CN105071963A - Multi-service traffic estimation method facing big data Internet - Google Patents
Multi-service traffic estimation method facing big data Internet Download PDFInfo
- Publication number
- CN105071963A CN105071963A CN201510474547.XA CN201510474547A CN105071963A CN 105071963 A CN105071963 A CN 105071963A CN 201510474547 A CN201510474547 A CN 201510474547A CN 105071963 A CN105071963 A CN 105071963A
- Authority
- CN
- China
- Prior art keywords
- matrix
- network
- network management
- management station
- flow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/02—Standardisation; Integration
- H04L41/0213—Standardised network management protocols, e.g. simple network management protocol [SNMP]
-
- 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
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
- H04L43/0888—Throughput
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 invention provides a multi-service traffic estimation method facing a big data Internet, and belongs to the technical field of communication networks. The method comprises the following steps: step 1 of acquiring a backbone network link load by a network management station by adopting a simple network management protocol; step 2 of acquiring a route matrix by the network management station according to the routing table information in a network topology and a router; step 3 of generating into one random Bernoulli matrix by the network management station, and determining partial directly measured end-to-end network traffics according to the matrix; step 4 of approximately describing a traffic matrix by using a principal component analysis method; step 5 of constructing a network tomography model according to the steps 1, 2 and 4; step 6 of constructing a linear measurement model according to the steps 3 and 4; step 7 of estimating the traffic matrix by the network management station through the method of constructing an optimizing model according to the models in steps 5 and 6. According to the multi-service traffic estimation method facing the big data Internet provided by the invention, the linear measurement model is constructed through the Bernoulli matrix and a traffic estimation model is construction by using a network tomography technology.
Description
Technical field
The invention belongs to technical field of communication network, particularly a kind of multi-business flow amount estimation method towards large data interconnection net.
Background technology
Along with the development of information and communication technology (ICT), Internet technology has been deep in society life, industrial production.The particularly proposition of " the Internet+", advances the popularization of Internet technology in the field such as industrial, agriculture further, facilitates social industry's IT application process.While Internet technology is deep into social every field gradually, the problem of the aspect such as mass data processing, information security becomes increasingly conspicuous, and these problems are had higher requirement to network management.
In recent years, the Internet was that different terminal uses provides diversified network service, and the Internet has become a complicated heterogeneous network in this context.Network Management Function is introduced in network to ensure the service quality of each user.For a network manager, first it need the state information understanding end to end network flow when an execution effective Network Management Function.In the middle of reality, end to end network flow information can be described by traffic matrix, and traffic matrix is the important input parameter of network management procedure one.
Although traffic matrix has extremely important effect, for a large scale backbone network, acquisition traffic matrix is also remarkable.This reason is multiple, and first, for a large scale backbone network, direct collection network flow information is not attainable.In this case, researcher is more prone to estimation network flow and non-immediate collection network flow information indirectly.In numerous network flow estimating methods, normally go to infer network traffics by other effective network informations, such as network tomography technology removes estimation network flow by link load and routing iinformation.But network tomography model has the Ill-posed characteristic of height, therefore estimation network stream quantifier elimination is appointed and is so faced lot of challenges.
Summary of the invention
For the deficiency that existing method exists, the present invention proposes a kind of multi-business flow amount estimation method towards large data interconnection net, obtains backbone traffic estimated value accurately with this, for realizing effective network management and solid foundation is established in the network planning.
Towards a multi-business flow amount estimation method for large data interconnection net, comprise the following steps:
Step 1: Network Management Station adopts Simple Network Management Protocol to obtain backbone link load;
Step 2: Network Management Station is according to routing table acquisition of information route matrix in network topology structure and router;
Step 3: Network Management Station generates random Bernoulli Jacob's matrix, and according to the end to end network flow that this matrix determining section is directly measured;
Step 4: utilize principal component analytical method to describe traffic matrix approx;
Step 5: build network tomography model according to step 1,2 and 4;
Step 6: build linear measurement model according to step 3 and 4;
Step 7: according to the model in step 5 and 6, Network Management Station is by building the method estimated flow matrix of optimal model.
