CN102364899A - Particle-swam-optimization-based virtual network mapping method and system - Google Patents
Particle-swam-optimization-based virtual network mapping method and system Download PDFInfo
- Publication number
- CN102364899A CN102364899A CN2011103640769A CN201110364076A CN102364899A CN 102364899 A CN102364899 A CN 102364899A CN 2011103640769 A CN2011103640769 A CN 2011103640769A CN 201110364076 A CN201110364076 A CN 201110364076A CN 102364899 A CN102364899 A CN 102364899A
- Authority
- CN
- China
- Prior art keywords
- particles
- particle
- mrow
- optimal position
- virtual network
- 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
- 238000013507 mapping Methods 0.000 title claims abstract description 147
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000005457 optimization Methods 0.000 title claims abstract description 30
- 239000002245 particle Substances 0.000 claims abstract description 173
- 238000012360 testing method Methods 0.000 claims description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 16
- 238000007689 inspection Methods 0.000 claims description 3
- 230000007774 longterm Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 7
- 238000011156 evaluation Methods 0.000 description 4
- 238000009827 uniform distribution Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000004088 simulation Methods 0.000 description 2
- 101150110390 Slc10a6 gene Proteins 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000003116 impacting effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a particle-swarm-optimization-based virtual network mapping method and a particle-swarm-optimization-based virtual network mapping system. The method comprises the following steps of: initializing position and speed parameters of particles; performing feasibility checking on the particles to obtain feasible particles, and determining an initial global optimal position and individual optimal positions of each particle; performing speed and position updating on the feasible particles, performing the feasibility checking on the updated particles to obtain feasible particles again, determining a global optimal position and individual optimal positions again, adding 1 to a current number of iterations, and repeating the operations of the step when the number of iterations is less than a maximum number of iterations; and when the number of iterations is equal to the maximum number of iterations, outputting the finally determined global optimal position as a mapping scheme. By the method and the system, the position and speed parameters of a particle swarm are initialized under the condition that a bottom-layer network does not support path splitting, and the particles are iterated to select an optimal virtual network mapping scheme, so the resource utilization efficiency of the bottom-layer network is improved.
Description
Technical Field
The present invention relates to the field of computer networks, and in particular, to a virtual network mapping method and system based on Particle Swarm Optimization (PSO).
Background
Network virtualization technology allows multiple heterogeneous virtual networks to coexist on top of a shared underlying physical network infrastructure, making it possible to deploy new network architectures, protocols, and applications without impacting existing networks, thus effectively introducing network technological innovation.
In a network virtualization environment, an infrastructure provider manages an operating underlay network, and a service provider can lease network resources to the infrastructure provider and create a virtual network to provide personalized end-to-end network services. The problem of allocating infrastructure provider's underlying network resources for a service provider's virtual network request with node and link resource constraints is known as the virtual network mapping (embedding, assignment) problem. The virtual network mapping problem is a Non-deterministic Polynomial (NP) difficult problem, even after all virtual nodes have been mapped, mapping virtual links with bandwidth resource constraints is still NP-hard. The virtual network mapping problem is one of the main challenges facing network virtualization technology, and has become a hot issue in this research field.
The Virtual Network mapping problem is similar to the Virtual Private Network (VPN) provisioning problem and the Network laboratory bed mapping problem. However, a typical VPN request only includes a traffic matrix between point pairs with bandwidth constraints, so that only a path satisfying the bandwidth constraints between the point pairs needs to be searched in the underlying network to solve the problem. The virtual network mapping problem is fundamentally different from the above problem in that the virtual network request includes not only bandwidth resource constraints but also node constraints (e.g., computing power and location requirements, etc.). For the network experimental bed mapping problem, although the node and link constraint conditions in the experimental network request topology need to be considered simultaneously when the mapping of the experimental network to the underlying network is constructed, the network experimental bed mapping problem is different from the virtual network mapping problem in that the underlying network does not allow different experimental networks to share the same underlying network nodes.
Because of the complexity of the virtual network mapping problem, early studies on the virtual network mapping problem all limited the problem to varying degrees: assuming that all virtual network requests are known; ignoring resource requirements of the virtual network nodes or links; assuming that the underlying network has sufficient resources, not considering the admission control requested by the virtual network; only the virtual network request with a special topological structure is concerned, such as the virtual network request with a star topological structure; the location requirements of the virtual network nodes are ignored. In the existing virtual network mapping problem, in the case that the underlying network supports path splitting, a mixed integer programming model of virtual network mapping is established, and a virtual network mapping algorithm of cooperative nodes and links is proposed, which comprehensively considers the above aspects, but still has the following problems: under the condition that the underlying network does not support path splitting, a corresponding mapping algorithm is not provided under a correct model; the existing proposed mapping algorithm is not only high in time overhead, but also the obtained mapping scheme is not always feasible, even if feasible, the mapping scheme is not always optimal, and network resources cannot be effectively utilized.
Disclosure of Invention
In view of the above, the main objective of the present invention is to provide a virtual network mapping method and system based on particle swarm optimization, which can improve the utilization efficiency of underlying network resources.
In order to achieve the purpose, the technical scheme of the invention is realized as follows:
a particle swarm optimization-based virtual network mapping method, the method comprising:
initializing the position parameters and the speed parameters of the particles;
carrying out feasibility test on the particles to obtain feasible particles, and determining an initial global optimal position and an individual optimal position of each particle;
updating the speed and the position of the feasible particle, carrying out feasibility inspection on the updated particle, obtaining the feasible particle again, re-determining the global optimal position and the individual optimal position of each particle, adding 1 to the current iteration number, and repeating the step when the iteration number is less than the preset maximum iteration number;
and when the iteration times are equal to the preset maximum iteration times, outputting the finally determined global optimal position as a mapping scheme.
Wherein the initializing the position parameter and the velocity parameter of the particle is:
the position parameter X of the particlesiInitialized to the ith virtual network mapping scheme, speed parameter ViAn adjustment decision for the virtual mapping scheme is initialized.
