CN100405782C - Optical burst exchange route choosing method based on resource forecast - Google Patents

Optical burst exchange route choosing method based on resource forecast Download PDF

Info

Publication number
CN100405782C
CN100405782C CNB2004100097395A CN200410009739A CN100405782C CN 100405782 C CN100405782 C CN 100405782C CN B2004100097395 A CNB2004100097395 A CN B2004100097395A CN 200410009739 A CN200410009739 A CN 200410009739A CN 100405782 C CN100405782 C CN 100405782C
Authority
CN
China
Prior art keywords
information
nodes
node
wavelength
burst
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CNB2004100097395A
Other languages
Chinese (zh)
Other versions
CN1770729A (en
Inventor
李正斌
于刚
李维
单樑
李勤兵
袁驰
童霆
徐安士
谢麟振
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.)
Comtech Technology Shenzhen Co ltd
Original Assignee
Peking 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 Peking University filed Critical Peking University
Priority to CNB2004100097395A priority Critical patent/CN100405782C/en
Publication of CN1770729A publication Critical patent/CN1770729A/en
Application granted granted Critical
Publication of CN100405782C publication Critical patent/CN100405782C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention discloses a route selection method for optical burst exchange routes based on resource predictions, which comprises the following steps. Three state information of first level nodes is read in a database, which comprise a transmission distance L1, a refresh cycle T1, and the busy degree P1 of a wavelength and occupation information psi 1 of the wavelength. Corresponding information of second level nodes is read, each parameter is respectively substituted in a model based on a linear or a nonlinear discrete information prediction. A calculated result is a correction factor epsilon (0< epsilon < 1) of the confidence level of the state information of the corresponding nodes. Instant resource information of the node is computed with the original of wavelength information of the node psi (psi=0, 1). If psi is 0, the result is corrected to 0+epsilon. When psi is 1, the result is corrected to 1+epsilon. Each instant resource information of all the nodes in the first level nodes are used as the objects to form an array A. In the same way, each instant resource information of all the nodes in the second level nodes are used as the objects to form an array B. And Q=A*B' is computed. Two nodes corresponding to the maximum Q value are selected as the routes of the largest confidence level for burst packet transmission. The present invention can consumedly lower the blocking probability and the packet loss ratio of burst packages, and immensely improves the utility ratio of bandwidth.