Network Management Station described in step 1 adopts Simple Network Management Protocol to obtain backbone link load, and method is: when with matrix
lwhen representing link load, one is had
nindividual node and
pthe backbone network of bar link, intercepts
tthe link load data of individual time slot, then
lit is one
p×
tmatrix.
Network Management Station described in step 2 is according to routing table acquisition of information route matrix in network topology structure and router, and method is: when with
rwhen representing route matrix, one is had
nindividual node and
pthe backbone network of bar link, route matrix
rit is one
p×
n 2matrix.
Network Management Station described in step 3 generates random Bernoulli Jacob's matrix, and according to the end to end network flow that this matrix determining section is directly measured, concrete steps are as follows:
Step 3-1: use matrix
mrepresent a traffic matrix, then traffic matrix is one
n 2×
tmatrix, each end to end network flow obtains a mark, be respectively 1 to
n 2;
Step 3-2: Network Management Station generates one
q×
n 2random Bernoulli Jacob's matrix, use symbol
brepresent, its element be respectively 1 or 0.Matrix
bin each element independent same distribution, and
b q,n ~ Bern (
z), wherein
q=1,2 ...,
q,
n=1,2 ...,
n 2,
zfor working as
b q,n probability when=1;
Step 3-3: to Bernoulli Jacob's matrix
beach row on element get union, namely
,
Wherein,
w n for
bin the result of calculation of union of element of each row;
Step 3-4: when
w n value when equaling 1, then Network Management Station is by controlling the method for router port, utilizes router NetFlow functional measurement to be numbered
nend to end network flow;
Step 3-5: the end to end network flow directly measured is put into traffic matrix
min, then traffic matrix
min contain by directly measuring and known end to end network flow and not measured unknown flow rate;
Step 3-6: obtain a linear system according to step 3-5, namely
,
Wherein, matrix
ybe called measured value, from linear relationship above,
yonly and traffic matrix
min known end to end network flow relevant, and to have nothing to do with unknown flow rate, therefore measured value
yfor being known us, by the end to end network flow directly measured and Bernoulli Jacob's matrix
bbe multiplied and calculate.
The principal component analytical method that utilizes described in step 4 describes traffic matrix approx, and concrete steps are as follows:
Step 4-1: Network Management Station collects historical traffic matrix, and is expressed as
m *;
Step 4-2: utilize the method for singular value decomposition to decompose historical traffic matrix
m *transposition, namely
,
Wherein, Σ
*for diagonal matrix, the value on its diagonal element is matrix
m * Tsingular value.
v *be an orthogonal matrix, matrix
u *describe historical traffic matrix dynamic-change information;
Step 4-3: utilize principal component Description Matrix M approx
*
t , namely
,
Wherein, diagonal matrix Σ
* kfor extracting
ksingular value matrix after individual principal component, this step is equivalent to retain
kindividual maximum singular value, and all the other little singular values are set to 0;
Step 4-4: use historical traffic matrix
m *principal component approx traffic matrix is described
m, namely
,
Wherein, matrix
u tdescribe traffic matrix
mdynamic variation characteristic.
Build network tomography models according to step 1,2 and 4 described in step 5, method is: network tomography model representation is
。
Build linear measurement models according to step 3 and 4 described in step 6, method is: the model representation of this structure is:
。
Described in step 7 according to the model in step 5 and 6, Network Management Station is by building the method estimated flow matrix of optimal model, and concrete steps are as follows:
Step 7-1: the model construction Optimized model in step 5 and 6, as follows
;
Step 7-2: obtain matrix by the method for the optimal model in solution procedure 7-1
u t estimated value, be expressed as Es{
u t , then traffic matrix estimated value Es{
m}=
v *Σ
* k es{
u t .
Advantage of the present invention: a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention, adopts compressed sensing technology and network tomography technology, on the one hand, overcomes the deficient of network tomography technology and determines characteristic; On the other hand, the great expense incurred problem directly measuring all end to end network flows is avoided.The inventive method obtains good traffic matrix estimated value, and in evaluated error, had obvious improvement, evaluated error reduces 16%.
