CN110177019A - A kind of control method slowing down network congestion - Google Patents
A kind of control method slowing down network congestion Download PDFInfo
- Publication number
- CN110177019A CN110177019A CN201910494944.1A CN201910494944A CN110177019A CN 110177019 A CN110177019 A CN 110177019A CN 201910494944 A CN201910494944 A CN 201910494944A CN 110177019 A CN110177019 A CN 110177019A
- Authority
- CN
- China
- Prior art keywords
- matrix
- network congestion
- node
- network
- indicate
- 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.)
- Granted
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
- H04L41/14—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
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a kind of control methods for slowing down network congestion.The present invention includes the following steps: step 1, establishes the state-space model of network congestion；Step 2, sytem matrix representation method is proposed；Step 3, design angular domain condition solves the problems, such as that system mode is constrained；Step 4, the state feedback control law of planned network blocking node.The present invention realizes effective control to network congestion phenomenon by technologies such as constraint control, non-linear control designs, reduces the space hold to buffer area, and the memory space resource of each buffer area of reasonable distribution, solves network congestion problem.
Description
Technical field
The invention belongs to fields of automation technology, propose a kind of controller design method for slowing down network congestion phenomenon.
By technologies such as constraint control, nonlinear Controls, effective control to network congestion phenomenon is realized, reduce the space to buffer area
It occupies, can be used for network communication field.
Background technique
Network congestion frequently problem and is concerned during network communication, and many network communications are all different
Suffer from this problem to degree.Network congestion will cause information transmission delay, and transmission speed reduces, and even result in network
Communication paralysis.During congestion, it may appear that loss of data, time delay increase, and throughput degradation results even in " congestion when serious
Collapse ".Usually, two class methods can alleviate network congestion phenomenon: the first kind is opened loop control, that is, passes through design one
Good algorithm avoids congestion.When carrying out congestion control, the current state of network is not considered.Second class is closed loop control
System, that is, give feedback mechanism, control congestion according to the current state of network.
The main reason for leading to network congestion be transmitted in packet switching network grouping number it is too many.In Internet traffic
The increasingly increased epoch, once communication peak period is encountered, in advance without the resource-sharing of any negotiation and request permissive mechanism
In network, several IP groupings will reach router simultaneously and it is expected to forward through the same output port.However, not all grouping
It can receive processing simultaneously, it is necessary to have a service order, the caching on intermediate node provides centainly for the grouping of waiting for service
Protection.But this situation has certain duration, and when spatial cache is depleted, router can abandon grouping, and data is caused to lose
It loses, or even generates " congestion collapse ".In the state of this sustained overload, network performance also will sharply decline.
With the introducing of intelligent network resource allocation system, that is, use advanced electronic information, communication, automatic control, meter
The technologies such as calculation machine network are distributed data streams to up to the time, so that network congestion problem by reasonably controlling, scheduler buffer
It is obviously improved.But the unconfined increase of output port memory space, it will lead to data packet forwarding time-out.Therefore, originally
Invention is based on positive system Feedback Control Design method, it can be achieved that the closed-loop control that node storage space is called when network congestion.
Summary of the invention
The purpose of the present invention is provide one kind and slow down network for the network congestion phenomenon often occurred in current network communication
The Feedback Control Design method of congestion.Specific technical solution is as follows:
A kind of control method slowing down network congestion, includes the following steps:
Step 1 establishes the state-space model of network congestion system, and specific method is:
Acquisition storage forward node delta data first, the state for establishing network congestion grid using these data are empty
Between model, form is as follows:
Wherein, f (x (t))=(f1(x1), f2(x2) ..., fn(xn))TIndicate the state in t moment network storage space, n
Indicate the info class of data flow, u (t) ∈ RmFor the state of t moment buffer area node, m indicates the info class of buffer data stream,
RmReal column vector is tieed up for m；A, B are sensor collected node data and the weighting matrixs that form in real time；Consider real system
Positivity, i.e. x (t), u (t) are non-negative always, it is assumed here that network congestion system is a kind of positive system model, meets A matrix
All off diagonal elements are all non-negative, B >=0, and " >=" is for each element in B matrix, i.e., all elements are all non-in B matrix
It is negative, σ (ti) indicate to work as t ∈ [ti, ti+1) when, σ (ti) a subsystem is in tiMoment is activated, in ti+1Moment is left.