Description

Light burst-switched route selection method based on resources
Technical field
The present invention relates to a kind of light burst-switched route selection method based on resources.
Background technology
Along with increasing rapidly of IP operation in the global range, just increase with unprecedented speed to transmitting guipure demand wide and the switching system capacity.At present, along with the maturation of close wavelength-division multiplex technology (DWDM), make the magnanimity bandwidth (every optical fiber can surpass 10Tbps) that makes full use of optical fiber become possibility.Thereby switching equipment partly becomes the bottleneck of communication network.The speed of the ip router of commercialization at present can reach 10~120Gbps, still has very big gap but compare with the transmission capacity of DWDM, realizes that full light exchange is a feasible program that solves this bottleneck.At existing communication pattern in the communication network, people have proposed three kinds of schemes to how realizing exchange and transmission in the WDM optical-fiber network:
1) light path exchanges light wavelength route Switch Solutions (OCS) just, and the wavelength route belongs to coarseness exchange, and its routing is relative with the link setup time all long, but the huge bandwidth resources of optical-fiber network are not fully utilized, and lacks flexibility.
2) light packet switching (OPS), but, therefore also be in conceptual phase at present owing to lack high-speed light logical device, optical cache memory etc., do not put in the practical application.
3) light burst-switched (OBS), this is a kind of scheme of comprehensive, compromise, it has merged the advantage of OPS and OCS, overcome the shortcoming of the two simultaneously, comparatively speaking, it is more flexible than wavelength route, the bandwidth availability ratio height, needing the light buffer memory again unlike OPS, is a kind of switching technology of at present most possible practicability.
In the exchange of light burst packets, at source node, the transmission flow of a common burst packets is: the group bag, routing sends signaling, sends data.At present, burst packets length is generally the ms magnitude, and the routing decision-making time is generally the ns magnitude, as shown in Figure 1, establishes t 1Constantly, source node receives the state updating message of other nodes, be generally the s magnitude to receive the time that other node state updating message are experienced next time, this time is far longer than the routing decision-making time, and the time interval that this routing table refreshes is uncertain, has therefore brought very big uncertain factor to Route Selection.Source node A is at t 2During routing constantly, if the state information of each node does not refresh as yet at this moment, the common practice is to think at t 1The constantly idle idle probability of channel when this burst packets arrives this channel is 1, and the idle probability in promptly selected path is that 1 confidence level will reduce along with the time gap that refreshed from last time increases, as shown in Figure 2.Abscissa is the time among the figure, ordinate is a Reliability of Information in the route information table, initial point is certain refreshes constantly of route information table, and this chart understands that flushing to time interval internal routing table Reliability of Information that refreshing information next time arrives from this increases in time and reduce.Method of the present invention is the forecasting mechanism that adds a routing on common routing algorithm basis.So whole network has been treated as a quasistatic network, yet other nodes also may be in routing, if other nodes are also chosen that channel of source node A choosing between twice state refresh, when two burst packets arrive this channel simultaneously, will make a bursty data packet congestion so.This route selecting method in real network, can cause bigger blocking rate and packet loss when particularly network is bigger and in the light burst exchange network of employing Just in Time agreement.
Summary of the invention
At the existing problem and shortage of burst packets routing in the above-mentioned existing light burst-switched, the purpose of this invention is to provide a kind of light burst-switched route selection method that can calculate and select for use immediately non-busy channel immediately to the situation that takies of node based on resources.
The present invention is achieved in that a kind of light burst-switched route selection method based on resources, may further comprise the steps,
1) source node reads the three condition information of one-level node: the distance L of transmission in associated databases when carrying out the group bag of light burst packets 1, the period T that refreshes 1, wavelength busy extent P 1With wavelength occupied information ψ 1Similarly, read the three condition information of two-level node: the distance L of transmission 2, the period T that refreshes 2, wavelength busy extent P 2And ψ 2
2) each parameter in the step 1) is updated to the model of predicting based on linear or non-linear discrete message respectively, as the Wiener filtering method, kalman filter method, and the consecutive numbers mathematic(al) function is as associating Gaussian membership function, the bell membership function, sigmoid function type membership function, the medium information prediction mechanism of trapezoidal membership function or Triangleshape grade of membership function, the result who calculates is the confidence level modifying factor ε (0<ε<1) of the state information of respective nodes, and with former wavelength information ψ (ψ=0 of this node, 1) calculates the instant resource information of this node together, be specially, if ψ was modified to 0+ ε at 0 o'clock, ψ was modified to 1-ε at 1 o'clock;
3) each the instant resource information with all nodes in the one-level node is that object is formed array A, and each the instant resource information with all nodes in the two-level node is that object is formed array B equally, and asks Q=A * B '; Obtain one group of confidence level state information through revising.
4) select the maximum confidence path of two corresponding nodes of a maximum maximum confidence Q value as the burst packets transmission.
Further, the wavelength occupied information that described node sends to the whole network state resource database is two states only, i.e. " 0 " one state, and wherein, " 0 " attitude is represented to take, and can utilize probability is 0; " 1 " expression is idle, and can utilize probability is 1.