Wherein, the feasibility test is carried out on the particles, and the feasible particles are obtained by:
detecting the capability constraint of the underlying network nodes related to the position parameters of the particles;
and when the CPU capacity of the underlying network node meets the CPU capacity of the virtual network node and is positioned in the range of the position requested by the virtual network node, detecting the bandwidth and connectivity constraint conditions of the position parameter by adopting a shortest path algorithm to obtain the fitness of the particle.
Wherein the determining the initial global optimal position and the individual optimal position of each particle is as follows:
and determining the position of the particle with the minimum fitness in all the particles as an initial global optimal position, and determining the position of the particle corresponding to the minimum fitness in the obtained multiple fitness of each particle as an initial individual optimal position of the particle.
Further, after obtaining viable particles, the method further comprises:
and re-initializing the position parameters and the speed parameters of the particles except the feasible particles.
Wherein the re-determining the global optimal position and the individual optimal position of each particle is:
calculating the fitness of each particle, and when the obtained fitness is smaller than the fitness corresponding to the previously determined individual optimal position, re-determining the individual optimal position as the position corresponding to the current fitness;
and when the fitness corresponding to the redetermined individual optimal position is smaller than the fitness corresponding to the previously determined global optimal position, redetermining the global optimal position as the redetermined individual optimal position.
Further, the method further comprises: and outputting the finally determined fitness value corresponding to the global optimal position to obtain the bandwidth overhead corresponding to the mapping scheme.
A particle swarm optimization-based virtual network mapping system, the system comprising: the device comprises an initialization unit, a feasibility checking unit, an updating unit and an output unit; wherein,
the initialization unit is used for initializing the position parameter and the speed parameter of the particles;
the feasibility testing unit is used for carrying out feasibility testing on the initialized particles and determining an initial global optimal position and an individual optimal position of each particle; carrying out feasibility test on the particles updated by the updating unit to obtain feasible particles, re-determining the global optimal position and the individual optimal position of each particle, and adding 1 to the current iteration number;
the updating unit is used for updating the speed and the position of the feasible particles obtained by the feasibility checking unit when the current iteration number is smaller than the preset maximum iteration number;
and the output unit is used for outputting the global optimal position finally determined by the feasibility testing unit as a mapping scheme when the current iteration number is equal to the preset maximum iteration number.
The initialization unit is further configured to reinitialize the position parameter and the velocity parameter of the particle other than the feasible particle.
The output unit is further configured to output a fitness value corresponding to the finally determined global optimal position, so as to obtain a bandwidth overhead corresponding to the mapping scheme.
The method initializes the position parameters and the speed parameters of the particle swarm under the condition that the underlying network does not support path splitting, realizes the selection of the optimized virtual network mapping scheme by iterating the particles, improves the resource utilization efficiency of the underlying network, can provide space for receiving more virtual network requests, can well balance the optimization degree and the time complexity of the virtual network mapping scheme, and can be widely applied to a backbone network or data center network environment supporting a network virtualization technology.
Drawings
FIG. 1 is a schematic diagram of an implementation process of a particle swarm optimization-based virtual network mapping method according to the present invention;
FIG. 2 is a schematic diagram of a mapping scheme obtained by applying the particle swarm optimization-based virtual network mapping method of the present invention;
fig. 3 is a schematic diagram of long-term average operation yield curves of an underlying network of three virtual network mapping methods;
FIG. 4 is a schematic diagram of a virtual network request acceptance rate curve of three virtual network mapping methods;
FIG. 5 is a diagram illustrating long-term average profit-to-cost ratio curves of an underlying network for three virtual network mapping methods;
FIG. 6 is a schematic diagram showing a comparison of the operation times of three virtual network mapping methods;
fig. 7 is a schematic structural diagram of a virtual network mapping system based on particle swarm optimization according to the present invention.
Detailed Description
First, a brief introduction is made to concepts involved in the field of virtual network mapping.
Bottom layer network: the underlying network topology can be marked as a weighted undirected graph
Wherein N issRepresenting a set of underlying network nodes, LsRepresenting a set of underlying network links.
And
respectively represent bottom nodes ns(ns∈Ns) And ls(ls∈Ls) A collection of properties. Bottom node nsHaving the attribute of the currently available computing power CPU (n) of the nodes) And physical location Loc (n)s) (ii) a Underlying Link lsHaving the attribute of the currently available bandwidth resource BW (l) of the links)。
Virtual network request: similar to the underlying network, the virtual network topology can also be marked as a weighted undirected graphWherein N isvIs a set of virtual nodes, LvIs a set of virtual links that are,
and
respectively representing virtual nodes nv(nv∈Nv) With virtual links lv(lv∈Lv) The resource constraints of (2). Generally speaking, the resource constraint of a virtual node mainly considers the computing power requirement of the virtual node, and the resource constraint of a virtual link mainly considers the bandwidth resource requirement of the virtual link. For a virtual network request, a triplet VNR may be used(i)(Gv，ta，td) Is shown in which t isaIndicating the arrival time of the virtual network request, tdIndicating the time that the virtual network lasts in the underlying network. When the ith virtual network request arrives, the underlying networkIt should be allocated the corresponding resources that meet its node and link resource requirements. When the virtual network leaves the underlying network, the resources allocated to it will be released. In addition, the request should be delayed from mapping or directly rejected when there are insufficient underlying network resources.
Virtual network mapping: the virtual network mapping problem is defined as mapping: m: gv(Nv，Lv)→Gs(N′s，P′s) In which P issRepresents a loop-free path of all underlying networks and
the mapping can be decomposed into the following two steps: mapping the virtual node to a bottom layer node meeting the node resource constraint condition; and mapping the virtual link to the bottom loop-free path meeting the bandwidth resource constraint condition.
The main evaluation indexes are as follows: the main evaluation indexes of the virtual network mapping include the long-term average operation income of the underlying network, the virtual network request acceptance rate and the long-term average income and expense ratio of the virtual network mapping. Furthermore, for online virtual network mapping, time complexity should also be considered. The evaluation index will be described further below.
