimal value of k-th measurement index of all-network node;
To the key measure vector XiWith reference vector YiImmeasurable tempering process is carried out, is obtained key after nondimensionalization
Measure vectorAnd the reference vector obtained after nondimensionalization
Wherein,
Set up matrix of differences
According toCalculate the coefficient of association r of each measurement index of each network nodeik, wherein,
DmaxAnd DminFor the maximum in the matrix of differences Δ and minima, ρ is resolution ratio, 0 < ρ < 1;
According toCalculate the comprehensive evaluation value R of each network nodei.
10. device as claimed in claim 6, it is characterised in that the comprehensive evaluation value computing unit specifically for:
According to m measurement index of n network node, the key measure vector X of each network node is determinedi=(xi1,
xi2,...,xim);
The measurement index of Integrated comparative all-network node, obtains reference vector Y of measurement indexi=(y1,y2,...,ym), its
In, ykIt is the optimal value of k-th measurement index of all-network node;
To the key measure vector XiWith reference vector YiImmeasurable tempering process is carried out, the key after nondimensionalization is obtained
Property measure vectorAnd the reference vector obtained after nondimensionalization
Wherein,
Set up matrix of differences
According toCalculate the coefficient of association r of each measurement index of each network nodeik, wherein,
DmaxAnd DminFor the maximum in the matrix of differences Δ and minima, ρ is resolution ratio, 0 < ρ < 1;
Set up index matrix:And be normalized, obtain normalization matrix:Wherein,xmax(j) and xminJ () is respectively all-network node
J-th measurement index optimal value and worst-case value;
According toDetermine the entropy S of j-th measurement indexj, wherein,According toDetermine the entropy weight w of j-th measurement indexj, obtain the entropy vector W=(w of m measurement index1,w2,...,
wm);
According toCalculate the comprehensive evaluation value R of each network nodei.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610880461.1A CN106506192A (en) | 2016-10-09 | 2016-10-09 | A kind of method and apparatus of identification network key node |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610880461.1A CN106506192A (en) | 2016-10-09 | 2016-10-09 | A kind of method and apparatus of identification network key node |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106506192A true CN106506192A (en) | 2017-03-15 |
Family
ID=58293757
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610880461.1A Pending CN106506192A (en) | 2016-10-09 | 2016-10-09 | A kind of method and apparatus of identification network key node |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106506192A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107154870A (en) * | 2017-04-28 | 2017-09-12 | 毕晨晓 | A kind of flow monitoring method based on power automatic system |
CN108009710A (en) * | 2017-11-19 | 2018-05-08 | 国家计算机网络与信息安全管理中心 | Node test importance appraisal procedure based on similarity and TrustRank algorithms |
CN110474806A (en) * | 2019-08-19 | 2019-11-19 | 国网甘肃省电力公司经济技术研究院 | A kind of power telecom network key node recognition methods |
CN111371624A (en) * | 2020-03-17 | 2020-07-03 | 电子科技大学 | Tactical communication network key node identification method based on environment feedback |
CN113342523A (en) * | 2021-06-04 | 2021-09-03 | 中国人民解放军军事科学院评估论证研究中心 | Method, device, equipment and medium for analyzing equilibrium of combat system structure |
CN115277438A (en) * | 2022-08-11 | 2022-11-01 | 国网河南省电力公司信息通信公司 | Power communication network node importance evaluation method based on multi-factor evaluation index |
CN116132446A (en) * | 2022-12-30 | 2023-05-16 | 南京工程学院 | Network key node identification method and application thereof |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103200118A (en) * | 2013-03-13 | 2013-07-10 | 南京理工大学 | Complex network protective method oriented towards adding of protective links of key nodes |
CN103281336A (en) * | 2013-06-19 | 2013-09-04 | 上海众恒信息产业股份有限公司 | Network intrusion detection method |
CN105069290A (en) * | 2015-08-03 | 2015-11-18 | 同济大学 | Parallelization critical node discovery method for postal delivery data |
CN105301602A (en) * | 2015-11-06 | 2016-02-03 | 北京空间飞行器总体设计部 | Navigation satellite integrity key point comprehensive identification method based on grey relational degree |
-
2016
- 2016-10-09 CN CN201610880461.1A patent/CN106506192A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103200118A (en) * | 2013-03-13 | 2013-07-10 | 南京理工大学 | Complex network protective method oriented towards adding of protective links of key nodes |
CN103281336A (en) * | 2013-06-19 | 2013-09-04 | 上海众恒信息产业股份有限公司 | Network intrusion detection method |