The present invention thinks and carves t at a time when routing 1Idle channel is from t 1Constantly begin, its idle probability just differs and is decided to be 1, and it is time dependent, normally from t 1Constantly far away more, the confidence level of idle probability is more little.Routing should choose current time from the source node to the destination node between the channel of idle probability maximum (product maximum) send the burst packets data.The present invention will reduce the blocking probability and the packet loss of burst packets greatly, greatly improve bandwidth availability ratio.
Description of drawings
Below in conjunction with accompanying drawing, the present invention is made detailed description.
Fig. 1 is a state information refresh time schematic diagram;
Fig. 2 is a state information confidence level schematic diagram;
Fig. 3 is the routing schematic diagram;
Fig. 4 is a wavelength forecasting mechanism flow chart of the present invention;
Fig. 5 is the whole network state information database example of the present invention.
Embodiment
As shown in Figure 2, the present invention is at t 2During routing, think and carve t at a time 1The channel of the state free time of refreshing is from t 1Constantly begin, its idle probability just differs and is decided to be 1, and the confidence level of its idle probability is from t 1In time, is carved into and refreshes t constantly next time 3, be time dependent, normally from t 1Constantly far away more, idle probability is that 1 confidence level is more little.Add forecasting mechanism when routing after, the blocking probability of burst packets and packet loss can reduce.Compare with former data communication flow process, the flow process that the source node after the present invention improves sends a burst packets has become: group bag → prediction → routing → transmission signaling → transmission burst packet.
The present invention can adopt discrete message Forecasting Methodology and/or mathematical function Forecasting Methodology that node is taken situation to resource state information to carry out the resource status prediction.The present invention adopts linear and nonlinear prediction model such as Wiener filtering, Kalman filtering etc. to the discrete message forecasting mechanism.Then choose the resource status forecasting problem that fuzzy membership function carries out for the mathematical prediction function.For a fuzzy number, degree of membership shows the fog-level of this numerical value.The present invention provides the factor that influences the insincere degree of resource state information, reduces the following aspects:
1, the ability of node processing information.If the node processing ability is lower, then processing delay may not catch up with the variation of dynamic calling.
2, the distance of status information transmission.The nodal distance source node is far away more, and then information is can not reliability high more.
3, the speed that refreshes of node status information.Refresh cycle is short more, and speed is fast more, and information credibility is high more.
4, the historical information of each node in fringe node the whole network state information bank.
Among the present invention, each core node only has two states to the wavelength occupied information that the whole network state resource database sends, i.e. " 0 " " 1 " two condition.The present invention's definition " 0 " attitude is represented to take, and can utilize probability is 0; " 1 " expression is idle, and can utilize probability is 1.Therefore can distribute according to " 0 " one state of each node wave length resource historical information in the whole network resource database, judge the busy extent of corresponding wavelength, be about to occupied probability.
As shown in Figure 3, when the source node burst packets arrives, show in the whole network state information database after last time refreshes that the one-level node has only A1 to use, two-level node has only B2 to use, therefore routing is among the figure shown in the arrow, in fact be updated to the present moment from last the whole network state information database, the state of interior network is always in dynamic change during this period of time, therefore can not think this moment this selected path with probability 1 at leisure, and other 8 paths are exactly occupied with probability 1.
As shown in Figure 4, route selecting method of the present invention is to read the three condition information of one-level node in associated databases: the distance L of transmission 1, the period T that refreshes 1, wavelength busy extent P 1With wavelength occupied information ψ 1Similarly, read the three condition information of two-level node: the distance L of transmission 2, the period T that refreshes 2, wavelength busy extent P 2And ψ 2Above-mentioned each parameter is updated to linearity respectively, non-linear discrete message forecast model such as Wiener filtering, Kalman filtering etc., or mathematical function is as associating Gaussian membership function, the bell membership function, sigmoid function type membership function, in the anticipation functions such as trapezoidal membership function or Triangleshape grade of membership function, the result who calculates is the modifying factor ε (0<ε<1) of the state information of respective nodes, and with former wavelength information ψ (ψ=0 of this node, 1) calculates the instant resource information of this node together, be specially, if ψ was modified to 0+ ε at 0 o'clock, ψ was modified to 1-ε at 1 o'clock, had provided the example of state table in the whole network state information database according to the flow process of forecasting mechanism as Fig. 5.Can be clear that by chart node N is because the status information transmission distance and the wavelength that refresh are busier, the information fuzzy degree that therefore turns back to the whole network slip condition database is very high, with a low credibility, and therefore revised state is opposite with raw information; Instant resource information with all 3 nodes in the one-level node is that object is set up array A, A = 1 - &epsiv; A 1 0 + &epsiv; A 2 0 + &epsiv; A 3 ; Similarly, the instant resource information with all 3 nodes in the two-level node is an object
Set up array B, B = 0 + &epsiv; B 1 1 - &epsiv; B 2 0 + &epsiv; B 3 ; The transposition that matrix A be multiply by B obtains,
Q = A &times; B &prime; = ( 1 - &epsiv; A 1 ) &CenterDot; &epsiv; B 1 ( 1 - &epsiv; A 1 ) ( 1 - &epsiv; B 2 ) ( 1 - &epsiv; A 1 ) &CenterDot; &epsiv; B 3 &epsiv; A 2 &CenterDot; &epsiv; B 1 &epsiv; A 2 &CenterDot; ( 1 - &epsiv; B 2 ) &epsiv; A 2 &CenterDot; &epsiv; B 3 &epsiv; A 3 &CenterDot; &epsiv; B 1 &epsiv; A 3 &CenterDot; ( 1 - &epsiv; B 2 ) &epsiv; A 3 &CenterDot; &epsiv; B 3 .
By each Q among the comparator matrix Q IjSize, select a maximum Q Ij, i of its correspondence and j are respectively the paths that one-level node and two-level node need be selected.Like this, by a forecasting mechanism, the blocking rate and the packet loss that can reduce the data burst bag reduce, and improve the communication efficiency of network.