The revenue from accepting a virtual network request at time t may be defined as the sum of the computational and bandwidth resource requirements of the virtual network request:
the overhead of accepting a virtual network request at time t may be defined as the sum of the resources allocated to the virtual network request by the underlying network:
wherein
When the underlying link lsTo virtual link lvWhen allocating bandwidth resourcesOtherwise
Is represented bysTo lvThe value of the allocated bandwidth.
The long term average operational yield of the underlay network can be expressed by:
the virtual network request acceptance rate may be expressed as:
wherein
Indicating the number of virtual networks successfully mapped from time T-0 to time T,
representing the total number of virtual network requests from time T-0 to time T.
The underlying network long term average revenue to overhead ratio can be expressed as:
the virtual network mapping method disclosed by the invention introduces the following related concepts:
binary variable
In the mapping process, if the physical link (i, j) carries the virtual link (u, v), the variable takes a value of 1, otherwise, the variable takes a value of 0.
Binary variableIn the mapping process, if virtual node u is mapped onto physical node i, the changeThe quantity takes the value of 1, otherwise the value takes the value of 0.
For a virtual network mapping request, the CPU overhead is the same for different mapping schemes, but the bandwidth overhead is different, so we take the following equation as the objective function of the model:
in terms of node constraints, the underlying network node's CPU power must be able to meet the CPU power requirements of the virtual network node, and the underlying network node must be located within D from the requested location of the virtual node. The formalized description of the constraints for nodes and links is given by:
if virtual nodes u and v map to underlying network nodes i and j, respectively, during the virtual network node mapping phase, then virtual link (u, v) will be mapped onto an underlying physical path P from node i to node j during the virtual network link mapping phase. At the source point i, the outflow rate is 1 and the inflow rate is 0, so
At the junction j, the outflow rate is 0 and the inflow rate is 1, and thereforeOn the other nodes of the path P, the outgoing and incoming flows are both 1, so <math>
<mrow>
<munder>
<mi>Σ</mi>
<mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>∈</mo>
<msub>
<mi>L</mi>
<mi>s</mi>
</msub>
</mrow>
</munder>
<msubsup>
<mi>f</mi>
<mi>ij</mi>
<mi>uv</mi>
</msubsup>
<mo>-</mo>
<munder>
<mi>Σ</mi>
<mrow>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>∈</mo>
<msub>
<mi>L</mi>
<mi>s</mi>
</msub>
</mrow>
</munder>
<msubsup>
<mi>f</mi>
<mi>ji</mi>
<mi>uv</mi>
</msubsup>
<mo>=</mo>
<mn>0</mn>
<mo>.</mo>
</mrow>
</math> The connectivity constraints are as follows:
for the same virtual network request, one underlying network node can only bear one virtual node, and one virtual network node can only be mapped to one underlying network node. Thus, variable
And
with the following constraints:
(10)
(11)
finally, the PSO algorithm is briefly described: the algorithm is a global random search algorithm based on group intelligence and proposed by Kennedy and Eberhart, and has the advantages of high execution speed, high efficiency and the like compared with similar optimization algorithms. In the algorithm, each particle moves in a solution space at a certain speed and moves to a historical optimal position XpbAnd neighborhood history best position XgbAnd aggregating to realize the evolution of the candidate solution. The velocity and position update formula for the particles is as follows:
Vi+1＝wVi+c1r1(Xpb-Xi)+c2r2(Xgb-Xi) (12)
x in the formulae (12) and (13)iIndicates the current position, V, of the ith particleiIndicating the current velocity of the ith particle. W in equation (12) represents the weight of the particle to maintain inertia; c. C1、c2Representing the acceleration of the particles, which respectively represents the trend of the particles moving to the self optimal position and the global optimal position; r is1、r2Are random numbers uniformly generated between (0, 1).
The basic idea of the invention is as follows: initializing the position parameters and the speed parameters of the particles; carrying out feasibility test on the particles to obtain feasible particles, and determining an initial global optimal position and an individual optimal position of each particle; updating the speed and the position of the feasible particle, carrying out feasibility inspection on the updated particle, obtaining the feasible particle again, re-determining the global optimal position and the individual optimal position of each particle, adding 1 to the current iteration number, and repeating the step when the iteration number is less than the preset maximum iteration number; and when the iteration times are equal to the preset maximum iteration times, outputting the finally determined global optimal position as a mapping scheme. The fitness function (6) is labeled as f (X), where the position vector X represents a possible mapping scheme. The processing procedure of the fitness function f (x) is to firstly check the node capacity constraint, specifically check through the formula (7), and then judge the feasibility of the current mapping scheme by adopting a shortest path algorithm so as to check the bandwidth and connectivity constraint conditions. When the mapping scheme is feasible, the value of (f) (x) represents the overhead of the virtual network corresponding to the mapping scheme; when the mapping scheme is not feasible, the value of f (x) is set to + ∞.
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings by way of examples.
The PSO algorithm is mainly used for solving the problem of multi-objective optimization in the continuous domain initially, and when solving the discretization problem with respect to virtual network mapping and the like, the parameters and the like of the particles need to be reset according to specific problems, where the specific settings are as follows:
1. position of the particle: position vector of particle
Is set to the ith possible mapping scheme. D indicates that the virtual network request contains D virtual network nodes in total.
Take a positive integer whose value indicates that the jth virtual node is from its underlying networkAnd the numbers of the bottom network nodes selected from the candidate node list.
2. Velocity of the particles: velocity vector of particle
Configured as an adjustment decision for the mapping scheme for guiding the current mapping scheme to be adjusted to a more optimal mapping scheme. Wherein
Is a binary variable if
Indicating that the jth virtual node needs to reselect a node map from its underlying network candidate node list.
3. Subtraction Θ: xiΘXjFor calculating the difference between the two mapping schemes. If the mapping scheme XiAnd XjHaving the same value in the same dimension, the difference results in a 1, otherwise 0. For example, (1, 2, 3, 4, 5) Θ (1, 3, 2, 4, 6) ═ 1, 0, 0, 1, 0.