Step 2 designing system matrix representation method is mentioned due to the uncertainty of network-caching area node resource service condition
More cell spaces do not know form to indicate the uncertainty of system out, and specific method is:
Collect one group of data on sytem matrix vertex in real time using sensor, sytem matrix A, B → [A (t), B (t)] are
A kind of more born of the same parents' three-dimensional-structures:
Wherein, A (t), B (t) indicate t moment weighting constant matrix, n be positive integer representation sensor acquisition vertex matrix
Number, p=1,2 ..., n, [Ap,Bp] represent matrix A, the sytem matrix of p-th of node of B.
In real life, since the resource of store-and-forward node is limited, thus planned network buffer area node provides step 3
Source be it is controlled, specific method is:
Wherein,δP andFor two positive numbers, p indicates p-th of network node, i.e. state change of the system in p-th of node
It is by memory space and limited bandwidth.
Step 4 planned network congestion control state feedback control law, comprises the concrete steps that:
The 4.1 linear remaining positive type Lyapunov functions of construction one shaped like:
Vp(x (t))=x (t)Tν(p)
Its derivative are as follows:
Wherein, for any time t, ν(p)Indicate that p-th of n ties up real column vector and each element is positive number.
4.2, which measure existing more cell spaces for step 2, does not know, and following optimization problem is made to have solution:
Constant ζ if it existsp, μ > 0, λ > 1, vector z(p)∈Rs,So that:
ν(p)> 0,
To any (p, q) ∈ S × S, p ≠ q is set up, and n indicates the number of vertex matrix；T indicates that vector or matrix transposition, z indicate a s dimension
Each element is both less than zero in real column vector and column, " > ", " >=", " < ", and "≤" is also for single in vector or matrix
Element size relationship；Ap,BpIt defines in step 2,It defines in step 3, vpIt is defined in step 5.1, vpWith vqUnanimously, full
Sufficient p, q ∈ 1,2 ..., n } and p ≠ q；
The 4.3 Lyapunov functions and its derivative constructed according to step 4.1 are more if step 4.2 optimization problem has solution
Cell space closed-loop system be it is stable, following inequality relation can be obtained:
4.4 condition according to optimization problem designed by step 4.2 can obtain following result:
4.5 combine the condition in the optimization problem in step 4.3 and step 4.4, can obtain following inequality relation:
4.6 to sum up step 4.1 to step 4.5 can obtain, web impact factor restrain form it is as follows:
Beneficial effects of the present invention are as follows:
Method of the present invention is directed to network transmission congestion problems, establishes the state-space model of buffer area node resource, constructs
One linear Lyapunov function, devises state feedback controller, is effectively utilized meshwork buffering area, and it is existing to reduce congestion
As.
The method of the present invention is based on positive system and proposes state feedback controller design, reduces traffic congestion phenomenon.
Adoption status feedback of the present invention implements the closed loop control of meshwork buffering area resource to network congestion phenomenon
System.By the means such as data acquisition to buffer resource, model foundation, state and constraint control, performance evaluation, establish
A kind of state feedback controller design method slowing down network congestion phenomenon can alleviate information transmission peak using this method and make
At congestion problems, guarantee information flow it is not overtime, do not lose under the premise of, system have good operational effect.
Specific embodiment
The invention will be further described below.
