CN105490834B - A kind of probe deployment method based on vertex covering and weak vertex cover - Google Patents
A kind of probe deployment method based on vertex covering and weak vertex cover Download PDFInfo
- Publication number
- CN105490834B CN105490834B CN201510807537.3A CN201510807537A CN105490834B CN 105490834 B CN105490834 B CN 105490834B CN 201510807537 A CN201510807537 A CN 201510807537A CN 105490834 B CN105490834 B CN 105490834B
- Authority
- CN
- China
- Prior art keywords
- node
- vertex
- probe
- network
- state
- 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.)
- Expired - Fee Related
Links
Classifications
-
- 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
- H04L41/12—Discovery or management of network topologies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/12—Network monitoring probes
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to a kind of probe deployment method based on vertex covering and weak vertex cover, including：Determine the deployed position of the quantity and the test probe of test probe in a network.The present invention proposes a kind of probe deployment technology based on vertex covering and weak vertex cover, it the advantage is that, the technology is improved based on the greedy algorithm to minimal covering problem on the basis of the algorithm, it is proposed that the probe deployment algorithm based on vertex covering and weak vertex cover.Minimum Vertex Covering is used first so that under conditions of it can obtain each link flow, required flow monitor number is minimum.It is improved to minimum weak vertex cover later.Emulation shows that compared with the deployment scheme based on Minimum Vertex Covering the number of probes that this programme uses is less, and algorithm is simpler, time-consuming shorter, is a kind of more excellent Network Traffic Monitoring probe deployment scheme.
Description
Technical field
The invention belongs to the network flow detection techniques based on probe, are a kind of network layers that can be realized to data network
The probe deployment method of complete detection.
Background technology
For the business demand of Network Traffic Monitoring, the detection mode of NetFlow can be used.A certain node in a network
Setting probe can obtain the flow of all chain roads being connected with this node.Therefore, all links in network in order to obtain
Network flow, generally can by certain switching nodes (router) be configured monitor realize, we will consider the problems of
It is：Monitor is set on network which node, could be caused under conditions of it can obtain each link flow, required stream
It is minimum to measure monitor number.
By probe, the comprehensive monitoring to business to the network layer of data network is realized.It is not only of high cost if all standing,
And big amount measurement data can be also generated when measuring, cause noise.In the network of n network element composition, between k network element of measurement
Traffic characteristic, need to choose k network element from n, furthermore, it is contemplated that flowing to problem, also need the arrangement group of k node of consideration
It closes, measurement scheme just needs C (n, k) P (k, k).It is if complicated according to the time caused by the measurement method and scheme of this all standing
It spends for factorial grade, space complexity is also very huge.Needs around business monitoring set out, and how research disposes suitable spy
Needle is to meet different business monitoring requirements, to solve the problems, such as deployed position and deployment capacity.
To understand the state of development of the prior art, retrieval is carried out to existing paper and patent, has compared and analyzes, has been screened
Go out the higher technical information of following degree related to the present invention：
Technical solution 1：Patent No. CN103138833A's《Network code pairing relationship in PON based on flow monitoring
Management method》Patent is related to field of communication technology, is mainly completed by three steps：First, OLT set up that there are bi-directional data biographies
The network code connection relation of defeated ONU pairs, after OLT detects the interaction of this ONU pairs bidirectional traffic, you can start network
Cataloged procedure；Second, OLT use certain calculation based on all ONU pairs there are bidirectional data transfers of ONU identifiers statistics
Bi-directional data flow size is converted into corresponding impact factor by method；Third, OLT monitor the situation of change of impact factor in real time,
If when the impact factor IFn for having neither part nor lot in ONU pairs of network code is more than the impact factor IFp for participating in ONU pairs of network code,
The corresponding network codes pair of IFp are removed, while are connected for the corresponding ONU of IFn to establishing network code.
Technical solution 2：Patent No. CN102781014A's《Intelligent flow based on wireless sensor network and 3G network
Monitoring method》Patent, be related to the data on flows of sensor acquisition and the nowhere seamless link technology of sensor network information,
802.15.4 protocol conversion technology between agreement and ICP/IP protocol.Data collecting module collected different fluid medium, caliber etc.
The parameter of dynamic change, the parameter value of acquisition are sent to the ZigBee channel radios of wireless sensor network by Asynchronous Transfer Mode
Believe module, then be transmitted to processing and operation that microprocessor module carries out parameter, host computer prison is transmitted to finally by 3G gateways
Remote flow monitoring, management are realized in control center.