Claims (2)

1. the light burst-switched route selection method based on resources may further comprise the steps,
When 1) the source node burst packets arrives, in associated databases, read the three condition information of one-level node: the distance L of transmission 1, the period T that refreshes 1, wavelength busy extent P 1With wavelength occupied information ψ 1Similarly, read the three condition information of two-level node: the distance L of transmission 2, the period T that refreshes 2, wavelength busy extent P 2With wavelength occupied information ψ 2
2) each parameter in the step 1) is updated to respectively based on linear and nonlinear prediction model, or mathematical function is as associating Gaussian membership function, the bell membership function, sigmoid function type membership function, in trapezoidal membership function or the Triangleshape grade of membership function anticipation function, the result who calculates is the modifying factor ε of the state information of respective nodes, and calculate the instant resource information of this node with the former wavelength information ψ of this node, 0<ε<1 wherein, ψ=0,1, be specially, if ψ was modified to 0+ ε at 0 o'clock, ψ was modified to 1-ε at 1 o'clock;
3) each the instant resource information with all nodes in the one-level node is that object is formed array A, and each the instant resource information with all nodes in the two-level node is that object is formed array B equally, and asks Q=A * B ', and wherein B ' is the transposition of B;
4) select the path of two corresponding nodes of a maximum Q value as the burst packets transmission.
2. the light burst-switched route selection method based on resources as claimed in claim 1 is characterized in that, the wavelength occupied information that described node sends to the whole network state resource database is two states only, i.e. " 0 " one state, wherein, " 0 " attitude is represented to take, and can utilize probability is 0; " 1 " expression is idle, and can utilize probability is 1.
CNB2004100097395A 2004-11-02 2004-11-02 Optical burst exchange route choosing method based on resource forecast Expired - Fee Related CN100405782C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB2004100097395A CN100405782C (en) 2004-11-02 2004-11-02 Optical burst exchange route choosing method based on resource forecast

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB2004100097395A CN100405782C (en) 2004-11-02 2004-11-02 Optical burst exchange route choosing method based on resource forecast

Publications (2)

Publication Number Publication Date
CN1770729A CN1770729A (en) 2006-05-10
CN100405782C true CN100405782C (en) 2008-07-23

Family

ID=36751737

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2004100097395A Expired - Fee Related CN100405782C (en) 2004-11-02 2004-11-02 Optical burst exchange route choosing method based on resource forecast

Country Status (1)

Country Link
CN (1) CN100405782C (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102299738B (en) * 2011-07-01 2015-05-13 华为技术有限公司 Method for acquiring optical network link performance parameters and device thereof
CN103955584B (en) * 2014-05-12 2017-08-25 合肥工业大学 The upper bound optimization method of network-on-chip restructuring caching based on Multi-path route
CN111935018B (en) * 2020-07-23 2022-03-08 北京华云安信息技术有限公司 Springboard network path generation method capable of configuring networking rules autonomously
CN112770199B (en) * 2021-01-08 2022-06-17 杭州奥克光电设备有限公司 Management platform for intelligent optical traffic resources

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4894818A (en) * 1987-10-22 1990-01-16 Kokusai Denshin Denwa Kabushiki Kaisha Optical packet switching system using multi-stage combination of light triggering switches
CN1381963A (en) * 2002-06-06 2002-11-27 上海交通大学 Multi-grain optical router based on optical burst switch
CN1419353A (en) * 2002-10-24 2003-05-21 上海交通大学 Optical code division multiple control channel based optical burst package switching system
CN1510851A (en) * 2002-12-24 2004-07-07 ض� Method and device for controlling and dispatching data in WDM light switching network
CN1516412A (en) * 2002-12-23 2004-07-28 ���Ͽع����޹�˾ Congestion control in optical burst exchange network