The control method for slowing down network congestion of the invention, includes the following steps
Step 1 establishes the state-space model of network congestion system, and specific method is:
Acquisition storage forward node delta data first, the state for establishing network congestion grid using these data are empty
Between model, form is as follows:
Wherein, f (x (t))=(f1(x1), f2(x2) ..., fn(xn))TIndicate the state in t moment network storage space, n
Indicate the info class of data flow, u (t) ∈ RmFor the state of t moment buffer area node, m indicates the info class of buffer data stream,
RmReal column vector is tieed up for m；A, B are sensor collected node data and the weighting matrixs that form in real time；Consider real system
Positivity, i.e. x (t), u (t) are non-negative always, it is assumed here that network congestion system is a kind of positive system model, meets A matrix
All off diagonal elements are all non-negative, B >=0, and " >=" is for each element in B matrix, i.e., all elements are all non-in B matrix
It is negative, σ (ti) indicate to work as t ∈ [ti, ti+1) when, σ (ti) a subsystem is in tiMoment is activated, in ti+1Moment is left.
Step 2 designing system matrix representation method is mentioned due to the uncertainty of network-caching area node resource service condition
More cell spaces do not know form to indicate the uncertainty of system out, and specific method is:
Collect one group of data on sytem matrix vertex in real time using sensor, sytem matrix A, B → [A (t), B (t)] are
A kind of more born of the same parents' three-dimensional-structures:
Wherein, A (t), B (t) indicate t moment weighting constant matrix, n be positive integer representation sensor acquisition vertex matrix
Number, p=1,2 ..., n, [Ap,Bp] represent matrix A, the sytem matrix of p-th of node of B.
In real life, since the resource of store-and-forward node is limited, thus planned network buffer area node provides step 3
Source be it is controlled, specific method is:
Wherein,δP andFor two positive numbers, p indicates p-th of network node, i.e. state change of the system in p-th of node
It is by memory space and limited bandwidth.
Step 4 planned network congestion control state feedback control law, comprises the concrete steps that:
The 4.1 linear remaining positive type Lyapunov functions of construction one shaped like:
Vp(x (t))=x (t)Tν(p)
Its derivative are as follows:
Wherein, for any time t, ν(p)Indicate that p-th of n ties up real column vector and each element is positive number.
4.2, which measure existing more cell spaces for step 2, does not know, and following optimization problem is made to have solution:
Constant ζ if it existsp, μ > 0, λ > 1, vector z(p)∈Rs,So that:
ν(p)> 0,
To any (p, q) ∈ S × S, p ≠ q is set up, and n indicates the number of vertex matrix；T indicates that vector or matrix transposition, z indicate a s dimension
Each element is both less than zero in real column vector and column, " > ", " >=", " < ", and "≤" is also for single in vector or matrix
Element size relationship；Ap,BpIt defines in step 2,It defines in step 3, vpIt is defined in step 5.1, vpWith vqUnanimously, full
Sufficient p, q ∈ 1,2 ..., n } and p ≠ q；
The 4.3 Lyapunov functions and its derivative constructed according to step 4.1 are more if step 4.2 optimization problem has solution
Cell space closed-loop system be it is stable, following inequality relation can be obtained:
4.4 condition according to optimization problem designed by step 4.2 can obtain following result:
4.5 combine the condition in the optimization problem in step 4.3 and step 4.4, can obtain following inequality relation:
4.6 to sum up step 4.1 to step 4.5 can obtain, web impact factor restrain form it is as follows:
Claims (5)
1. a kind of control method for slowing down network congestion, it is characterised in that include the following steps:
Step 1, the state-space model for establishing network congestion；
Step 2 proposes sytem matrix representation method；
Step 3, design angular domain condition solve the problems, such as that system mode is constrained；
The Feedback Control Laws of step 4, planned network blocking node.