Technical solution 3：Patent No. CN103036741A's《The determining method and device of flow monitoring baseline》Patent relates to
And a kind of determining method and device of flow monitoring baseline, the historical traffic number in predetermined period is determined in rectangular coordinate plane
According to figure, the first reference axis in the rectangular coordinate plane represents the time in the predetermined period, the rectangular coordinate plane
The second interior reference axis represents the historical traffic data；According to the historical traffic data figure and first reference axis and institute
The region that the second reference axis surrounds is stated, baseline drawing area is determined in the coordinate plane；In the baseline drawing area,
Receive the mouse event that user triggers according to the historical traffic data figure；Area is drawn in the baseline according to the mouse event
Coordinate position in domain determines flow monitoring baseline value.The device includes：First processing module, receives mould at Second processing module
Block and third processing module.
Technical solution 1 can determine whether to be suitble to participate in network code according to ONU uninterrupted features in real time, so as to find
The optimum network coding pair of network code is participated in, improves the utilization rate of network bandwidth, further reduces the congestion of network.Only
It is encoded on the basis of flow monitoring, how there is no primary studies carries out flow monitoring.Technical solution 2 using it is wired with
The mode being wirelessly combined saves the manpower and materials for being largely used to tour, meter reading, improves the degree of automation management level,
Realize low-power consumption, low cost, automation, networking, the intelligence of flow monitoring.But the intelligent flow monitoring side of the program
Method is based on wireless sensor network and 3G network, and cost is higher, and be easy to cause data congestion, and there is no consider optimization problem.
Technical solution 3 can accurately determine flow monitoring baseline value, avoid reporting by mistake and fail to report.Focus on studying flow detection limiting value
Determine, there is no the methods of specific research flow monitoring.
Invention content
The present invention is precisely in order to a kind of probe based on vertex covering and weak vertex cover for overcoming drawbacks described above and providing
Dispositions method.It is an object of the invention to study the probe deployment method based on Greedy strategy, by asking Minimum Vertex Covering
The greedy algorithm of topic is improved, and the Greedy strategy that method is limited by branch realizes probe deployment problem.It can ensure algorithm more
Simplicity takes shorter and less using number of probes.
Present invention firstly provides the probe deployment scheme based on Minimum Vertex Covering, experiment shows that the algorithm can acquire most
The optimal solution of small vertex covering, but its computation complexity is high, spends the time long, can not be applied in extensive problem.And
The optimal solution of the optimal solution of Minimum Vertex Covering not necessarily Network Traffic Monitoring probe deployment.Thus in Minimum Vertex Covering
Probe deployment scheme on the basis of improved, it is proposed that the probe deployment scheme based on minimum weak vertex cover, with based on
The deployment scheme of Minimum Vertex Covering is compared, and the number of probes that this programme uses is less, and algorithm is simpler, time-consuming shorter, is
A kind of more excellent Network Traffic Monitoring probe deployment scheme.
The present invention is achieved by the following technical solution：
A kind of probe deployment method based on vertex covering and weak vertex cover, the invention is characterised in that, including：It determines to survey
Sound out the deployed position of the quantity and the test probe of needle in a network.
In the probe deployment method based on fixed point covering, determine to test the quantity of probe and the test probe in network
In deployed position the step of be：
S1. it sorts to the degree of nodes all in network, is met the degree of preceding k-1 node simultaneously and less than number of links
With degree and more than link said conditions the k of preceding k node；
S2., artwork data is configured to the node of a solution space tree, solution is judged whether there is using strategy is delimited, if nothing
K is added 1 by solution, reenters S2, if there is that may have solution to be then inserted into Priority Queues；
S3. if Priority Queues is not sky, then first feasible node is just taken out from Priority Queues, into S4, such as
K is then added 1 by fruit Priority Queues for sky, reenters S2；
S4. judge whether present node meets the condition of solution, exited if meeting just output solution, if being unsatisfactory for just entering
S5；
S5. check whether present node can extend, it is impossible to it is just continued cycling through if extension into S3, if can extend
Words then extend, and then verification expands to whether left and right node has solution, the expanding node for having solution is inserted into Priority Queues, then
It is continued cycling through into S3.
In the probe deployment method based on minimum weak vertex cover, the quantity of test probe and the test probe are determined
The step of deployed position in a network is：
S1. all degree are first deleted as 1 node, that is, delete the row that the sum of all row elements are 1 in incidence matrix；
S2. the most node of a number of links included is chosen, is denoted as vi；
S3. leave out v in incidence matrixiRow where corresponding row and the element that the row intermediate value is 1；Then in remaining pass
Other rows of the sum of all row elements no more than 1 and these row intermediate values are deleted in connection matrix successively again as corresponding to 1 element
Row, until cannot delete new row and column again；
S4. S2, the operation of S3, until all links are all incorporated into are repeated.
