CN103001884A - Distributed energy-saving routing traffic adjusting method - Google Patents

Distributed energy-saving routing traffic adjusting method Download PDF

Info

Publication number
CN103001884A
CN103001884A CN2012105373911A CN201210537391A CN103001884A CN 103001884 A CN103001884 A CN 103001884A CN 2012105373911 A CN2012105373911 A CN 2012105373911A CN 201210537391 A CN201210537391 A CN 201210537391A CN 103001884 A CN103001884 A CN 103001884A
Authority
CN
China
Prior art keywords
link
bandwidth
way link
ingress router
energy
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
Application number
CN2012105373911A
Other languages
Chinese (zh)
Other versions
CN103001884B (en
Inventor
李丹
余逸荣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Original Assignee
Tsinghua University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tsinghua University filed Critical Tsinghua University
Priority to CN201210537391.1A priority Critical patent/CN103001884B/en
Publication of CN103001884A publication Critical patent/CN103001884A/en
Application granted granted Critical
Publication of CN103001884B publication Critical patent/CN103001884B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention discloses a distributed energy-saving routing traffic adjusting method and belongs to the technical field of internet energy conservation. The method is characterized in that in an energy-saving routing method combining the shortest path tree and generating trees, when routers detect that utilization ratios of adjacent generating tree links are changed obviously, the routers conduct flooding on the current link utilization ratio situation, so that all routers can obtain the latest utilization ratio conditions of all generating tree links, and available bandwidth allowances of routers on all generating tree links can be calculated through an energy-saving scheduling algorithm. Routers can select the global shortest paths or generating tree paths again for streams starting form routers, conducts the phase of 'shifting out generating trees' to reduce overloaded link loads firstly and then conducts 'shifting in generating trees' to achieve traffic aggregation. Each of the routers conducts bandwidth calculation and traffic scheduling decision independently, and the good balance is achieved between traffic aggregation energy conservation and link overloading through distributed coordination.