Accompanying drawing explanation
Fig. 1 is the network topology structure of a kind of multi-business flow amount estimation method specific embodiment towards large data interconnection net of the present invention;
Fig. 2 is a kind of multi-business flow amount estimation method structural representation towards large data interconnection net of the present invention;
Fig. 3 is a kind of multi-business flow amount estimation method flow chart towards large data interconnection net of the present invention;
Fig. 4 to be a kind of multi-business flow amount estimation method label towards large data interconnection net of the present invention be 99 end to end network flow actual value and estimated value contrast schematic diagram;
Fig. 5 to be a kind of multi-business flow amount estimation method label towards large data interconnection net of the present invention be 105 end to end network flow actual value and estimated value contrast schematic diagram;
Fig. 6 is the relative root-mean-square error schematic diagram of a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention;
Fig. 7 is the cumulative distribution function schematic diagram of the relative root-mean-square error of a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in further detail.
The present embodiment is estimated Abilene backbone network data, and Abilene backbone network topological structure as shown in Figure 1, it comprises 12 nodes and 54 one way links (comprising 24 peripheral links and 30 inner link).Therefore traffic matrix
min the number of end-to-end network traffics be
n 2=12
2=144, the quantity of link load
p=54.For Bernoulli Jacob's matrix
b, its line number
q=144, and
z=0.01.In principal component analysis process, get principal component quantity
k=6.
A kind of multi-business flow amount estimation method towards large data interconnection net of the present embodiment, concrete steps as shown in Figure 2.
Step 1: Network Management Station adopts Simple Network Management Protocol to obtain backbone link load;
When with matrix
lwhen representing link load, one is had to the backbone network of 12 nodes and 54 links, intercept the link load data of 1516 time slots, then
lit is the matrix of 54 × 1516.
Step 2: Network Management Station is according to routing table acquisition of information route matrix in network topology structure and router;
When with
rwhen representing route matrix, one is had to the backbone network of 12 nodes and 54 links, route matrix
rit is the matrix of 54 × 144.
Step 3: Network Management Station generates random Bernoulli Jacob's matrix, and according to the end to end network flow that this matrix determining section is directly measured, concrete steps are as follows:
Step 3-1: use matrix
mrepresent a traffic matrix, then traffic matrix is the matrix of 144 × 1516, and each end to end network flow obtains an identification code, is respectively 1 to 144;
Step 3-2: Network Management Station generates 144 × 144 random Bernoulli Jacob's matrixes, uses symbol
brepresent, its element be respectively 1 or 0.Matrix
bin each element independent same distribution, and
b q,n ~ Bern (0.01), wherein
q=1,2 ..., 144,
n=1,2 ..., 144,
zfor working as
b q,n probability when=1;
Step 3-3: to Bernoulli Jacob's matrix
beach row on element get union, namely
,
Wherein,
w n for
bin the result of calculation of union of element of each row;
Step 3-4: when
w n value when equaling 1, then Network Management Station is by controlling the method for router port, utilizes router NetFlow functional measurement to be numbered
nend to end network flow;
Step 3-5: the end to end network flow directly measured is put into traffic matrix
min, then traffic matrix
min contain by directly measuring and known end to end network flow and not measured unknown flow rate;
Step 3-6: obtain a linear system according to step 3-5, namely
,
Wherein, matrix
ybe called measured value, from linear relationship above,
yonly and traffic matrix
min known end to end network flow relevant, and to have nothing to do with unknown flow rate, therefore measured value
yfor being known us, by the end to end network flow directly measured and Bernoulli Jacob's matrix
bbe multiplied and calculate.
Step 4: utilize principal component analytical method to describe traffic matrix approx, concrete steps are as follows:
Step 4-1: Network Management Station collects historical traffic matrix, and is expressed as
m *;
Step 4-2: utilize the method for singular value decomposition to decompose historical traffic matrix
m *transposition, namely
,
Wherein, Σ
*for diagonal matrix, the value on its diagonal element is matrix
m *
t singular value.
v *be an orthogonal matrix, matrix
u *describe historical traffic matrix dynamic-change information;
Step 4-3: utilize principal component Description Matrix approx
m *
t , namely
,
Wherein, diagonal matrix Σ
* k for extracting the singular value matrix after 10 principal components, this step is equivalent to the maximum singular value of reservation 10, and all the other little singular values are set to 0;
Step 4-4: use historical traffic matrix
m *principal component approx traffic matrix is described
m, namely
,
Wherein, matrix
u t describe traffic matrix
mdynamic variation characteristic.