4. Addition 
For obtaining an adjustment decision for the mapping scheme. Wherein P isiViAnd PjVjRespectively represent by PiProbability of maintaining ViThe sum of the values of the dimensions is PjProbability of maintaining VjValue of each dimension, and Pi+Pj1 (P is more than or equal to 0 and less than or equal to 1). For example, <math>
<mrow>
<mn>0.1</mn>
<mrow>
<mo>(</mo>
<mn>1,0,0,1,1</mn>
<mo>)</mo>
</mrow>
<mo>⊕</mo>
<mn>0.9</mn>
<mrow>
<mo>(</mo>
<mn>1,0,1,0,1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mn>1,0</mn>
<mo>,</mo>
<mo>*</mo>
<mo>,</mo>
<mo>*</mo>
<mo>,</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
</math> where it indicates that the dimension is uncertain as 0 or 1, e.g., the first in the addition results in this example indicates that the dimension is 0 with a probability of 0.1 and 1 with a probability of 0.9.
5. Multiplication
For obtaining a new mapping scheme. Mapping scheme XiAccording to adjustment decision ViAnd adjusting the virtual node mapping scheme. For example,
indicating that the mapping scheme of the second virtual network node in the mapping scheme needs to be adjusted.
According to the definition, the position and speed updating basic formula of the particle swarm optimization algorithm is obtained again, and the method comprises the following steps:
wherein, P1，P2And P3Is a constant number, and P1+P2+P3＝1。
The following describes in detail the implementation process of the virtual network mapping method of the present invention with reference to fig. 1, where the method includes the following steps:
Specifically, the method comprises the steps of setting the size number N of particle swarms according to an underlying network, comprehensively considering mapping accuracy and time complexity to determine the maximum iteration number MG, and setting the current iteration number of particles to be 0; simultaneously mapping the position vector of the particle with a possible virtual network mapping scheme, and initializing a position parameter XiAnd a speed parameter ViHere, as described above, the position parameter X of the particleiRepresenting the ith possible virtual network mapping scheme, the speed parameter ViFor representing adjustment decisions for the virtual network mapping scheme.
Here, the feasibility test is performed on the particles by calculating the fitness f (X) of each particlei) Wherein the fitness f (X)i) Representing virtual network mapping using mapping scheme XiBandwidth overhead of the corresponding virtual network;
calculating the fitness f (X) of each particlei) Is first examined for a position parameter X of the particleiThe capability constraints of the underlying network nodes involved are checked specifically by the above equation (7); and when the CPU capacity of the underlying network node meets the CPU capacity of the virtual network node and is located in the range of the position requested by the virtual network node, detecting the feasibility of the current mapping scheme by adopting a shortest path algorithm, specifically checking bandwidth and connectivity constraint conditions. When the mapping scheme is feasible, the particle fitness f (X) corresponding to the mapping schemei) Representing the cost of the virtual network mapping scheme, and determining the position of the particle with the minimum fitness in all the particles as an initial global optimal position Xgb. Because the particle is in continuous motion, for the same particle, the particle may be mapped into a plurality of virtual network mapping schemes, and then a plurality of fitness degrees are obtained through calculation, and the position of the particle corresponding to the minimum fitness degree in the plurality of fitness degrees is determined as the initial individual optimal position X of the particlepb。
103, updating the speed and the position of the obtained feasible particles;
specifically, the speed update may be performed according to the above equation (14), and the position update may be performed according to the equation (15); and randomly selecting candidate nodes meeting constraint conditions in the underlying network in the speed and position updating process.
104, carrying out feasibility test again on each particle after the speed and the position are updated in the step 103, obtaining feasible particles again, and determining the totalLocal optimum position XgbAnd individual optimal position X of each particlepbAdding 1 to the current iteration number;
specifically, the feasibility test is carried out by calculating the fitness of each particle, and when the obtained fitness is smaller than the fitness corresponding to the previously determined individual optimal position, the individual optimal position is determined as the position corresponding to the current fitness again;
and when the fitness corresponding to the redetermined individual optimal position is smaller than the fitness corresponding to the previously determined global optimal position, redetermining the global optimal position as the redetermined individual optimal position.
Specifically, the above-mentioned re-determining the global optimal position and the individual optimal position of each particle may be expressed as: if f (X)i)＜f(Xpb) Then Xpb＝Xi(ii) a If f (X)pb)＜f(Xgb) Then Xgb＝Xpb；
In the step, the current iteration times are added by 1 to obtain new current iteration times.
specifically, a mapping scheme corresponding to the minimum fitness value of the particle is output; in addition, the step also comprises the step of enabling the global optimal position X to be usedgbAnd outputting the corresponding fitness value to obtain the bandwidth overhead corresponding to the mapping scheme.
In addition, the method further comprises the following steps: reinitializing particles except for feasible particles in the particle swarm to generate an initial position parameter XiAnd a speed parameter ViAnd then step 102 is executed.
The method is further described with reference to the mapping scheme shown in fig. 2, which is obtained by applying the particle swarm optimization-based virtual network mapping method of the present invention.
The virtual network mapping method based on particle swarm optimization obtains an optimal virtual network mapping scheme on the basis of the topological structure and resource capacity conditions of an underlying network and a virtual network request, the left side of fig. 2 is a received virtual network request, and the right side is the condition that the virtual network request is mapped to the underlying network, wherein the virtual network request on the left side comprises three virtual nodes a, b and c, the number in a rectangle near the node represents the computing resource requirement of the virtual node, and the number near the link represents the bandwidth resource requirement of the virtual link; the topology of the right underlying network includes A, B, C, D, E, F six network nodes, data in a rectangle near a node represents the available computing resources of the underlying node, and data near a link represents the available bandwidth resources of the underlying link.