Before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：
Model is established using optimal queue and space tree；After model is established, the quantity and the probe that determine test probe exist
Deployed position in network model.
In the present invention, model is established using optimal queue and space tree, including:
Safeguard that the node of present feasible, each node maintenance are also an option that in the case of the node with a Priority Queues
Vertex number k1, need the remaining number of edges e covered, the state state on vertex, vertex the information such as number of edges Degree, these
Node sorts from big to small by the number of edges on vertex, wherein for the state on vertex, state there are three types of vertex in the strategy, respectively
State s1, unselected state s2, the state s3 that can be selected having selected for, wherein, s1 pairs of the state having selected for
The left sibling in solution space tree number is answered, selects the node, then which is set for state s1 has been selected；Unselected state
Right node in s2 homographic solution space trees, does not select the node, then sets the node not select state s2；It can select
Father node in state s3 homographic solution space trees, selects the node, then sets the node as that can select state s3；S1 and s2
All it is the state having determined, the selection of node can only be chosen from s3.
Before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：
Establish the detection model based on minimum weak vertex cover；After model is established, the quantity of test probe and the probe are determined
Deployed position in network detection model.
In the present invention, the detection model based on minimum weak vertex cover is established, including：
The incidence matrix of a node and link is established, the matrix is using link as row, using node as row, if the link includes
The node, then for value for 1, be otherwise 0.
By probe, the comprehensive monitoring to business to the network layer of data network is realized.It is not only of high cost if all standing,
And big amount measurement data can be also generated when measuring, cause noise.In the network of n network element composition, between k network element of measurement
Traffic characteristic, need to choose k network element from n, furthermore, it is contemplated that flowing to problem, also need the arrangement group of k node of consideration
It closes, measurement scheme just needs C (n, k) P (k, k).It is if complicated according to the time caused by the measurement method and scheme of this all standing
It spends for factorial grade, space complexity is also very huge.
Needs around business monitoring set out, and how research disposes suitable probe monitors need to meet different business
It asks, to solve the problems, such as deployed position and deployment capacity.It is required that the deployment examples solved are " certain data network topological diagram ".
1. the link metric in data network in hypothesis problem is all equal, it is assumed that is 1；
2. the communication routing rule in hypothesis problem between network element uses shortest path algorithm；
3. in minimum weak vertex cover model, it is assumed that a network element can monitor directly relevant all links
Flow, and meet flow conservation；
4. in minimal set cover model, it is assumed that arbitrary two network element can pass through all chains on its path of communication monitoring
The state on road, network monitor order are happened between probe and probe；
5. the number of degrees due to network edge node are usually smaller, if the position of detection site is selected in the edge of network
Place, then the number of the detection sent out from the detection site and the ability of detection will receive corresponding limitation, thus false
If probe location will not be at the edge of network.
For the business demand of Network Traffic Monitoring, the detection mode of NetFlow can be used.A certain node in a network
Setting probe can obtain the flow of all chain roads being connected with this node；Therefore, all links in network in order to obtain
Network flow, generally can by certain switching nodes (router) be configured monitor realize, we will consider the problems of
It is：Monitor is set on network which node, could be caused under conditions of it can obtain each link flow, required stream
It is minimum to measure monitor number.
This can be solved the problems, such as using Minimum Vertex Covering.It is complete that vertex cover problem has proved to be NP, can only
Approximate optimal solution is calculated with approximate data.We are herein improved the greedy algorithm of minimal covering problem, pass through
The Greedy strategy that branch limits method realizes probe deployment problem.
While in view of network monitor mode, in addition the constraint of flow conservation, it can using minimum weak vertex cover
Less probe is disposed to realize the Network Traffic Monitoring of whole network.Similary minimum weak vertex cover problem also has proved to be
NP is complete and can only seek approximate optimal solution using approximate data.We realize probe deployment used here as Greedy strategy
Problem.
For the evaluation index of flow monitoring test:
1. the number of the probe of deployment；
2. link coverage rate.
We need the network for needing to dispose probe to be converted into non-directed graph G (V, E), wherein V=(v1,v2,…,vn)
For network node collection (router can be seen as in an ip network), E=(e1,e2,…,em) for the link set in network.Its
In, n=| V |, m=| E | the number of G interior joints and link is represented respectively.Use ek=(vi,vj) representekIt is connecting nodeviAnd vj
A link.
The degree of the number of links, i.e. node v of node v is represented with Degree (v).