Step 5: build network tomography model according to step 1,2 and 4; Network tomography model representation is
。
Step 6: build linear measurement model according to step 3 and 4; The model representation of this structure is
。
Step 7: according to the model in step 5 and 6, Network Management Station is by building the method estimated flow matrix of optimal model, and concrete steps are as follows:
Step 7-1: the model construction Optimized model in step 5 and 6, as follows
;
Step 7-2: obtain matrix by the method for the optimal model in solution procedure 7-1
u t estimated value, be expressed as Es{
u t , then traffic matrix estimated value Es{
m}=
v *Σ
* k es{
u t .
For verifying validity of the present invention, first we analyze the ability that a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention follows the tracks of end to end network changes in flow rate trend.Simultaneously, in specific embodiment, a kind of for the present invention multi-business flow amount estimation method towards large data interconnection net and three classical traffic matrix methods of estimation are compared, three kinds of methods are respectively sparse regularization singular value decomposition method (SparsityRegularizedSingularValueDecomposition, SRSVD), principal component analytical method (PrincipalComponentAnalysis, PCA) and Gravity Models method (Tomogravity).
Select two end to end network flows in this embodiment and compare their actual value and estimated value.Fig. 4 depicts end to end network flow actual value and the estimated value that label is 99.Can be found out by simulation result, all methods can follow the tracks of the variation tendency of No. 99 end to end network flows.However, SRSVD is by larger evaluated error at 1200 time slots.Tomogravity method, when estimating this end to end network flow, is estimated to have occurred deficient estimation in the period whole.PCA method is maximum at all method medial error.Following analysis label is the end to end network flow of 105, and as shown in Figure 5, four kinds of methods all obtain comparatively ideal estimated result.The shake of usual tracking flow is very difficult.However, four kinds of methods all can follow the tracks of the variations in detail of this end to end network flow, such as, shake between time slot 770-800.
In this embodiment, in order to describe the evaluated error of a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention quantitatively, introduce relative root-mean-square error (RelativeRootMeanSquaredError, RRMSE) and compare four kinds of methods as a tolerance.RRMSE is defined as
,
Wherein,
m n (
t) for label be
nend to end network flow at time slot
ttime flow value.Es{
m n (
t) be
m n (
t) estimated value.Fig. 6 gives the relative root-mean-square error of four kinds of methods.Therefrom can find out, the RRMSE of a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention is starkly lower than SRSVD method and Tomogravity method.The RRMSE of PCA method has larger fluctuation.The mean value of four kinds of method RRMSE is respectively 0.23,0.29,0.26 and 0.27.Be depicted as the cumulative distribution function of four kinds of method RRMSE in the figure 7.Fig. 7 illustrates in most of the cases, and the RRMSE of a kind of multi-business flow amount estimation method towards large data interconnection net of the present invention is minimum.In other words, herein carry algorithm RRMSE there is no large fluctuation.
Claims (4)
1., towards a multi-business flow amount estimation method for large data interconnection net, it is characterized in that: comprise the steps:
Step 1: Network Management Station adopts Simple Network Management Protocol to obtain backbone link load;
Step 2: Network Management Station is according to routing table acquisition of information route matrix in network topology structure and router;
Step 3: Network Management Station generates random Bernoulli Jacob's matrix, and according to the end to end network flow that this matrix determining section is directly measured;
Step 4: utilize principal component analytical method to describe traffic matrix approx;
Step 5: build network tomography model according to step 1,2 and 4;
Step 6: build linear measurement model according to step 3 and 4;
Step 7: according to the model in step 5 and 6, Network Management Station is by building the method estimated flow matrix of optimal model.