As can be seen from fig. 2, in the obtained virtual network mapping scheme, the node mapping scheme is { a → E, b → C, C → D }, and the link mapping scheme is { (a, b) → (E, C), (a, C) → (E, D), (b, C) → (C, D) }.
The particle swarm optimization-based virtual network mapping method and the existing D-VinE-LB and D-VinE-SP virtual network mapping method have the following evaluation indexes: the long-term average operation yield (see the above formula (3)) of the underlying network, the virtual network request acceptance rate (see the above formula (4)), the long-term average profit-cost ratio (see the above formula (5)) of the underlying network, and the time-cost aspect are compared:
the experimental setup was as follows: the underlying network topology is provided with 100 nodes and approximately 500 links. The underlying network node CPU resources and bandwidth resources are subject to a uniform distribution of 50-100. Assuming that the arrival of virtual network requests within each 100 time units obeys a poisson process with an average value of 5, the lifetime of each virtual network obeys an exponential distribution with an average lifetime of 500 time units.
For each virtual network request, the virtual network nodes obey a uniform distribution of 2-20, with each pair of virtual network nodes connected with a probability of 0.5. The CPU resource and link bandwidth resource demands of the virtual network nodes are subject to uniform distribution of 0-50.
The network topology and its additional location information can be randomly generated using the GT-ITM tool, the x and y variables of the location coordinates obey a uniform distribution of 0-100, and the location constraints D of all virtual network mapping requests are taken as constants.
About 50000 time units are operated in each simulation experiment, and each simulation experiment comprises 2500 virtual network requests, and for the virtual network mapping method provided by the invention, the specification number N of a particle swarm is preset to be 5, and the maximum iteration number MG executed by an algorithm is 20. P in formula (14)1，P2And P3Are set to 0.1, 0.2 and 0.7, respectively.
Specific experimental results refer to fig. 3 to 6, where fig. 3 shows comparison of long-term average operation revenue of the underlying network for the three virtual network mapping methods, fig. 4 shows comparison of virtual network request acceptance rates for the three virtual network mapping methods, and fig. 5 shows comparison of long-term average revenue cost performance of the underlying network for the three virtual network mapping methods; wherein
The performance dotted line obtained by the particle swarm optimization-based virtual network mapping method provided by the invention is shown,
representing a performance curve obtained by a D-VinE-LB virtual network mapping method,
representing a performance curve obtained by a D-VinE-SP virtual network mapping method; FIG. 6 shows a run-time comparison of three virtual network methods, of "VNE-R-PSO ' represents the particle swarm optimization-based virtual network mapping method provided by the invention, D-VinE-LB ' represents the D-VinE-LB virtual network mapping method, and D-VinE-SP ' represents the D-VinE-SP virtual network mapping method.
As can be seen from fig. 3 and 4, both the particle swarm optimization-based virtual network mapping algorithm and the D-ViNE-SP virtual network mapping method provided by the invention are based on the situation that the underlying network does not support path splitting, and compared with the D-ViNE-SP virtual network mapping method, the particle swarm optimization-based virtual network mapping method significantly improves the long-term average operation yield and the virtual network request acceptance rate of the underlying network, wherein the average yield is improved by about 22%, and the request acceptance rate is improved by about 10%.
Even compared with the D-VinE-LB virtual network mapping method which is suitable for the characteristic that the underlying network supports path splitting, the virtual network mapping method based on particle swarm optimization still has great advantages, for example, when the time unit 40000 is adopted, the long-term average operation income and the virtual network request acceptance rate of the underlying network are respectively improved by about 8% and 7%. The main reason is that the solutions obtained by the two compared methods based on relaxation and rounding techniques are not necessarily optimal solutions or even feasible solutions, but the virtual network mapping method based on particle swarm optimization can obtain approximately global optimal solutions, thereby remarkably reducing the virtual network mapping overhead and providing possibility for an underlying network to accept more virtual networks.
As shown in fig. 5, compared with the D-ViNE-SP virtual network mapping method and the D-ViNE-LB virtual network mapping method, the virtual network mapping method based on particle swarm optimization significantly improves the long-term average profit-to-cost ratio of the underlying network, which is shown in fig. 5 to be improved by about 20% and about 9%, respectively.
Fig. 6 shows that the virtual network mapping method based on particle swarm optimization reduces the required time of virtual network mapping by about 45% compared with the D-ViNE-SP virtual network mapping method, and reduces the virtual network mapping time by about 53% compared with the D-ViNE-LB virtual network mapping method. The virtual network mapping method based on particle swarm optimization can effectively balance the running time and the solving quality by setting the number of particle swarms, the iteration times and other iteration termination conditions, so that the utilization efficiency of underlying network resources can be improved.
The invention also provides a virtual network mapping system based on particle swarm optimization, which comprises the following components: an initialization unit 71, a feasibility verification unit 72, an update unit 73, and an output unit 74; wherein,
the initialization unit 71 is configured to initialize the position parameter and the velocity parameter of the particle;
the feasibility checking unit 72 is configured to perform feasibility checking on the initialized particles, and determine an initial global optimal position and an individual optimal position of each particle; performing feasibility test on the particles updated by the updating unit 73 to obtain feasible particles, re-determining the global optimal position and the individual optimal position of each particle, and adding 1 to the current iteration number;
an updating unit 73, configured to update the speed and the position of the feasible particle obtained by the feasibility checking unit 72 when the current iteration number is smaller than a preset maximum iteration number;
and an output unit 74, configured to output the global optimal position finally determined by the feasibility checking unit 72 as a mapping scheme when the current iteration number is equal to a preset maximum iteration number.
The initialization unit 71 is further configured to reinitialize the position parameters and the velocity parameters of the particles other than the feasible particles.
The output unit 74 is further configured to output a fitness value corresponding to the finally determined global optimal position, so as to obtain a bandwidth overhead corresponding to the mapping scheme.
Wherein the initialization unit 71 is specifically configured to determine the position parameter X of the particleiInitialized to the ith virtual network mapping scheme, speed parameter ViAn adjustment decision for the virtual mapping scheme is initialized.