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4894818A (en) * 1987-10-22 1990-01-16 Kokusai Denshin Denwa Kabushiki Kaisha Optical packet switching system using multi-stage combination of light triggering switches
CN1381963A (en) * 2002-06-06 2002-11-27 上海交通大学 Multi-grain optical router based on optical burst switch
CN1419353A (en) * 2002-10-24 2003-05-21 上海交通大学 Optical code division multiple control channel based optical burst package switching system
CN1516412A (en) * 2002-12-23 2004-07-28 ���Ͽع����޹�˾ Congestion control in optical burst exchange network
CN1510851A (en) * 2002-12-24 2004-07-07 ض� Method and device for controlling and dispatching data in WDM light switching network

Also Published As

Publication number Publication date
CN1770729A (en) 2006-05-10

Similar Documents

Publication Publication Date Title
CN1887024B (en) System and method for the automatic setup of switched circuits based on traffic prediction in a telecommunications network
Batayneh et al. On routing and transmission-range determination of multi-bit-rate signals over mixed-line-rate WDM optical networks for carrier ethernet
CN101552933B (en) Optical network self-adapting route system for low/middle orbit double-layer satellite and calculating method of agent route
Chan Optical flow switching networks
Ramamurthy et al. Translucent optical WDM networks for the next-generation backbone networks
CN1996921B (en) Method, route device and business network for establishing the business connection
Yu et al. A deep learning based RSA strategy for elastic optical networks
CN100405782C (en) Optical burst exchange route choosing method based on resource forecast
CN100407707C (en) Network configuration for loading high-speed data business and trans mitting method
CN109450587A (en) Spectral integration processing method, device and storage medium
Nleya et al. A Node‐Regulated Deflection Routing Framework for Contention Minimization
Harai et al. Optical fiber-delay-line buffer management in output-buffered photonic packet switch to support service differentiation
CN101753450B (en) Method for joint optimization of resources in three-layered network
Yang et al. Sparse regeneration in a translucent WDM optical network
Karamitsos et al. A resource reservation protocol with linear traffic prediction for OBS networks
Koubàa Routing, protection and traffic engineering in WDM all-optical networks
CN107181694A (en) The route frequency spectrum distributing method that a kind of utilization multithreading is realized
Puttasubbappa Optical burst switching: challenges, solutions and performance evaluation
Wang et al. A hybrid control architecture for connection management in translucent WDM networks
Todimala et al. A dynamic partitioning protection routing technique in WDM networks
Lee Analysis of Operation Areas for Automatically Tuning Burst Size-based Loss Differentiation Scheme Suitable for Transferring High Resolution Medical Data
Wang et al. Topology aggregation method in WSON networks
Ganguly Implementation and modeling of a scheduled optical flow switching (OFS) network
Luo et al. A new method for solving routing and wavelength assignment problems under inaccurate routing information in optical networks with Conversion Capability
Fourneau et al. Analysis and optimization of aggregation in a reconfigurable optical add/drop multiplexer

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
ASS Succession or assignment of patent right

Owner name: KETONG COMMUNICATION TECHNOLOGY (SHENZHEN) CO., LT

Free format text: FORMER OWNER: PEKING UNIVERSITY

Effective date: 20100519

C41 Transfer of patent application or patent right or utility model
COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 100871 NO.5 YIHEYUAN ROAD, HAIDIAN DISTRICT, BEIJING TO: 518057 C1001, CHUANGWEI BUILDING, SOUTH ZONE OF HIGH-TECH INDUSTRIAL PARK, NANSHAN DISTRICT, SHENZHEN CITY, ROOM C1003

TR01 Transfer of patent right

Effective date of registration: 20100519

Address after: 518057 c1003 building, SKYWORTH building, South District, Nanshan District hi tech Industrial Park, Shenzhen,

Patentee after: Comtech Technology (Shenzhen) Co.,Ltd.

Address before: 100871 Beijing the Summer Palace Road, Haidian District, No. 5

Patentee before: Peking University

CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20080723

Termination date: 20211102