CN105069290A (en) * | 2015-08-03 | 2015-11-18 | 同济大学 | Parallelization critical node discovery method for postal delivery data |
CN105301602A (en) * | 2015-11-06 | 2016-02-03 | 北京空间飞行器总体设计部 | Navigation satellite integrity key point comprehensive identification method based on grey relational degree |
Non-Patent Citations (1)
Title |
---|
《复杂网络关键节点识别技术研究》: "《复杂网络关键节点识别技术研究》", 31 January 2012 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107154870A (en) * | 2017-04-28 | 2017-09-12 | 毕晨晓 | A kind of flow monitoring method based on power automatic system |
CN107154870B (en) * | 2017-04-28 | 2020-07-24 | 毕晨晓 | Flow monitoring method based on electric power automation system |
CN108009710A (en) * | 2017-11-19 | 2018-05-08 | 国家计算机网络与信息安全管理中心 | Node test importance appraisal procedure based on similarity and TrustRank algorithms |
CN110474806A (en) * | 2019-08-19 | 2019-11-19 | 国网甘肃省电力公司经济技术研究院 | A kind of power telecom network key node recognition methods |
CN110474806B (en) * | 2019-08-19 | 2022-04-01 | 国网甘肃省电力公司经济技术研究院 | Power communication network key node identification method |
CN111371624A (en) * | 2020-03-17 | 2020-07-03 | 电子科技大学 | Tactical communication network key node identification method based on environment feedback |
CN111371624B (en) * | 2020-03-17 | 2022-10-14 | 电子科技大学 | Tactical communication network key node identification method based on environment feedback |
CN113342523A (en) * | 2021-06-04 | 2021-09-03 | 中国人民解放军军事科学院评估论证研究中心 | Method, device, equipment and medium for analyzing equilibrium of combat system structure |
CN113342523B (en) * | 2021-06-04 | 2023-10-03 | 中国人民解放军军事科学院评估论证研究中心 | Battle architecture balance analysis method, device, equipment and medium |
CN115277438A (en) * | 2022-08-11 | 2022-11-01 | 国网河南省电力公司信息通信公司 | Power communication network node importance evaluation method based on multi-factor evaluation index |
CN115277438B (en) * | 2022-08-11 | 2023-07-07 | 国网河南省电力公司信息通信公司 | Power communication network node importance evaluation method based on multi-factor evaluation index |
CN116132446A (en) * | 2022-12-30 | 2023-05-16 | 南京工程学院 | Network key node identification method and application thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106506192A (en) | A kind of method and apparatus of identification network key node | |
CN103874200B (en) | A kind of floor recognition methods and system | |
CN109882019B (en) | Automobile electric tail door opening method based on target detection and motion recognition | |
CN106469299A (en) | A kind of vehicle search method and device | |
CN105426908B (en) | A kind of substation's attributive classification method based on convolutional neural networks | |
CN104992177A (en) | Internet porn image detection method based on deep convolution nerve network | |
CN108932535B (en) | A kind of edge calculations clone's node recognition methods based on machine learning | |
CN104077560B (en) | Fingerprint comparison method | |
CN105488528A (en) | Improved adaptive genetic algorithm based neural network image classification method | |
CN103888541B (en) | Method and system for discovering cells fused with topology potential and spectral clustering | |
CN106557740B (en) | The recognition methods of oil depot target in a kind of remote sensing images | |
CN110033332A (en) | A kind of face identification method, system and electronic equipment and storage medium | |
CN108875907B (en) | Fingerprint identification method and device based on deep learning | |
CN102722577B (en) | The determination method and device of index changeable weight | |
CN108256436A (en) | A kind of radar HRRP target identification methods based on joint classification | |
CN107506350A (en) | A kind of method and apparatus of identification information | |
CN103310235B (en) | A kind of steganalysis method based on parameter identification and estimation | |
CN104581748B (en) | A kind of method and apparatus identifying scene within a wireless communication network | |
CN106572387B (en) | Video sequence alignment schemes and system | |
CN103678483A (en) | Video semantic analysis method based on self-adaption probability hypergraph and semi-supervised learning | |
CN106372656B (en) | Obtain method, image-recognizing method and the device of the disposable learning model of depth | |
CN107145778A (en) | A kind of intrusion detection method and device | |
CN107360577A (en) | A kind of frequency spectrum sensing method and device based on machine learning | |
CN107451605A (en) | A kind of simple target recognition methods based on channel condition information and SVMs | |
CN114998958A (en) | Face recognition method based on lightweight convolutional neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170315 |
|
WD01 | Invention patent application deemed withdrawn after publication |