Sequence label equipment such as the all-router of primitive network figure, interchangers according to 1,2 ... ..., then counts
The figure is converted into non-directed graph by all link informations.The network has 160 nodes, 185 links.It is a sparse graph (side
Item number m be much smaller than n2Figure), non-directed graph can be represented with corresponding adjacent chained list, it is contemplated that ordinary circumstance, in order to solve
Various figures represents non-directed graph using adjacency matrix.
The probe deployment scheme based on Minimum Vertex Covering is introduced first：
According to being defined below, problem model can very easily be described
Define 1：Given non-directed graph G (V, E), wherein V is vertex set, and E is side collection, and S is the subset of V, if according to being pushed up in S
The flow on the associated each side of point, it may be determined that the flow of any limit in E, then it is measurement sets of the figure G about flow to claim S.
The target for effectively measuring problem is exactly the minimum measurement set for seeking given figure G about flow.It can be converted to and ask definition 2
The minimal covering problem of middle figure G.
Define 2 (minimal covering problems)：A fixed non-directed graph G (V, E) is showed, asks a most boy of vertex set V
Collect S so that e=(u, v) ∈ E, and the point that any side in u ∈ S or v ∈ S, i.e. E at least contains in this subset is used as top
Point, that is to say, that vertex covering in S side collection E.
Its mathematical description is：
Min cx wherein xi∈S
Wherein c=[1,1 ..., 1] and x=[x1,x2,…,xn] be all a length of n vector, (aij)n×nIt is the adjoining square of figure
Battle array, and
It is complete that minimal covering problem has proved to be NP, up to the present there is not yet polynomial time algorithm
Solve, we can remove to be obtained with greedy algorithm approximate optimal solution, i.e., the nodes of each iteration all degree of selection maximums as probe,
Then all coupled nodes can exclude, and meet the optimal situation for choosing that probe is all situation instantly every time, but
Allow for it is each it is optimal may not be able to obtain on the whole optimal, in order to find smaller solution, pass through following restriction
Strategy, we are improved greed, and more excellent solution is obtained by branch-bound method.
It sorts from big to small to the degree of all nodes, k node before taking meets
When, it is the boundary that possible meet vertex covering.
By above-mentioned constraint, we are first obtained a smallest possible optimal solution k, will ask before greedy algorithm is used
Topic is converted into judgement non-directed graph meets Minimum Vertex Covering with the presence or absence of k node.It is gone by Greedy strategy and limiting policy
Judgement, k is optimal solution if meeting, if being unsatisfactory for k adding 1, continues to judge.
Before specific decision process is introduced, model is established using optimal queue and space tree first, it is excellent with one
The node of first queue maintenance present feasible, each node maintenance the vertex number k being also an option that in the case of the node1, need
The remaining number of edges e to be covered, the state state on vertex, vertex the information such as number of edges Degree, the sequences of these nodes follows greedy
Heart strategy sorts from big to small by the number of edges on vertex.For the state on vertex.State there are three types of vertex in the strategy, respectively
The state S1, unselected state S2, the state S3 that can be selected having selected for.Wherein, S1 pairs of the state having selected for
The left sibling in solution space tree number is answered, selects the node, then which is set for state S1 has been selected；Unselected state
Right node in S2 homographic solution space trees, does not select the node, then sets the node not select state S2.It can select
Father node in state S3 homographic solution space trees, selects the node, then sets the node as that can select state S3.S1 and S2
All it is the state having determined, the selection of node can only be chosen from S3.
First in optimal queue node is taken out to be extended：
The node is set as S1, while the information of the node is changed according to formula 4, the node is then put into space tree
Left sibling；The node is set as S2 again, the right node of space tree is directly placed into without changing other nodal informations.
k1=k1- 1, e=e-Degree, Degree=0 4
Pass through constraintsFirst right node is judged, if satisfied, the point is inserted into optimal queue；
It chooses left sibling again to be judged, if satisfied, the node is inserted into optimal queue.
Iteration to previous step is exactly specific decision process.
Corresponding flow chart is as shown in Figure 1.
It is improved on the basis of the probe deployment algorithm based on Minimum Vertex Covering, it is proposed that based on minimum weak vertex
The probe deployment algorithm of covering：
On the basis of defining 1, since probe is arranged on the switching equipment such as router or interchanger, kept according to flow
It is permanent, then figure G also meets following two constraints：
1. the arbitrary vertex v in the vertex set V of couple figure G, degree Degree (v) >=2；
2. the arbitrary vertex v in the vertex set V of couple figure G, meets stream conservation equation, that is, flows into=flow out.
Although the distortion that following reason can will lead to stream conservation equation, such as：Switching equipment is source or the remittance of data, and not only
Only it is headend；Multicast causes the data of output port to replicate；Switching equipment packet delay in itself or loss.But
Several studies have shown that relative error possessed by stream conservation equation is less than 0.05%.