2. a kind of web impact factor according to claim 1 and processing method, it is characterised in that step 1 is specific as follows:
Acquisition storage forward node delta data first, the state space mould of network congestion grid is established using these data
Type, form are as follows:
Wherein, f (x (t))=(f1(x1), f2(x2) ..., fn(xn))TIndicate the state in t moment network storage space, n indicates number
According to the info class of stream, u (t) ∈ RmFor the state of t moment buffer area node, m indicates the info class of buffer data stream, RmFor m dimension
Real column vector；A, B are sensor collected node data and the weighting matrixs that form in real time；Consider the positivity of real system,
That is x (t), u (t) are non-negative always, it is assumed here that network congestion system is a kind of positive system model, and it is all non-to meet A matrix
Diagonal entry is all non-negative,It is for each element in B matrix, i.e., all elements are all non-negative in B matrix, σ
(ti) indicate to work as t ∈ [ti, ti+1) when, σ (ti) a subsystem is in tiMoment is activated, in ti+1Moment is left.
3. a kind of network congestion system control according to claim 2 and processing method, it is characterised in that step 2 is specific such as
Under:
Collect one group of data on sytem matrix vertex in real time using sensor, sytem matrix A, B → [A (t), B (t)] are a kind of
More born of the same parents' three-dimensional-structures:
Wherein, A (t), B (t) indicate t moment weighting constant matrix, n be positive integer representation sensor acquisition vertex matrix
Number, p=1,2 ..., n, [Ap,Bp] represent matrix A, the sytem matrix of p-th of node of B.
4. a kind of network congestion system control according to claim 3 and processing method, it is characterised in that step 3 is specific such as
Under:
Wherein,δ pWithFor two positive numbers, p indicates p-th of network node, i.e., system the state change of p-th of node be by
Memory space and limited bandwidth.
5. a kind of network congestion system control according to claim 4 and processing method, it is characterised in that step 4 is specific such as
Under:
The 4.1 linear remaining positive type Lyapunov functions of construction one shaped like:
Vp(x (t))=x (t)Tν(p)
Its derivative are as follows:
Wherein, for any time t, ν(p)Indicate that p-th of n ties up real column vector and each element is positive number；
4.2, which measure existing more cell spaces for step 2, does not know, and following optimization problem is made to have solution:
Constant ζ if it existsp, μ > 0, λ > 1, vector z(p)∈Rs,So that:
To any (p, q) ∈ S × S, p ≠ q is set up, and n indicates the number of vertex matrix；er=[1 ..., 1]T∈Rr, T expression vector or matrix transposition, z indicate a s
It ties up each element in real column vector and column and is both less than zero,It is also in vector or matrix
Individual element size relation；Ap,BpIt defines in step 2,It defines in step 3, vpIt is defined in step 5.1, vpWith vq
Unanimously, meet p, q ∈ { 1,2 ..., n } and p ≠ q；
The 4.3 Lyapunov functions and its derivative constructed according to step 4.1, if step 4.2 optimization problem has solution, more born of the same parents
Body closed-loop system be it is stable, following inequality relation can be obtained:
4.4 condition according to optimization problem designed by step 4.2 can obtain following result:
4.5 combine the condition in the optimization problem in step 4.3 and step 4.4, can obtain following inequality relation:
4.6 to sum up step 4.1 to step 4.5 can obtain, web impact factor restrain form it is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910494944.1A CN110177019B (en) | 2019-06-06 | 2019-06-06 | Control method for slowing down network congestion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910494944.1A CN110177019B (en) | 2019-06-06 | 2019-06-06 | Control method for slowing down network congestion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110177019A true CN110177019A (en) | 2019-08-27 |
| CN110177019B CN110177019B (en) | 2021-11-30 |
Family
ID=67697164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910494944.1A Active CN110177019B (en) | 2019-06-06 | 2019-06-06 | Control method for slowing down network congestion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110177019B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650496A (en) * | 2019-09-24 | 2020-01-03 | 杭州电子科技大学 | Digital communication network congestion control method for suppressing interference |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103414245A (en) * | 2013-06-04 | 2013-11-27 | 浙江工业大学 | Quantization-based wide-area power system output feedback control method |