Description

Distributed energy-conservation routing traffic method of adjustment
Technical field
GreenNet
Background technology
GreenNet is the important research direction of Next Generation Internet, and the power-saving mechanism of network level mainly is divided into centralized and distributed two large classes now.Distributed energy-conservation flow scheduling algorithm needs each source router to carry out independent decision-making according to the rule of setting usually, by the link load information analysis to collecting, guaranteeing under the network resilience constraints as much as possible traffic aggregation to less link, to make more link have an opportunity to enter into resting state.
The distributed energy-conservation routing traffic adjustment algorithm of shortest path tree and spanning tree combination is a kind ofly dynamically to determine that according to the link load situation needs whether carry out the flow adjustment algorithm of moving in the path.The objective of the invention is between offered load elasticity and energy-saving efficiency, to average out, both avoided link load too high, improve again energy-conservation efficient as far as possible.
Summary of the invention
The object of the invention is to design the dispatching algorithm of moving for every stream, calculate this according to the link budget information of stream process and flow whether needs carry out the path switching, and in network topology, switch fast between each bar link dynamically.
The present invention is characterized in:
1. on network topology, realize according to the following steps successively:
Step (1), distribute its idle bandwidth on one way link l for each ingress router according to the following steps:
Step (1.1), ISP ISP is that the above one way link l of online subgraph sets load higher limit X l,
Step (1.2), each ingress router i broadcasts mutually the following information on the described one way link l of self associated: total bandwidth C by LSA mechanism LSA l, total load u l, and energy-conservation flow p l, i=1,2 ... Z, Z are the sum of ingress router i,
Step (1.3), i ingress router judged the total load u on the one way link l lWhether at [X l-α, X l] in the scope, α be one to X lStable adjustment threshold value, span be [0,1), calculate respectively the bandwidth budget b of self by following formula I, l, b I, lBe the bandwidth budget at i the ingress router of one way link l,
Figure BDA0000257179951
X lBe the higher limit of the upper load of one way link l,
E lBe the upper expectation load value of setting of one way link l,
t I, lFor take i router as ingress router and through the big or small summation of all the energy-conservation flows on the one way link l,
p lBe the size of the energy-conservation total flow that flows through on the one way link l,
Z lBe the set of the ingress router that comprises one way link l on the shortest path tree,
Each bandwidth estimated value b that step (2) obtains according to step (1.3) I, l, i ∈ Z, carry out according to the following steps the path and switch:
Every stream selected path in last round of flow is adjusted on each link of the online subgraph of step (2.1) record,
The bandwidth budget that each ingress router of step (2.2) correlates for self is the one way link of negative value, order of magnitude order from big to small according to flow shifts out online subgraph whole or in part to the stream on the described one way link of flowing through, until satisfy till the bandwidth estimated value that step (1.3) obtains, and the bandwidth budget of renewal relevant link
The bandwidth budget that each ingress router of step (2.3) correlates for self on the occasion of one way link, check the stream on employed each bar shortest path in the last round of flow adjustment, according to the size order from small to large of flow absolute value whole or in part stream move each described bandwidth budget on the occasion of one way link on, until be no more than till the corresponding bandwidth budget, and upgrade immediately the bandwidth budget of relevant one way link.
The thinking of method proposed by the invention is: utilize Open Shortest Path First (OSPF) Routing Protocol to have flooding mechanism now, design distributed link utilization information broadcasting and reception mechanism, realize the energy consumption perception of network-wide basis, namely each ingress router has identical network link loads distribution.Research determines according to the link load situation whether stream is moved according to the stream dispatching algorithm of network traffics dynamic implementation.Design dynamically distributes bandwidth reserved according to the link load situation for the link bandwidth budget allocation algorithm of each ingress router, and determines that with this whether needing to carry out flow moves.Design judges according to the link budget information of stream process this flows whether needs carry out the path switching for the dispatching algorithm of moving of every stream.
The invention has the advantages that:
1. the employing distributed algorithm need not central controller and carries out centralized dispatching;
2. need not to collect accurate traffic demand information, each ingress router calculates the link bandwidth budget according to network link utilance situation and self traffic demand;
3. each ingress router is according to the bandwidth budget, for every flowable state ground take oneself as entrance carries out Path selection, obtains good balance by the principle of " go out greatly little advance " before stream shifting frequency and network energy-saving efficient.
Description of drawings
Fig. 1. distributed energy-conservation routing traffic method of adjustment schematic flow sheet.
Embodiment