2. a kind of multi-business flow amount estimation method towards large data interconnection net according to claim 1, it is characterized in that: the Network Management Station described in step 3 generates random Bernoulli Jacob's matrix, and according to the end to end network flow that this matrix determining section is directly measured, specifically comprise the steps:
Step 3-1: use matrix
mrepresent a traffic matrix, then traffic matrix is one
n 2×
tmatrix, each end to end network flow obtains a mark, be respectively 1 to
n 2;
Step 3-2: Network Management Station generates one
q×
n 2random Bernoulli Jacob's matrix, use symbol
brepresent, its element be respectively 1 or 0, matrix
bin each element independent same distribution, and
b q,n ~ Bern (
z), wherein
q=1,2 ...,
q,
n=1,2 ...,
n 2,
zfor working as
b q,n probability when=1;
Step 3-3: to Bernoulli Jacob's matrix
beach row on element get union, namely
,
Wherein,
w n for
bin the result of calculation of union of element of each row;
Step 3-4: when
w n value when equaling 1, then Network Management Station is by controlling the method for router port, utilizes router NetFlow functional measurement to be numbered
nend to end network flow;
Step 3-5: the end to end network flow directly measured is put into traffic matrix
min, then traffic matrix
min contain by directly measuring and known end to end network flow and not measured unknown flow rate;
Step 3-6: obtain a linear system according to step 3-5, namely
,
Wherein, matrix
ybe called measured value, from linear relationship above,
yonly and traffic matrix
min known end to end network flow relevant, and to have nothing to do with unknown flow rate, therefore measured value
yfor being known us, by the end to end network flow directly measured and Bernoulli Jacob's matrix
bbe multiplied and calculate.
3. a kind of multi-business flow amount estimation method towards large data interconnection net according to claim 1, is characterized in that: build linear measurement models according to step 3 and 4 described in step 6, and concrete steps are as follows:
The model representation of this structure is
。
4. a kind of multi-business flow amount estimation method towards large data interconnection net according to claim 1, it is characterized in that, described in step 7 according to the model in step 5 and 6, Network Management Station is by building the method estimated flow matrix of optimal model, and concrete steps are as follows:
Step 7-1: the model construction Optimized model in step 5 and 6, as follows
;
Step 7-2: obtain matrix by the method for the optimal model in solution procedure 7-1
u t estimated value, be expressed as Es{
u t , then traffic matrix estimated value Es{
m}=
v *Σ
* k es{
u t .
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510474547.XA CN105071963A (en) | 2015-08-06 | 2015-08-06 | Multi-service traffic estimation method facing big data Internet |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510474547.XA CN105071963A (en) | 2015-08-06 | 2015-08-06 | Multi-service traffic estimation method facing big data Internet |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105071963A true CN105071963A (en) | 2015-11-18 |
Family
ID=54501250
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510474547.XA Pending CN105071963A (en) | 2015-08-06 | 2015-08-06 | Multi-service traffic estimation method facing big data Internet |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105071963A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108075928A (en) * | 2017-12-15 | 2018-05-25 | 中盈优创资讯科技有限公司 | Network traffics Universal Simulation Model and method |
CN109088796A (en) * | 2018-09-19 | 2018-12-25 | 哈尔滨工业大学 | Network flow Matrix Prediction Method based on network tomography technology |
CN110635973A (en) * | 2019-11-08 | 2019-12-31 | 西北工业大学青岛研究院 | Backbone network flow determining method and system based on reinforcement learning |
CN110855485A (en) * | 2019-11-08 | 2020-02-28 | 西北工业大学青岛研究院 | Method and system for determining network flow of IP backbone network |
CN113379092A (en) * | 2020-03-09 | 2021-09-10 | 西北工业大学青岛研究院 | Backbone network multi-service traffic estimation method and system facing big data |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102801629A (en) * | 2012-08-22 | 2012-11-28 | 电子科技大学 | Traffic matrix estimation method |
EP2613480A1 (en) * | 2010-08-30 | 2013-07-10 | Nec Corporation | Communication quality monitoring system, communication quality monitoring method, and storage medium |
CN103607292A (en) * | 2013-10-28 | 2014-02-26 | 国家电网公司 | Fast distributed monitoring method for electric-power communication network services |