Define 3 (weak vertex covers)：It is assumed that non-directed graph G (V, E) mesh satisfaction there are Degree (v) >=2 to arbitrary v ∈ V, claimIt is the weak vertex cover collection of figure G, can makes side all in E that can be labeled and if only if performing following operation：
(1) all sides associated with vertex in S are marked；
(2) if oneself is labeled on the associated sides of Degree (v) -1 of some vertex v, that remaining correlation is marked
The side of connection；
(3) (2) step is repeated, until cannot mark new side again.
Obviously, thus weak vertex cover S can be obtained by the flow of each link in figure G.The general of incidence matrix is utilized herein
The model established and solve weak vertex cover problem is read, is first provided thus defined below.
4 (incidence relations) are defined in non-directed graph G (V, E), if v ∈ V are one of vertex of e ∈ E, claim to deposit between v and e
In incidence relation, it is denoted as vRe
Define the incidence matrix A=(a of 5 (incidence matrix) figure G (V, E)ij) be
Refer to the n × m matrixes being defined as below:
According to the definition of incidence matrix, it is known that a subset of V forms a covering of figure G, is included and if only if at it
Node corresponding to incidence matrix row in each column at least there are one 1.
According to above analysis, weak vertex cover problem can be solved using greedy algorithm, it should be noted that our net
Boundary point degree in network figure is 1, and is unsatisfactory for constraint 1, so we need to pre-process non-directed graph.
The vertex v for being 1 for degree under initial situationi, only exist there are one side with the point and be associated with, which is denoted as ej, ejIt is another
One vertex is denoted as vj。
Work as vjOnly there are one the adjacent vertices v that degree is 1iAs long as obtain vjThe flow information of other dependence edges just can be according to flow
Calculate eijFlow.That is vjMeet minimum weak vertex cover, only calculating the degree on the side needs eijIn addition；Or vjIt is probe
Also e can be directly monitored during deployed positionijFlow.
Work as vjHave more (k | k ＞ 1) a degree for 1 adjacent vertices, if not node vjProbe is set as, then at least needs to set k
Probe, so only by node vjProbe is set as, just can guarantee that deployment probe mesh number is minimum.
According to being analyzed above, it is known that the node that initial situation degree is 1 can not set probe, can directly be rejected, but
When a node spent for 1 is given up in pretreatment, the degree of other coupled nodes is not subtracted 1.
In view of the above-mentioned flow monitoring model based on minimum weak vertex cover, it is as follows to design its algorithm flow：
Step1：All degree are first deleted as 1 node, that is, delete the row that the sum of all row elements are 1 in incidence matrix；
Step2：The most node of a number of links included is chosen, is denoted as vi；
Step3：Leave out v in incidence matrixiRow where corresponding row and the element that the row intermediate value is 1；Then remaining
Incidence matrix in delete successively again the sum of all row elements no more than 1 other rows and these row intermediate values it is right for 1 element
The row answered, until cannot delete new row and column again；
Step4：Repeat Step2, the operation of Step3, until all links are all incorporated into.
Corresponding algorithm flow chart is as shown in Figure 2.
The key problem in technology point of the present invention is：
1st, this algorithm is improved the greedy algorithm of minimal covering problem, and the Greedy strategy of method is limited by branch
Realize probe deployment problem；
2nd, while in view of network monitor mode, in addition the constraint of flow conservation, uses minimum weak vertex cover energy
Less probe is enough disposed to realize the Network Traffic Monitoring of whole network.
Beneficial effects of the present invention：This patent proposes a kind of probe deployment skill based on vertex covering and weak vertex cover
Art the advantage is that the technology is changed based on the greedy algorithm to minimal covering problem on the basis of the algorithm
Into, it is proposed that the probe deployment algorithm based on vertex covering and weak vertex cover.Minimum Vertex Covering is used first so that can
Under conditions of obtaining each link flow, required flow monitor number is minimum.It is improved to minimum weak vertex cover later.
Emulation shows that compared with the deployment scheme based on Minimum Vertex Covering the number of probes that this programme uses is less, and algorithm is simpler
It is single, it is time-consuming shorter, it is a kind of more excellent Network Traffic Monitoring probe deployment scheme.
Description of the drawings
Fig. 1 is the probe deployment algorithm flow chart based on Minimum Vertex Covering；
Fig. 2 is the algorithm flow chart based on minimum weak vertex cover.
Specific embodiment
A kind of probe deployment method based on vertex covering and weak vertex cover, the invention is characterised in that, including：It determines to survey
Sound out the deployed position of the quantity and the test probe of needle in a network.
