CN101155131A - Method for establishing label switched path of minimized path preemption cost - Google Patents
Method for establishing label switched path of minimized path preemption cost Download PDFInfo
- Publication number
- CN101155131A CN101155131A CNA2006101414368A CN200610141436A CN101155131A CN 101155131 A CN101155131 A CN 101155131A CN A2006101414368 A CNA2006101414368 A CN A2006101414368A CN 200610141436 A CN200610141436 A CN 200610141436A CN 101155131 A CN101155131 A CN 101155131A
- Authority
- CN
- China
- Prior art keywords
- path
- router
- label switched
- link
- bandwidth
- 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
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides an establishing method of label switching path of minimizing path occupy cost in a multi protocol switching network. A link is wiped out when the not-reserved bandwidth is smaller than the bandwidth r of label switching path established by request in network topology. In the residual network topology, the residual bandwidth of each link is used as routing tolerance, and a priority algorithm of shortest path based on obligation is operated to find an optimal path of the said label switching path. For the said selected link on the optimal path, the residual bandwidth is analyzed whether it is more than or equal to the bandwidth r of label switching path established by request if yes the occupation is needless, if not, the occupation is necessary, the occupation formula is Y=alpha* p+ epsilon /|(b-r)|+sigma*n + theta*b. The invention not only establishes a label switching path for a new LSP request in minimal occupy cost, and obtains load balance and effectively increases the level of network resource utilization.
Description
Technical field
The present invention relates to data communication field, relate in particular to a kind of label exchange route setting method that in multi-protocol label switching network, minimizes the path preemption cost.
Background technology
At IP multiple service supporting net, traffic engineering based on multiprotocol label switching (mpls) particularly is provided on the backbone network, load allocating with maximization network resource utilization, equalizing network, and the differentiated service of multiple business such as realization voice, video and data, be present IP communication network field active subject the most.The core of the traffic engineering of MPLS is RSVP and management.Become rare and exist in the network environment of Business Stream of a large amount of delay sensitives in bandwidth resources, seize the business (label switching path LSP) that to guarantee high priority and can be routed on the favourable relatively path, with the service quality of the business of guaranteeing such.The low priority traffice stream of being seized (label switched path) will be re-routed to the path of suboptimum or be removed.Realize RSVP and management in the traffic engineering (DS-TE) of difference sensitivity, preemptive priority has become a very attractive strategy.And, be that the ASON (ASON) of controlling core has begun commercialization with GMPLS, seize the recovery time that can reduce under the high-priority service fault condition.
In the MPLS traffic engineering, the new LSP request of setting up has two important parameters: bandwidth and preempting priority.In order to minimize preemption cost, we can select the LSP that seized by optimizing a target function.This target function can be the one or more combination in following 3 parameters:
1) seizes the LSP of lowest priority.The service quality (QoS) of high priority LSP business can access and better meet.
2) seize minimized number LSP.Need rerouting or reversed LSP number to reduce to minimum.
3) seize and still can satisfy new LSP and set up request minimized number bandwidth.Network resource utilization will improve.
At present commercial and have LSP that the router of MPLS traffic engineering selects lowest priority usually as seizing object, the result causes the waste of massive band width resource and plurality purpose LSP to re-route or remove.Propose to seize the preemptive policy of the sequence of importance of optimizing criterion in some documents, as LSP number, bandwidth and the priority of being seized by certain; Or LSP priority, bandwidth and the LSP number seized.Actual conditions are, operator may constantly adjust these 3 parameters even the sequence of importance of multi-parameter more according to user and Internet resources applicable cases.J.C.de Oliveira, C.Scoglio, I.F.Akyildiz., G.Uhl, " Newpreemption policies for DiffServ-aware traffic engineering tominimize rerouting in MPLS networks; " IEEE/ACM Trans.Networking, vol.12, no.4, pp.733-745, the document of Aug.2004 has proposed a kind of agile and all-purpose preemptive policy, is called V-PREPT (a kind of general minimizing seized and optimized formula).This preemptive policy combines 3 main optimization criterions of seizing: the LSP number of being seized, the LSP priority of being seized and the LSP bandwidth summation of being seized.The novelty of strategy V-PREPT is: the needs according to operator are adjusted three importance degrees of optimizing criterion flexibly, do not need to arrange special optimization criterion order.