| CN103929777A (en) * | 2014-05-08 | 2014-07-16 | 西安电子科技大学 | Vehicle network data distribution congestion control method based on congestion game |
| CN109195179A (en) * | 2018-06-04 | 2019-01-11 | 杭州电子科技大学 | A kind of distributed congestion control of WSN network and power distribution method |
-
2019
- 2019-06-06 CN CN201910494944.1A patent/CN110177019B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103414245A (en) * | 2013-06-04 | 2013-11-27 | 浙江工业大学 | Quantization-based wide-area power system output feedback control method |
| CN103929777A (en) * | 2014-05-08 | 2014-07-16 | 西安电子科技大学 | Vehicle network data distribution congestion control method based on congestion game |
| CN109195179A (en) * | 2018-06-04 | 2019-01-11 | 杭州电子科技大学 | A kind of distributed congestion control of WSN network and power distribution method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650496A (en) * | 2019-09-24 | 2020-01-03 | 杭州电子科技大学 | Digital communication network congestion control method for suppressing interference |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110177019B (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Charny et al. | Congestion control with explicit rate indication | |
| Ryu et al. | Advances in internet congestion control | |
| Benmohamed et al. | Feedback control of congestion in packet switching networks: The case of multiple congested nodes | |
| Benmohamed et al. | A control-theoretic ABR explicit rate algorithm for ATM switches with per-VC queueing | |
| CN110662256B (en) | Multi-path cross-protocol transmission data packet scheduling method and system | |
| CN101803313A (en) | Methods and apparatus for providing congestion information | |
| CN102461093B (en) | Method of managing a traffic load | |
| CN102014075A (en) | Traffic policing method and device | |
| Ramamurthy et al. | A multidimensional framework for congestion control in B-ISDN | |
| CN110177019A (en) | A kind of control method slowing down network congestion | |
| Barrera et al. | Statistical approach for congestion control in gateway routers | |
| Zhao et al. | Bctcp: A feedback-based congestion control method | |
| Gibbens et al. | Distributed control and resource marking using best-effort routers | |
| Ramamurthy et al. | Analysis of the adaptive rate control for ABR service in ATM networks | |
| Gusat et al. | Extended ethernet congestion management (e2cm): Per path ecm-a hybrid proposal | |
| Wechta et al. | Simulation-based analysis of the interaction of end-to-end and hop-by-hop flow control schemes in packet switching LANs | |
| Yuan et al. | A generalized fast tcp scheme | |
| Yan et al. | Variable structure control in active queue management for TCP with ECN | |
| Athuraliya et al. | Optimization flow control with Newton-like algorithm | |
| Kotian et al. | Study on Different Mechanism for Congestion Control in Real Time Traffic for MANETS | |
| Misra et al. | Jointly coordinating ECN and TCP for rapid adaptation to varying bandwidth | |
| Siu et al. | Performance of TCP over ATM with time-varying available bandwidth | |
| Manikandan et al. | Active queue management based congestion control protocol for wireless networks | |
| Ishizuka et al. | Performance analysis of TCP over ABR in high-speed WAN environment | |
| Dong et al. | Predictive Queue-based Low Latency Congestion Detection in Data Center Networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220628 Address after: 510000 unit 501, 5 / F, Huihua commercial building, No. 80, Xianlie Middle Road, Yuexiu District, Guangzhou City, Guangdong Province, self compiled g0398 Patentee after: The Guangzhou Intellectual Property Service Co.,Ltd. Address before: 310018 no.1158, No.2 street, Xiasha Higher Education Park, Hangzhou City, Zhejiang Province Patentee before: HANGZHOU DIANZI University |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220907 Address after: 8A12, 7th Floor, Building 11, Dongshuijing Hutong, Dongcheng District, Beijing 100000 Patentee after: Yunwang Anxin (Beijing) Technology Co.,Ltd. Address before: 510000 unit 501, 5 / F, Huihua commercial building, No. 80, Xianlie Middle Road, Yuexiu District, Guangzhou City, Guangdong Province, self compiled g0398 Patentee before: The Guangzhou Intellectual Property Service Co.,Ltd. |