The feasibility checking unit 72 is specifically configured to detect a capability constraint of an underlying network node to which a position parameter of a particle relates; and when the CPU capacity of the underlying network node meets the CPU capacity of the virtual network node and is positioned in the range of the position requested by the virtual network node, detecting the bandwidth and connectivity constraint conditions of the position parameter by adopting a shortest path algorithm to obtain the fitness of the particle.
The feasibility checking unit 72 is specifically configured to determine a position of a particle with the minimum fitness among all the particles as an initial global optimal position, and determine a position of a particle corresponding to the minimum fitness among the obtained multiple fitness of each particle as an initial individual optimal position of the particle.
The feasibility checking unit 72 is specifically configured to calculate the fitness of each particle updated by the updating unit, and when the obtained fitness is smaller than the fitness corresponding to the previously determined individual optimal position, re-determine the individual optimal position as the position corresponding to the current fitness; and when the fitness corresponding to the redetermined individual optimal position is smaller than the fitness corresponding to the previously determined global optimal position, redetermining the global optimal position as the redetermined individual optimal position.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.
Claims (10)
1. A virtual network mapping method based on particle swarm optimization is characterized by comprising the following steps:
initializing the position parameters and the speed parameters of the particles;
carrying out feasibility test on the particles to obtain feasible particles, and determining an initial global optimal position and an individual optimal position of each particle;
updating the speed and the position of the feasible particles, carrying out feasibility inspection on the updated particles, obtaining the feasible particles again, re-determining the global optimal position and the individual optimal position of each particle, adding 1 to the current iteration number, and repeating the step when the iteration number is less than the preset maximum iteration number;
and when the iteration times are equal to the preset maximum iteration times, outputting the finally determined global optimal position as a mapping scheme.
2. The method of claim 1, wherein the initializing the position parameters and the velocity parameters of the particles is to:
the position parameter Xi of the particle is initialized to the ith virtual network mapping scheme, and the speed parameter Vi is initialized to the adjustment decision of the virtual mapping scheme.
3. The method of claim 1, wherein the feasibility test is performed on the particles to obtain feasible particles as follows:
detecting the capability constraint of the underlying network nodes related to the position parameters of the particles;
and when the CPU capacity of the underlying network node meets the CPU capacity of the virtual network node and is positioned in the range of the position requested by the virtual network node, detecting the bandwidth and connectivity constraint conditions of the position parameter by adopting a shortest path algorithm to obtain the fitness of the particle.
4. The method of claim 3, wherein determining the initial global optimal position and the individual optimal position for each particle is:
and determining the position of the particle with the minimum fitness in all the particles as an initial global optimal position, and determining the position of the particle corresponding to the minimum fitness in the obtained multiple fitness of each particle as an initial individual optimal position of the particle.
5. The method of claim 1, wherein after obtaining viable particles, the method further comprises:
and re-initializing the position parameters and the speed parameters of the particles except the feasible particles.
6. The method of claim 1, wherein the re-determining the global optimal position and the individual optimal position for each particle is:
calculating the fitness of each particle, and when the obtained fitness is smaller than the fitness corresponding to the previously determined individual optimal position, re-determining the individual optimal position as the position corresponding to the current fitness;
and when the fitness corresponding to the redetermined individual optimal position is smaller than the fitness corresponding to the previously determined global optimal position, redetermining the global optimal position as the redetermined individual optimal position.
7. The method of claim 1, further comprising: and outputting the finally determined fitness value corresponding to the global optimal position to obtain the bandwidth overhead corresponding to the mapping scheme.
8. A particle swarm optimization-based virtual network mapping system, the system comprising: the device comprises an initialization unit, a feasibility checking unit, an updating unit and an output unit; wherein,
the initialization unit is used for initializing the position parameter and the speed parameter of the particles;
the feasibility testing unit is used for carrying out feasibility testing on the initialized particles and determining an initial global optimal position and an individual optimal position of each particle; carrying out feasibility test on the particles updated by the updating unit to obtain feasible particles, re-determining the global optimal position and the individual optimal position of each particle, and adding 1 to the current iteration number;
the updating unit is used for updating the speed and the position of the feasible particles obtained by the feasibility checking unit when the current iteration number is smaller than the preset maximum iteration number;
and the output unit is used for outputting the global optimal position finally determined by the feasibility testing unit as a mapping scheme when the current iteration number is equal to the preset maximum iteration number.
9. The system of claim 8, wherein the initialization unit is further configured to re-initialize the position parameters and the velocity parameters of the particles other than the feasible particles.