The network that constitutes for the LSP by peanut, using V-PREPT to optimize LSP that the formula online in real time selects to be seized is a kind of method comparatively fast and accurately.Yet, for the large scale network of forming by a large amount of LSP, use the computation complexity of this optimization formula and computing time to increase along with the increase exponentially of N, prove on the mathematics that this is a NP-complete problem, be unfavorable for that obviously the online in real time of router is calculated.Design certain heuritic approach and will help reducing complexity, reduce computing time, but may bring approximate optimization result, even optimize accurate result that formula obtains with V-PREPT and compare and greatly differ from each other.J.C.de Oliveira, C.Scoglio, I.F.Akyildiz., G.Uhl, " New preemption policies forDiffServ-aware traffic engineering to minimize rerouting in MPLSnetworks; " IEEE/ACM Trans.Networking, vol.12, no.4, pp.733-745, the document of Aug.2004 has also proposed a kind of heuritic approach, though this algorithm can reduce computing time, brings low selection accuracy.When new LSP bandwidth on demand increased, this accuracy reduced along with the increase of selecteed LSP number.
Minimize preemptive policy V-PREPT and can use and constituting by single link, or requested LSP is made up of the multistage link and all links all move the occasion of identical LSP for requested LSP.Such as the label switched path that is requested to set up is LSP1, it is made of the multistage link, as L1-L2-L3-L4-L5, L1 wherein, L2, L3, L4, L5 is 5 sections links that join in turn, L1, L2, L3, L4, L5 is moving LSP2 at this moment, these two LSP of LSP3, do not move other LSP, at this moment if carry out the words of preemptive priority, because LSP2, LSP3 is by L1, L2, L3, L4, move on the entire path that L5 constitutes, therefore, this LSP that is made of the multistage link minimizes and seizes the LSP that constitutes with single link to minimize preemption scenario identical.Yet in real network, LSP tends to cross over a plurality of network nodes, and promptly LSP is made of the multistage link, and the LSP that moves on every section link also has nothing in common with each other.Such as requested label switched path is LSP1, and it is made of the multistage link, as L1-L2-L3-L4-L5, wherein LSP2 moves on L1, L2, LSP3 moves on L2, L3, L4, and LSP4 moves on L1, and LSP5 moves on L1, L2, L3, L4, L5, that is to say, the last operation of L1 has LSP2, LSP4, LSP5, and the last operation of L2 has LSP2, LSP3, LSP5, and the last operation of L3 has LSP3, LSP5, the last operation of L4 has LSP3, LSP5, and the last operation of L5 has LSP5.In this case, problem is just complicated.Adopt V-PREPT on every section link, to select the LSP that is seized,, might not cause being requested LSP seizing on entire path and minimize though can make seizing of single link minimize.One selected, be requested on the path that LSP must pass through, the number of links that operation each bar LSP is thereon crossed over may have nothing in common with each other: some LSP may only cross over one section link in this path, and some LSP may cross over multistage even all links.Obviously, seize those LSP that cross in this path than multilink and will reduce the LSP number of being seized in the entire path, total path preemption cost also could reduce!
Selecting the path of a minimum preemption cost also is a NP-complete problem.At present, the Path selection that seldom has document that minimizing of latter event seized is studied.B.Szviatovszki, A.Szentesi, and A.Juttner. " Minimizing re-routing inMPLS networks with preemption-aware constraint-basedrouting ", Computer Commuuications 25 (1076-1083), 2002 documents have proposed a kind of new Path selection solution of seizing that minimizes based on constraint Open Shortest Path First (CSPF) algorithm.This solution is utilized a kind of based on constraint Open Shortest Path First (CSPF) algorithm, it is dijkstra's algorithm, this CSPF algorithm is well known in the prior art, its basic idea is: use certain index (for example link cost, link remaining bandwidth, chain-circuit time delay etc.) of every link in the network topology as route metric, just can search out optimal path for label switched path by this algorithm.In the document, with the link cost of all links in the network as first route metric, seek optimal path by this algorithm, and the bandwidth of being seized because of the limit priority number of seizing affected label switched path and this label switched path measured as secondary route, when the optimum that obtains based on first route metric is set up more than one of path, by selecting an optimized LSP to set up the path based on second tolerance.After selecting optimized LSP and setting up the path, whether the remaining bandwidth of each section link of checking this optimal path is more than or equal to the bandwidth of LSP to be set up, if the remaining bandwidth of certain section link then just need not be seized low priority LSP more than or equal to the required bandwidth of LSP to be set up on this section link; If less than the required bandwidth of LSP to be set up, on this section link, seize the bandwidth of existing LSP according to priority order from low to high, till the bandwidth summation of being seized can satisfy the bandwidth of LSP to be set up, so just drawn the LSP number that to seize for one section link, gather constituting the LSP number of seizing that each section link that this selected optimal L SP sets up the path obtains then, this has just obtained total cost that the path is seized.