In the probe deployment method based on fixed point covering, determine to test the quantity of probe and the test probe in network
In deployed position the step of be：
S1. it sorts to the degree of nodes all in network, is met the degree of preceding k-1 node simultaneously and less than number of links
With degree and more than link said conditions the k of preceding k node；
S2., artwork data is configured to the node of a solution space tree, solution is judged whether there is using strategy is delimited, if nothing
K is added 1 by solution, reenters S2, if there is that may have solution to be then inserted into Priority Queues；
S3. if Priority Queues is not sky, then first feasible node is just taken out from Priority Queues, into S4, such as
K is then added 1 by fruit Priority Queues for sky, reenters S2；
S4. judge whether present node meets the condition of solution, exited if meeting just output solution, if being unsatisfactory for just entering
S5；
S5. check whether present node can extend, it is impossible to it is just continued cycling through if extension into S3, if can extend
Words then extend, and then verification expands to whether left and right node has solution, the expanding node for having solution is inserted into Priority Queues, then
It is continued cycling through into S3.
In the probe deployment method based on minimum weak vertex cover, the quantity of test probe and the test probe are determined
The step of deployed position in a network is：
S1. all degree are first deleted as 1 node, that is, delete the row that the sum of all row elements are 1 in incidence matrix；
S2. the most node of a number of links included is chosen, is denoted as vi；
S3. leave out v in incidence matrixiRow where corresponding row and the element that the row intermediate value is 1；Then in remaining pass
Other rows of the sum of all row elements no more than 1 and these row intermediate values are deleted in connection matrix successively again as corresponding to 1 element
Row, until cannot delete new row and column again；
S4. S2, the operation of S3, until all links are all incorporated into are repeated.
Before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：
Model is established using optimal queue and space tree；After model is established, the quantity and the probe that determine test probe exist
Deployed position in network model.
In the present invention, model is established using optimal queue and space tree, including:
Safeguard that the node of present feasible, each node maintenance are also an option that in the case of the node with a Priority Queues
Vertex number k1, need the remaining number of edges e covered, the state state on vertex, vertex the information such as number of edges Degree, these
Node sorts from big to small by the number of edges on vertex, wherein for the state on vertex, state there are three types of vertex in the strategy, respectively
State s1, unselected state s2, the state s3 that can be selected having selected for, wherein, s1 pairs of the state having selected for
The left sibling in solution space tree number is answered, selects the node, then which is set for state s1 has been selected；Unselected state
Right node in s2 homographic solution space trees, does not select the node, then sets the node not select state s2；It can select
Father node in state s3 homographic solution space trees, selects the node, then sets the node as that can select state s3；S1 and s2
All it is the state having determined, the selection of node can only be chosen from s3.
Before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：
Establish the detection model based on minimum weak vertex cover；After model is established, the quantity of test probe and the probe are determined
Deployed position in network detection model.
In the present invention, the detection model based on minimum weak vertex cover is established, including：
The incidence matrix of a node and link is established, the matrix is using link as row, using node as row, if the link includes
The node, then for value for 1, be otherwise 0.
As shown in table 1, the Minimum Vertex Covering model for method being limited by using branch disposes the network, only need to dispose 43
Probe can be realized as the traffic monitoring network coverage 100%, i.e., the probe deployment plan implementation 43 based on Minimum Vertex Covering
A probe can realize the flow monitoring on each side of network full figure.
1 evaluation index of table
| Network node number | Probe number | The traffic monitoring network coverage |
| 160 | 43 | 100% |
It should be pointed out that although branch limits method and can acquire the optimal solution of Minimum Vertex Covering, computation complexity
Height spends the time long, can not be applied in extensive problem.And the optimal solution of Minimum Vertex Covering not necessarily network
The optimal solution of flow monitoring probe deployment, so easier in order to find algorithm, time-consuming shorter and less using number of probes
Deployment scheme disposes probe below in conjunction with the property of network flow conservation by minimum weak vertex cover.
As shown in table 2, by using the method for minimum weak fixed point covering, only need to dispose 32 probes can be realized as flow
Monitor the network coverage 100%.In this kind of method, the network flow of part path is calculated by flow conservation
's.
2 evaluation index of table
| Network node number | Probe number | The traffic monitoring network coverage |
| 160 | 32 | 100% |
Compared with the deployment scheme based on Minimum Vertex Covering, the number of probes that this programme uses is less, and algorithm is simpler
It is single, it is time-consuming shorter, it is a kind of more excellent Network Traffic Monitoring probe deployment scheme.