In this Flow-rate adjustment algorithm, ingress router be every stream from two path candidates---dynamically select spanning tree path and the shortest path.The target of stream scheduling is traffic aggregation as much as possible on spanning tree, so that can enter the non-subgraph link number maximization of resting state.In algorithm for design, consider the following constraints: 1) load of spanning tree link need to remain on below certain threshold value, to reserve idle bandwidth reply flow spike; 2) the stream scheduling decision must be distributed, and namely each ingress router independent decision-making need not show with other routers to cooperate, and need not know the traffic matrix of whole network; 3) the algorithm speed of service wants enough fast, and ingress router could be made quick response to the network traffics variation like this.
The stream dispatching algorithm comprises that idle bandwidth distributes and two stages are switched in the path, sees Fig. 1.In the allocated bandwidth stage, because the flow on the direction of a concrete link might be simultaneously from a plurality of different ingress routers, need the idle bandwidth of this one way link of research how in these ingress routers, to distribute, study this one way link load over-expense part and how to divide at these ingress routers, ingress router need to be cut down the flow on this link.In the path switch step, each ingress router is investigated idle bandwidth or the overload bandwidth that obtains from distributing on every link, from controllable flow, choose wherein suitable data flow and carry out switching between the path, so that the expection of the trend of the load state on each link.
(1) allocated bandwidth stage algorithm.Represent the upper limit threshold of online subgraph uplink load with X, concrete numerical value is disposed voluntarily by Internet Service Provider (ISP).Each router is at the following information of subgraph by the coupled online subgraph link l of Link-state Advertisement (LSA) mechanism broadcasting: total bandwidth C l, total load u l, energy-conservation flow load p lEach ingress router i can receive the information of other all ingress router inundations.In addition, ingress router it is also understood that the big or small t of the upper energy-conservation flow of reserving from it of every online subgraph link l I, lIn order to control the path switching frequency, the load of online subgraph link should be controlled at [X l-α X l] in the scope, α is a less steady drift amount to X.Work as u lIn the time of in this scope, ingress router represents that the flow on this link does not need to adjust to the bandwidth budget zero setting on this link.Work as u lBeyond this scope, ingress router need to arrange the bandwidth budget, so that final effect is u lBring back to the expectation load
Figure BDA0000257179952
Work as u l>X lThe time, the method computation bandwidth budget that ingress router uses a kind of multiplication to reduce.Allow r lExpression reaches the ratio between the energy-conservation flow and current energy-conservation flow in the adjustment aim situation, even our target will make p so l* r l+ (u l-p l)=E l, then obtain
Figure BDA0000257179953
Then each ingress router calculates this ratio can calculate the bandwidth budget of oneself Work as u l<X lDuring-α, come the computation bandwidth budget with the method for adding gradually.Allow Z lThe set of the ingress router of online subgraph one way link l is used in representative.The bandwidth budget of ingress router i oneself is set to
Figure BDA0000257179955
That is to say, because each ingress router is not known the decision-making of other ingress routers, so it considers worst case during the computation bandwidth budget on an online subgraph, the router that namely might use this link all needs shared bandwidth, so can only divide idle bandwidth.Ingress router i the epicycle application to bandwidth be used after, more multi-band is wide if also need, and can re-start distribution in next round, through some take turns distribution after, its bandwidth demand can be met.Whole bandwidth allocation algorithm can be used equation description:
Figure BDA0000257179956
(2) path switch step algorithm.After being the good bandwidth budget of every strip line subgraph link calculation, ingress router need to be adjusted the flow path take it as entrance, satisfies the requirement of bandwidth budget so that final flow rate distributes.Be the link of negative value for budget, need to consider from moving out of online subgraph through choosing wherein a part the energy-conservation flow of this link.For budget be on the occasion of link, can hold more flows of moving online subgraph from shortest path.Therefore can be divided into two subs to the path switch step: flow shifts out the online subgraph stage, flow moves into the online subgraph stage.The strategy that we select " little advance to go out greatly " is implemented to move.In the immigration stage, the shortest path runoff of flow minimum is paid the utmost attention to be moved, the process of moving last till do not have stream can move or not move after the stream of excess budget not.On the contrary, in the stage of shifting out, failing to be convened for lack of a quorum of flow maximum preferentially moved, and the process of moving lasts till do not have stream to move or till the link budget requirement has been satisfied.Algorithm can be divided into three phases: 1) initial phase, record every path of flowing last round of selection.2) shift out the stage, ingress router checks from big to small that by flow all are last round ofly using streams of online subgraph, if a certain link budget is negative value on stream, then this stream should be moved out of online subgraph and upgrade immediately the budget of relevant link.3) the immigration stage, ingress router checks all last round of streams using shortest path from small to large by flow, does not cause link to excess budget after supposing to move this stream on the online subgraph, then implements to move and upgrade immediately the budget of relevant link.
We test 3 real topologys, the distributed energy-conservation routing traffic method of adjustment of stream of discovery spanning tree and shortest path tree combination can be tackled quick changes in flow rate effectively when adjusting flow, and adjust dynamically flow according to demand, illustrate that the present invention has reached the purpose of expection.