Yu Ke, Zhang Lin, Zhang Hui-min, " A preemption-aware pathselection algorithm for DiffServ/MPLS networks; " Proc.IEEE workinggroup on Oct.2004, pp.1374-1377.pp:129-133 has proposed the another kind of Path selection solution of seizing that minimizes, it and B.Szviatovszki, A.Szentesi, and A.Juttner. " Minimizing re-routing in MPLS networks withpreemption-aware constraint-based routing ", ComputerCommunications 25 (1076-1083), 2002. difference only be: it uses total bandwidth that every section link seized and the limit priority of being seized LSP to make up a kind of distance function as route metric, moves the best that the CSPF algorithm obtains being requested LSP and sets up the path.
Above-mentioned two kinds of shortcomings that minimize the Path selection solution of seizing are: use same preemptive policy, promptly always seize the LSP of lowest priority at first, and not considering that the bandwidth of the LSP that seized and these LSP cross over the link condition that selected optimal L SP sets up the path, the result causes non-minimum preemption cost Path selection.
In addition, above-mentioned B.Szviatovszki, A.Szentesi, and A.Juttner. " Minimizing re-routing in MPLS networks with preemption-awareconstraint-based routing ", Computer Communications 25 (1076-1083), 2002. technology can not realize the purpose of the load balancing and the utilization of resources efficiently owing to be subjected to the constraint of SPF principle.
Before describing content of the present invention, earlier technical foundation of the present invention is carried out following description.
Currently constitute by four main functional modules in the MPLS of real network deploy traffic engineering (MPLS-TE).First functional module is IGRP (IGP) expansion, traffic engineering that promptly Open Shortest Path First (OSPF-TE), or intermediate system-intermediate system traffic engineering (IS-IS TE), by the not bandwidth reserved (UB) of each priority on every link of they issues, these bandwidth information be stored in router traffic engineering database (TED) and regularly or trigger and upgrade.Second functional module is to calculate and select the path for a new LSP request.This process is called based on constraint route (CR), produces an explicit path (ER) based on constraint route calculation result, i.e. the tabulation of a new necessary LSR (LSR) of crossing over of LSP request.The 3rd functional module is for the operation of this new LSP bandwidth reserved with the label distribution protocol of distribution label, and current, the main flow label distribution protocol that traffic engineering is used is RSVP expansion, i.e. RSVP-TE.The 4th functional module is that each LSR on the explicit path is that this LSP implements to admit control, upgrades UB.
Because all types of service are subjected to the restriction of single bandwidth constraint, MPLS-TE can not guarantee for polytype business provides QoS that become more meticulous more, discrepant.Be suggested in nearly 2 years based on the allocated bandwidth of class of service and in conjunction with the DifferentiatedServices-aware MPLS traffic engineering (DS-TE) of DiffServ QoS characteristic.DS-TE has introduced type level (CT), bandwidth constraint (BC), traffic engineering classification new ideas such as (TE-Class).In the DS-TE network environment, TE-Class is main LSP attribute.In order to support DS-TE, original MPLS-TE agreement (OSPF-TE, IS-IS TE and RSVP-TE) further is expanded.The RSVP-TE of expansion carries a new object-LSP CT value, together with foundation and maintenance priority territory that SESSION_ATTRIBUTE object among the former RSVP-TE is comprised, a RSVP-TE PATH (PATH Message) includes the complete information that identifies TE-Class.The IGPs-TE of expansion comprises a new type-length-value (TLV) object, be used to provide bandwidth constraint (BC) information, and it is the not bandwidth reserved (UB) of 8 TE-Class that original TLV object that is used to carry the UB of 8 priority that belong to single BC is redefined.