10. The system of claim 8, wherein the output unit is further configured to output a fitness value corresponding to the finally determined global optimal position, so as to obtain a bandwidth overhead corresponding to the mapping scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103640769A CN102364899A (en) | 2011-11-16 | 2011-11-16 | Particle-swam-optimization-based virtual network mapping method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103640769A CN102364899A (en) | 2011-11-16 | 2011-11-16 | Particle-swam-optimization-based virtual network mapping method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102364899A true CN102364899A (en) | 2012-02-29 |
Family
ID=45691451
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011103640769A Pending CN102364899A (en) | 2011-11-16 | 2011-11-16 | Particle-swam-optimization-based virtual network mapping method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102364899A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102917441A (en) * | 2012-09-29 | 2013-02-06 | 北京邮电大学 | Target network selection method on basis of particle swarm algorithm for multi-mode terminals |
| CN103679564A (en) * | 2013-12-26 | 2014-03-26 | 国家电网公司 | Task allocation method applicable to power distribution network topology analysis distributed computation |
| CN103782556A (en) * | 2013-08-08 | 2014-05-07 | 华为技术有限公司 | Distributing method, controller, device, and system for virtual network |
| CN103841000A (en) * | 2014-03-12 | 2014-06-04 | 湖州师范学院 | Virtual link mapping algorithm based on minimum cost flow model |
| CN103856385A (en) * | 2013-12-11 | 2014-06-11 | 北京邮电大学 | Virtual network mapping method based on link priority |
| WO2014202016A1 (en) * | 2013-06-20 | 2014-12-24 | 中国科学院声学研究所 | Classification-based virtual network mapping method and system |
| CN104468308A (en) * | 2014-10-29 | 2015-03-25 | 北京邮电大学 | Cross-domain virtual network mapping method and system based on particle swarm optimization |
| CN104955056A (en) * | 2015-06-05 | 2015-09-30 | 大连理工大学 | Internet-of-vehicle road side unit deployment method based on particle swarm optimization |
| CN104994021A (en) * | 2015-07-21 | 2015-10-21 | 三星电子（中国）研发中心 | Method and device for determining optimal path |
| CN105610676A (en) * | 2016-02-18 | 2016-05-25 | 北京邮电大学 | Virtual mapping method and device for software-defined network |
| CN106130861A (en) * | 2016-06-15 | 2016-11-16 | 北京工业大学 | A kind of mapping method of virtual network of multi-objective particle swarm optimization based on microhabitat |
| CN106301924A (en) * | 2016-08-18 | 2017-01-04 | 北京邮电大学 | A kind of mapping method of virtual network and device |
| CN107135135A (en) * | 2017-05-02 | 2017-09-05 | 北京邮电大学 | A kind of mapping method of virtual network and device sorted based on bottom layer node |
| CN108337697A (en) * | 2018-02-12 | 2018-07-27 | 重庆邮电大学 | Wireless network virtualizes the Delay Mapping method of interior joint and link |
| CN108965020A (en) * | 2018-07-27 | 2018-12-07 | 北京邮电大学 | Cross-domain virtual network mapping method and device thereof, computer-readable medium |
| CN109347657A (en) * | 2018-09-12 | 2019-02-15 | 石家庄铁道大学 | The virtual data domain construction method of scientific and technological business is supported under SDN mode |
| WO2019052076A1 (en) * | 2017-09-18 | 2019-03-21 | 中国科学院声学研究所 | Method for generating nested container with no intersection and full coverage in the same layer and readable storage medium |
| CN112272120A (en) * | 2020-12-23 | 2021-01-26 | 江苏移动信息系统集成有限公司 | Optimization method and device of virtual network mapping algorithm |
| CN112953781A (en) * | 2021-03-31 | 2021-06-11 | 广东电网有限责任公司电力调度控制中心 | Particle swarm-based virtual service fault recovery method and device under network slice |
| CN117688823A (en) * | 2024-02-04 | 2024-03-12 | 北京航空航天大学 | Rock-soil particle track prediction method, electronic equipment and medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383736A (en) * | 2008-10-15 | 2009-03-11 | 中国科学院上海微系统与信息技术研究所 | Optimizing method for wireless sensor network node laying oriented to area monitoring |
| CN102129482A (en) * | 2010-01-13 | 2011-07-20 | 电子科技大学 | Chaotic discrete particle swarm optimization-based network on chip mapping method |
-
2011
- 2011-11-16 CN CN2011103640769A patent/CN102364899A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383736A (en) * | 2008-10-15 | 2009-03-11 | 中国科学院上海微系统与信息技术研究所 | Optimizing method for wireless sensor network node laying oriented to area monitoring |
| CN102129482A (en) * | 2010-01-13 | 2011-07-20 | 电子科技大学 | Chaotic discrete particle swarm optimization-based network on chip mapping method |
Non-Patent Citations (1)
| Title |
|---|
| 程祥等: "基于粒子群优化的虚拟网络映射算法", 《电子学报》 * |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102917441B (en) * | 2012-09-29 | 2015-06-24 | 北京邮电大学 | Target network selection method on basis of particle swarm algorithm for multi-mode terminals |
| CN102917441A (en) * | 2012-09-29 | 2013-02-06 | 北京邮电大学 | Target network selection method on basis of particle swarm algorithm for multi-mode terminals |
| WO2014202016A1 (en) * | 2013-06-20 | 2014-12-24 | 中国科学院声学研究所 | Classification-based virtual network mapping method and system |
| CN103782556A (en) * | 2013-08-08 | 2014-05-07 | 华为技术有限公司 | Distributing method, controller, device, and system for virtual network |
| CN103782556B (en) * | 2013-08-08 | 2016-12-07 | 华为技术有限公司 | The virtualization distribution method of network, controller, equipment and system |
| CN103856385B (en) * | 2013-12-11 | 2017-02-08 | 北京邮电大学 | Virtual network mapping method based on link priority |
| CN103856385A (en) * | 2013-12-11 | 2014-06-11 | 北京邮电大学 | Virtual network mapping method based on link priority |
| CN103679564A (en) * | 2013-12-26 | 2014-03-26 | 国家电网公司 | Task allocation method applicable to power distribution network topology analysis distributed computation |
| CN103679564B (en) * | 2013-12-26 | 2017-01-25 | 国家电网公司 | Task allocation method applicable to power distribution network topology analysis distributed computation |
| CN103841000A (en) * | 2014-03-12 | 2014-06-04 | 湖州师范学院 | Virtual link mapping algorithm based on minimum cost flow model |
| CN103841000B (en) * | 2014-03-12 | 2017-05-24 | 湖州师范学院 | Virtual link mapping algorithm based on minimum cost flow model |