But the anti-case of both probe deployments is all Passive Test, can utilize Port Mirroring, multichannel forwarding is with link
The data packet of itself transmission in the modes collection networks such as concatenation, including business data packet, signaling data bag and management information data
Packet etc. analyzes network performance, passively monitoring network performance, and having does not increase added flow, on the advantages of network influence very little.
Claims (5)
- A kind of 1. probe deployment method based on vertex covering and weak vertex cover, which is characterized in that including：Determine test probe Quantity and the test probe deployed position in a network；Wherein,The step of determining the quantity of test probe and the test probe deployed position in a network be：S1. it sorts to the degree of nodes all in network, is met the degree of preceding k-1 node simultaneously and less than number of links with before Degree and more than link said conditions the k of k node；S2., artwork data is configured to the node of a solution space tree, solution is judged whether there is using strategy is delimited, if will without solution K adds 1, reenters S2, if there is that may have solution to be then inserted into Priority Queues；S3. if Priority Queues is not sky, then first feasible node is just taken out from Priority Queues, into S4, if excellent K is then added 1 by first queue for sky, reenters S2；S4. judge whether present node meets the condition of solution, exited if meeting just output solution, if being unsatisfactory for just entering S5；S5. check present node whether can extend, it is impossible to just continued cycling through if extension into S3, if it can extend if Extension, then verification expand to whether left and right node has solution, the expanding node for having solution are inserted into Priority Queues, subsequently into S3 is continued cycling through；The step of determining the quantity of test probe and the test probe deployed position in a network be：S1. all degree are first deleted as 1 node, that is, delete the row that the sum of all row elements are 1 in incidence matrix；S2. the most node of a number of links included is chosen, is denoted as vi；S3. leave out v in incidence matrixiRow where corresponding row and the element that the row intermediate value is 1；Then in remaining association square Other rows of the sum of all row elements no more than 1 and these row intermediate values are deleted in battle array successively again as the row corresponding to 1 element, directly Until it cannot delete new row and column again；S4. S2, the operation of S3, until all links are all incorporated into are repeated.
- 2. a kind of probe deployment method based on vertex covering and weak vertex cover according to claim 1, feature exist In before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：Model is established using optimal queue and space tree；After model is established, the deployed position of the quantity and the probe of test probe in network model is determined.
- 3. a kind of probe deployment method based on vertex covering and weak vertex cover according to claim 2, feature exist In, model is established using optimal queue and space tree, including:The node of present feasible is safeguarded with a Priority Queues, each node maintenance the top being also an option that in the case of the node Count out k1, need the remaining number of edges e covered, the state state on vertex, vertex number of edges Degree information, these nodes press The number of edges on vertex sorts from big to small, wherein for the state on vertex, there are three types of states on vertex in the strategy, have respectively selected State s1, unselected state s2, the state s3 that can be selected selected, wherein, the state s1 homographic solutions having selected for are empty Between left sibling in tree number, select the node, then which be set for state s1 has been selected；Unselected state s2 is corresponded to Right node in solution space tree, does not select the node, then sets the node not select state s2；The state s3 that can be selected Father node in homographic solution space tree, selects the node, then sets the node as that can select state s3；S1 and s2 is Through determining state, the selection of node can only be chosen from s3.
- 4. a kind of probe deployment method based on vertex covering and weak vertex cover according to claims 1 or 2 or 3, It is characterized in that, before the deployed position of the quantity and the probe of test probe in a network is determined, the method further includes：Establish the detection model based on minimum weak vertex cover；After model is established, the deployed position of the quantity and the probe of test probe in network detection model is determined.