-
2015
- 2015-08-06 CN CN201510474547.XA patent/CN105071963A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2613480A1 (en) * | 2010-08-30 | 2013-07-10 | Nec Corporation | Communication quality monitoring system, communication quality monitoring method, and storage medium |
CN102801629A (en) * | 2012-08-22 | 2012-11-28 | 电子科技大学 | Traffic matrix estimation method |
CN103607292A (en) * | 2013-10-28 | 2014-02-26 | 国家电网公司 | Fast distributed monitoring method for electric-power communication network services |
Non-Patent Citations (2)
Title |
---|
LAISEN NIE: ""Traffic Matrix Estimation Approach Based on Partial Direct Measurements in Large-Scale IP Backbone Networks"", 《2015 IEEE 5TH INTERNATIONAL CONFERENCE ON ELECTRONICS INFORMATION AND EMERGENCY COMMUNICATION》 * |
LAISEN NIE等: ""A compressive sensing-based network tomography approach to estimating origin–destination flow traffic in large-scale backbone networks"", 《INTERNATIONAL JOURNAL OF COMMUNICATION SYSTEMS》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108075928A (en) * | 2017-12-15 | 2018-05-25 | 中盈优创资讯科技有限公司 | Network traffics Universal Simulation Model and method |
CN108075928B (en) * | 2017-12-15 | 2020-12-08 | 中盈优创资讯科技有限公司 | General simulation model and method for network flow |
CN109088796A (en) * | 2018-09-19 | 2018-12-25 | 哈尔滨工业大学 | Network flow Matrix Prediction Method based on network tomography technology |
CN110635973A (en) * | 2019-11-08 | 2019-12-31 | 西北工业大学青岛研究院 | Backbone network flow determining method and system based on reinforcement learning |
CN110855485A (en) * | 2019-11-08 | 2020-02-28 | 西北工业大学青岛研究院 | Method and system for determining network flow of IP backbone network |
CN110635973B (en) * | 2019-11-08 | 2022-07-12 | 西北工业大学青岛研究院 | Backbone network flow determining method and system based on reinforcement learning |
CN113379092A (en) * | 2020-03-09 | 2021-09-10 | 西北工业大学青岛研究院 | Backbone network multi-service traffic estimation method and system facing big data |
CN113379092B (en) * | 2020-03-09 | 2022-12-09 | 西北工业大学青岛研究院 | Backbone network multi-service traffic estimation method and system facing big data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105071963A (en) | Multi-service traffic estimation method facing big data Internet | |
Chong et al. | A simulation-based optimization algorithm for dynamic large-scale urban transportation problems | |
Tizghadam et al. | Betweenness centrality and resistance distance in communication networks | |
CN102724078B (en) | End-to-end network flow reconstruction method based on compression sensing in dynamic network | |
CN107181724A (en) | A kind of recognition methods for cooperateing with stream, system and the server using this method | |
CN113422695A (en) | Optimization method for improving robustness of topological structure of Internet of things | |
Qiao et al. | Efficient traffic matrix estimation for data center networks | |
CN109952743A (en) | System and method for low memory and low discharge expense big flow object detection | |
CN102546286A (en) | Method for on-line detection of shared network congestion paths | |
CN105024875A (en) | Method and system for measuring network flow rate of broadband classified users | |
JP6384125B2 (en) | Information processing apparatus and program | |
CN105678456B (en) | Method and system for automatically evaluating running state of electric energy metering device | |
CN113379092B (en) | Backbone network multi-service traffic estimation method and system facing big data | |
CN107528731A (en) | Network applied to NS3 parallel artificials splits optimized algorithm | |
CN105207905A (en) | Route selection method and device | |
Tornero et al. | A communication-aware topological mapping technique for NoCs | |
CN108322334A (en) | A kind of power communication system reliability estimation method | |
Niu et al. | Study on a new model for network traffic matrix estimation | |
Helm et al. | Flow-level Tail Latency Estimation and Verification based on Extreme Value Theory | |
Juva | Sensitivity of traffic matrix estimation techniques to their underlying assumptions | |
Shih et al. | Network topology discovery using finite mixture models | |
Nie et al. | A reconstructing approach to end‐to‐end network traffic based on multifractal wavelet model | |
CN105978714B (en) | The expansion planning method and expansion planning system of Internet data center | |
Karimov et al. | Models of network processes for describing operation of network protection tools | |
CN110232517A (en) | A kind of mobile intelligent perception user income choosing method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20151118 |