| CN104468308A (en) * | 2014-10-29 | 2015-03-25 | 北京邮电大学 | Cross-domain virtual network mapping method and system based on particle swarm optimization |
| CN104468308B (en) * | 2014-10-29 | 2018-06-22 | 北京邮电大学 | Cross-domain virtual network network mapping method and system based on particle cluster algorithm |
| CN104955056A (en) * | 2015-06-05 | 2015-09-30 | 大连理工大学 | Internet-of-vehicle road side unit deployment method based on particle swarm optimization |
| CN104994021B (en) * | 2015-07-21 | 2018-06-15 | 三星电子（中国）研发中心 | Determine the method and device of optimal path |
| CN104994021A (en) * | 2015-07-21 | 2015-10-21 | 三星电子（中国）研发中心 | Method and device for determining optimal path |
| CN105610676B (en) * | 2016-02-18 | 2019-03-22 | 北京邮电大学 | A kind of software defined network virtual map method and device |
| CN105610676A (en) * | 2016-02-18 | 2016-05-25 | 北京邮电大学 | Virtual mapping method and device for software-defined network |
| CN106130861A (en) * | 2016-06-15 | 2016-11-16 | 北京工业大学 | A kind of mapping method of virtual network of multi-objective particle swarm optimization based on microhabitat |
| CN106130861B (en) * | 2016-06-15 | 2019-09-27 | 北京工业大学 | A kind of mapping method of virtual network of the multi-objective particle swarm optimization based on microhabitat |
| CN106301924A (en) * | 2016-08-18 | 2017-01-04 | 北京邮电大学 | A kind of mapping method of virtual network and device |
| CN107135135A (en) * | 2017-05-02 | 2017-09-05 | 北京邮电大学 | A kind of mapping method of virtual network and device sorted based on bottom layer node |
| CN107135135B (en) * | 2017-05-02 | 2019-09-06 | 北京邮电大学 | A kind of mapping method of virtual network and device based on bottom layer node sequence |
| US11231954B2 (en) | 2017-09-18 | 2022-01-25 | Institute Of Acoustics, Chinese Academy Of Sciences | Method for generating nested container with no intersection and full coverage in the same layer and readable storage medium |
| WO2019052076A1 (en) * | 2017-09-18 | 2019-03-21 | 中国科学院声学研究所 | Method for generating nested container with no intersection and full coverage in the same layer and readable storage medium |
| JP2020533915A (en) * | 2017-09-18 | 2020-11-19 | 中国科学院声学研究所Ｉｎｓｔｉｔｕｔｅ Ｏｆ Ａｃｏｕｓｔｉｃｓ， Ｃｈｉｎｅｓｅ Ａｃａｄｅｍｙ Ｏｆ Ｓｃｉｅｎｃｅｓ | Nested container generation method and readable storage medium |
| CN108337697B (en) * | 2018-02-12 | 2021-06-18 | 重庆邮电大学 | Delay mapping method of nodes and links in wireless network virtualization |
| CN108337697A (en) * | 2018-02-12 | 2018-07-27 | 重庆邮电大学 | Wireless network virtualizes the Delay Mapping method of interior joint and link |
| CN108965020A (en) * | 2018-07-27 | 2018-12-07 | 北京邮电大学 | Cross-domain virtual network mapping method and device thereof, computer-readable medium |
| CN109347657A (en) * | 2018-09-12 | 2019-02-15 | 石家庄铁道大学 | The virtual data domain construction method of scientific and technological business is supported under SDN mode |
| CN112272120B (en) * | 2020-12-23 | 2021-05-11 | 江苏移动信息系统集成有限公司 | Virtual network mapping algorithm optimization method, equipment and storage medium |
| CN112272120A (en) * | 2020-12-23 | 2021-01-26 | 江苏移动信息系统集成有限公司 | Optimization method and device of virtual network mapping algorithm |
| CN112953781A (en) * | 2021-03-31 | 2021-06-11 | 广东电网有限责任公司电力调度控制中心 | Particle swarm-based virtual service fault recovery method and device under network slice |
| CN117688823A (en) * | 2024-02-04 | 2024-03-12 | 北京航空航天大学 | Rock-soil particle track prediction method, electronic equipment and medium |
| CN117688823B (en) * | 2024-02-04 | 2024-05-14 | 北京航空航天大学 | Rock-soil particle track prediction method, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102364899A (en) | Particle-swam-optimization-based virtual network mapping method and system | |
| Abidi et al. | Optimal 5G network slicing using machine learning and deep learning concepts | |
| Pandurangan et al. | On the distributed complexity of large-scale graph computations | |
| CN108566659B (en) | 5G network slice online mapping method based on reliability | |
| CN113708972B (en) | Service function chain deployment method and device, electronic equipment and storage medium | |
| CN110134493B (en) | Service function chain deployment algorithm based on resource fragment avoidance | |
| CA3080050A1 (en) | Training tree-based machine-learning modeling algorithms for predicting outputs and generating explanatory data | |
| CN105897584B (en) | Paths planning method and controller | |
| WO2018176385A1 (en) | System and method for network slicing for service-oriented networks | |
| CN110348571A (en) | A kind of neural network model training method, device, chip and system | |
| CN104104621B (en) | A kind of virtual network resource dynamic self-adapting adjusting method based on Nonlinear Dimension Reduction | |
| CN106961343A (en) | A kind of virtual map method and device | |
| Wang et al. | A reinforcement learning approach for online service tree placement in edge computing | |
| CN114050975B (en) | Heterogeneous multi-node interconnection topology generation method and storage medium | |
| Mai et al. | On the use of supervised machine learning for assessing schedulability: application to Ethernet TSN | |
| Zhou et al. | JPAS: Job-progress-aware flow scheduling for deep learning clusters | |
| CN110933000B (en) | Distributed data multi-stage aggregation method, device, server and storage medium | |
| Ometov et al. | On applicability of imagery-based CNN to computational offloading location selection | |
| Ling et al. | QoS and Fairness Oriented Dynamic Computation Offloading in the Internet of Vehicles based on Estimate Time of Arrival | |
| Kusetogullari et al. | A reduced uncertainty-based hybrid evolutionary algorithm for solving dynamic shortest-path routing problem | |
| CN113422726A (en) | Service chain deployment method and device, storage medium and electronic device | |
| Xia et al. | Learn to optimize: Adaptive VNF provisioning in mobile edge clouds | |
| CN104348695A (en) | Artificial immune system-based virtual network mapping method and system thereof | |
| Li et al. | Latency-aware batch task offloading for vehicular cloud: Maximizing submodular bandit | |
| Gamal et al. | Mapping and scheduling for non-uniform arrival of virtual network function (VNF) requests |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120229 |