- 5. a kind of probe deployment method based on vertex covering and weak vertex cover according to claim 4, feature exist In, the detection model based on minimum weak vertex cover is established, including：The incidence matrix of a node and link is established, the matrix is using link as row, using node as row, if the link includes the section Point, then corresponding value is 1, is otherwise 0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510807537.3A CN105490834B (en) | 2015-11-19 | 2015-11-19 | A kind of probe deployment method based on vertex covering and weak vertex cover |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510807537.3A CN105490834B (en) | 2015-11-19 | 2015-11-19 | A kind of probe deployment method based on vertex covering and weak vertex cover |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105490834A CN105490834A (en) | 2016-04-13 |
| CN105490834B true CN105490834B (en) | 2018-06-26 |
Family
ID=55677578
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510807537.3A Expired - Fee Related CN105490834B (en) | 2015-11-19 | 2015-11-19 | A kind of probe deployment method based on vertex covering and weak vertex cover |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105490834B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108737455B (en) * | 2017-04-13 | 2020-12-25 | 国家电网有限公司 | Network service identification device and method |
| CN107147534A (en) * | 2017-05-31 | 2017-09-08 | 国家电网公司 | A kind of probe deployment method of quantity optimization for power telecom network fault detect |
| CN107483309B (en) * | 2017-07-24 | 2020-10-23 | 中国空间技术研究院 | Distributed network system minimum weight vertex coverage solving method based on game theory |
| CN109714220B (en) * | 2017-10-25 | 2020-11-13 | 中国电信股份有限公司 | Network quality detection method and device |
| CN108198414A (en) * | 2017-12-27 | 2018-06-22 | 北斗七星（重庆）物联网技术有限公司 | A kind of method, apparatus, equipment and the storage medium of road monitoring point position distribution |
| CN109041071A (en) * | 2018-07-27 | 2018-12-18 | 北京国电通网络技术有限公司 | A kind of electric power wireless private network probe deployment method and apparatus |
| CN109299443B (en) * | 2018-09-04 | 2023-04-14 | 中山大学 | News text duplication eliminating method based on minimum vertex coverage |
| CN112422322A (en) * | 2020-10-28 | 2021-02-26 | 北京邮电大学 | Multi-objective optimization method based on probe scheduling and related equipment |
| CN113965400B (en) * | 2021-11-01 | 2023-06-30 | 电子科技大学长三角研究院(衢州) | Method for determining flow key points in communication network |
| CN116232956B (en) * | 2023-05-06 | 2023-07-18 | 国网智能电网研究院有限公司 | Network time delay in-band telemetry method, device, electronic equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103546319A (en) * | 2013-10-18 | 2014-01-29 | 中国南方电网有限责任公司 | Flow alarming method and system of network equipment |
| EP2882141A1 (en) * | 2013-12-04 | 2015-06-10 | Exfo Inc. | Network test system |
-
2015
- 2015-11-19 CN CN201510807537.3A patent/CN105490834B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103546319A (en) * | 2013-10-18 | 2014-01-29 | 中国南方电网有限责任公司 | Flow alarming method and system of network equipment |
| EP2882141A1 (en) * | 2013-12-04 | 2015-06-10 | Exfo Inc. | Network test system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105490834A (en) | 2016-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105490834B (en) | A kind of probe deployment method based on vertex covering and weak vertex cover | |
| Yoon et al. | Exploiting spatial correlation towards an energy efficient clustered aggregation technique (cag)[wireless sensor network applications] | |
| CN104125153B (en) | Method for discovering network topology and equipment | |
| CN108366386B (en) | Method for realizing wireless network fault detection by using neural network | |
| CN104468192B (en) | The method for routing that a kind of multiple dimensioned many weight link-qualities are assessed | |
| CN107483487B (en) | TOPSIS-based multi-dimensional network security measurement method | |
| CN104935476B (en) | A kind of network traffics matrix measuring method based on SDN | |
| CN107332631A (en) | A kind of method of use multi-attribute group decision making theoretical evaluation link-quality assessment models | |
| CN103281256B (en) | The end-to-end path packet loss detection method of chromatography Network Based | |
| CN105120477B (en) | A kind of mobile network optimization method and system | |
| CN103747537A (en) | Wireless sensor network outlier data self-adaption detecting method based on entropy measurement | |
| CN110049528A (en) | Mobile trust data collection method based on trust value effectiveness in a kind of Sensor Network | |
| CN105357132A (en) | Multi-domain ASON damage perception multicast routing method based on hypergraph model | |
| CN102883359A (en) | Method, device and system for measuring key nodes of wireless sensor network | |
| CN101610246B (en) | Method for inferring source-target flow component of link layer of local area network | |
| Qi et al. | Distributed combat system of systems network modeling | |
| CN106850253A (en) | A kind of method of the transmission time reliability measurement based on multimode network | |
| CN104093182A (en) | Method for acquiring a plurality of reliable communication paths based on field intensity in multi-layer wireless network | |
| CN106603294B (en) | A kind of synthesis vulnerability assessment method based on power communication web frame and state | |
| CN113411766A (en) | Intelligent Internet of things comprehensive sensing system and method | |
| Chen et al. | Using node clustering and genetic programming to estimate missing data for marine environmental monitoring | |
| CN111654413B (en) | Method, equipment and storage medium for selecting effective measurement points of network flow | |
| Zhao et al. | MPIDA: A sensor network topology inference algorithm | |
| Torres et al. | Modeling, analysis and simulation of ant-based network routing protocols | |
| Wan et al. | Capturing Spatial-Temporal Correlations with Attention Based Graph Convolutional Networks for Network Traffic Prediction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180626 Termination date: 20181119 |