Claims (1)

1. distributed energy-conservation routing traffic method of adjustment is characterized in that, realizes according to the following steps successively on network topology:
Step (1), distribute its idle bandwidth on one way link l for each ingress router according to the following steps:
Step (1.1), ISP ISP is that the above one way link l of online subgraph sets load higher limit X l,
Step (1.2), each ingress router i broadcasts mutually the following information on the described one way link l of self associated: total bandwidth C by LSA mechanism LSA l, total load u l, and energy-conservation flow p l, i=1,2 ... Z, Z are the sum of ingress router i,
Step (1.3), i ingress router judged the total load u on the one way link l lWhether at [X l-α, X l] in the scope, α be one to X lStable adjustment threshold value, span be [0,1), calculate respectively the bandwidth budget b of self by following formula I, l, b I, lBe the bandwidth budget at i the ingress router of one way link l,
Figure FDA0000257179941
X lBe the higher limit of the upper load of one way link l,
E is the upper expectation load value of setting of one way link l,
t , lFor take i router as ingress router and through the big or small summation of all the energy-conservation flows on the one way link l,
p lThe size of the energy-conservation total flow that flows through on the link l,
Z lComprise the set of the ingress router of one way link l on the shortest path tree,
Each bandwidth estimated value b that step (2) obtains according to step (1.3) I, l, i ∈ Z, carry out according to the following steps the path and switch:
Every stream selected path in last round of flow is adjusted on each link of the online subgraph of step (2.1) record,
The bandwidth budget that each ingress router of step (2.2) correlates for self is the one way link of negative value, order of magnitude order from big to small according to flow shifts out online subgraph whole or in part to the stream on the described one way link of flowing through, until satisfy till the bandwidth estimated value that step (1.3) obtains, and the bandwidth budget of renewal relevant link
The bandwidth budget that each ingress router of step (2.3) correlates for self on the occasion of one way link, check the stream on employed each bar shortest path in the last round of flow adjustment, according to the size order from small to large of flow absolute value whole or in part stream move each described bandwidth budget on the occasion of one way link on, until be no more than till the corresponding bandwidth budget, and upgrade immediately the bandwidth budget of relevant one way link.
CN201210537391.1A 2012-12-12 2012-12-12 Distributed energy-efficient routing flow method of adjustment Expired - Fee Related CN103001884B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210537391.1A CN103001884B (en) 2012-12-12 2012-12-12 Distributed energy-efficient routing flow method of adjustment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210537391.1A CN103001884B (en) 2012-12-12 2012-12-12 Distributed energy-efficient routing flow method of adjustment

Publications (2)

Publication Number Publication Date
CN103001884A true CN103001884A (en) 2013-03-27
CN103001884B CN103001884B (en) 2015-08-19

Family

ID=47930037

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210537391.1A Expired - Fee Related CN103001884B (en) 2012-12-12 2012-12-12 Distributed energy-efficient routing flow method of adjustment

Country Status (1)

Country Link
CN (1) CN103001884B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103412919A (en) * 2013-08-09 2013-11-27 杭州华为数字技术有限公司 Dispatching method and device for distributed file system
CN105721311A (en) * 2014-12-22 2016-06-29 帕洛阿尔托研究中心公司 System and method for efficient name-based content routing using link-state information in information-centric networks
CN107204933A (en) * 2016-03-16 2017-09-26 华为技术有限公司 Distributed stream amount adjustment method and device
WO2018006619A1 (en) * 2016-07-06 2018-01-11 华为技术有限公司 Method for transmitting load information about path and network node
CN108259342A (en) * 2018-03-09 2018-07-06 华洋通信科技股份有限公司 A kind of network part congestion processing method based on ospf protocol

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004019569A1 (en) * 2002-08-20 2004-03-04 Telia Ab (Publ) A method for load balancing using short path routing
CN101572620A (en) * 2008-04-29 2009-11-04 华为技术有限公司 Method for constructing multi-tree topology and root node
CN101841482A (en) * 2010-05-07 2010-09-22 清华大学 Energy-saving routing method and device for network of data center

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004019569A1 (en) * 2002-08-20 2004-03-04 Telia Ab (Publ) A method for load balancing using short path routing
CN101572620A (en) * 2008-04-29 2009-11-04 华为技术有限公司 Method for constructing multi-tree topology and root node
CN101841482A (en) * 2010-05-07 2010-09-22 清华大学 Energy-saving routing method and device for network of data center

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103412919A (en) * 2013-08-09 2013-11-27 杭州华为数字技术有限公司 Dispatching method and device for distributed file system
CN103412919B (en) * 2013-08-09 2016-12-28 杭州华为数字技术有限公司 The dispatching method of a kind of distributed file system and equipment
CN105721311A (en) * 2014-12-22 2016-06-29 帕洛阿尔托研究中心公司 System and method for efficient name-based content routing using link-state information in information-centric networks
CN105721311B (en) * 2014-12-22 2020-09-11 思科技术公司 Method and system for updating link state information
CN107204933A (en) * 2016-03-16 2017-09-26 华为技术有限公司 Distributed stream amount adjustment method and device
CN107204933B (en) * 2016-03-16 2021-03-23 华为技术有限公司 Distributed flow adjusting method and device
WO2018006619A1 (en) * 2016-07-06 2018-01-11 华为技术有限公司 Method for transmitting load information about path and network node
CN107592269A (en) * 2016-07-06 2018-01-16 华为技术有限公司 The method and network node of the load information of transmission path
CN107592269B (en) * 2016-07-06 2020-03-31 华为技术有限公司 Method and network node for transmitting load information of path
US10965585B2 (en) 2016-07-06 2021-03-30 Huawei Technologies Co., Ltd. Method for transmitting path load information and network node
CN108259342A (en) * 2018-03-09 2018-07-06 华洋通信科技股份有限公司 A kind of network part congestion processing method based on ospf protocol

Also Published As

Publication number Publication date
CN103001884B (en) 2015-08-19

Similar Documents

Publication Publication Date Title
CN104767694B (en) A kind of stream compression forwarding method towards Fat Tree data center network architectures
CN103001884B (en) Distributed energy-efficient routing flow method of adjustment
CN103650435B (en) Routing traffic method of adjustment, device and controller
CN106209669B (en) Towards SDN data center network maximum probability path stream scheduling method and device
Zhang et al. Optimizing network performance using weighted multipath routing
CN105141517B (en) A kind of flexible optical-fiber network Anycast business energy-saving routing method based on resource perception
CN104871490B (en) The multipath communication device of energy ecology and its method for distributing business for improving energy ecology can be improved
CN102710489A (en) Dynamic flow division scheduling system and method
CN103124420B (en) Wireless on-chip network structuring method
CN103051547A (en) Routing and spectrum resource allocation method and system for resource awareness in elastic optical path network
CN104079496A (en) Double-deck satellite load balancing method based on link cost conversion
CN104869076A (en) Load balance adjusting method and device
CN103412635A (en) Energy-saving method and energy-saving device of data center
CN103297341A (en) Method for configuring flow on nodes of internal router
CN103346979A (en) Flow distribution method and equipment in SPBM network
CN105282056A (en) Novel photoelectric hybrid exchange data center network architecture
Soorki et al. Label switched protocol routing with guaranteed bandwidth and end to end path delay in MPLS networks
CN102098731A (en) Hop-based flow adaptive dormancy scheduling method in wireless sensor network
CN102025621A (en) Routing selection method of cognitive network based on mapping mechanism
Borylo et al. Anycast routing for carbon footprint reduction in WDM hybrid power networks with data centers
CN102123089A (en) Tunnel establishing method and device
CN104754589A (en) Traffic awareness based dynamic networking method
CN109474523B (en) Networking method and system based on SDN
CN107529128B (en) Routing method and system of intelligent power distribution network wireless sensor network
CN109067646A (en) It is a kind of flow table capacity by limited time SDN network flow optimization scheme

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20150819

Termination date: 20151212

EXPY Termination of patent right or utility model