MPLS-TE router in the current network is regularly announced the UB of each priority on every the link that is attached thereto to other node in the network by IGPs-TE.LSR in the network does not have the attributes such as routing information, priority and bandwidth of all LSP.Be a new LSP when calculating a path based on CSPF, the LER of head end (LER) knows that chosen path is made of specific LSR.The LSP attribute that all result from LER comprises that CT, preempting priority, bandwidth and routing information all are stored in its TE Link manager.When LER uses RSVP-TE is a new LSP when setting up a strip label switching path, and it sends PATH message along downstream LSR, and the EXPLICIT_ROUTE object in the PATH message has been stipulated the LSR tabulation that this lsp tunnel must be crossed over.Each next jump LSR and receive PATH message, deletion it oneself from the EXPLICIT_ROUTE object.Each LSR in this LSP explicit path is that specific preempting priority LSP implements to admit control, and upgrades the UB of each priority on the link.In the DS-TE environment, different is that each LSR is that the LSP of specific T E-Class implements to admit control, and upgrades the UB of each TE-Class on the link.
Summary of the invention
An object of the present invention is to overcome the priority that prior art only considers to be seized LSP, do not consider the LSP bandwidth seized, the matching degree of this bandwidth and the required bandwidth of seizing, especially do not consider and these are crossed over the link condition that selected optimal L SP sets up the path by the LSP that seized, the result causes the problem of non-minimum preemption cost, a kind of bandwidth that also will consider the LSP that seized when considering to be seized the priority of LSP is provided, the matching degree of this bandwidth and the required bandwidth of seizing, especially consideration and these LSP that is seized cross over the link condition that selected optimal L SP sets up the path, make the label exchange route setting method that minimizes the path preemption cost of preemption cost minimum.
Another object of the present invention is to reach load balancing and network resource utilization efficiently.
The invention provides a kind of label exchange route setting method that in multi-protocol label switching network, minimizes the path preemption cost, comprising:
Step a. is in the time will setting up the label switched path of a certain priority between head end router and terminal router, wipe out in the multi-protocol label switching network topology not bandwidth reserved less than the link of the bandwidth r of the label switched path that is requested to set up, in the rest network topology, the head end router uses the remaining bandwidth of every link as route metric, operation is based on the shortest-path first algorithm (CSPF) of constraint, it is dijkstra's algorithm, for this label switched path is sought optimal path, if there are many such paths, select an optimal path of crossing over the minimum path of jumping figure as this label switched path;
Step b. is for every link on the optimal path selected in step a, analyze its remaining bandwidth whether more than or equal to the bandwidth r of the label switched path that is requested to set up, if the remaining bandwidth of all links is all more than or equal to the bandwidth r of the label switched path that is requested to set up on this optimal path, the head end router uses based on the RSVP of traffic engineering and sets up label switched path for the label switched path of this request, and upgrade this path what write down in each router of process is the not bandwidth reserved of that section link of starting point with this router, this path each router of process when the PATH message that receives from upstream router, write down the routing information of EXPLICIT_ROUTE object in this PATH message, i.e. each jumping in this router downstream in this label switched path, when next round has new label switched path request comes and needs under the situation of bandwidth-hogging resource, the routing information of these records is used for the degree that label switched path that more every label switched path of having set up is requested to set up with this next round is shared same link; If the remaining bandwidth of one or more snippets link of existence can not be more than or equal to r on the optimal path that should select, need contrast the low label switched path implementation of this priority that is requested label switched path on these links seizes, seizing by this link top router on the such link of certain bar carried out, concerning on it arbitrary wait to seize the label switched path, seize formula and be:
Y=α p+ ε/| (b-r) |+δ n+ θ b---formula 1
In the formula 1, p represents to wait to seize the priority of label switched path, b is a bandwidth of waiting to seize label switched path, 1/| (b-r) | the expression Bandwidth adaptation, n represents that label switched path to be seized is positioned at those links in this router downstream, link top and should selectes optimal path shared, and the link remaining bandwidth is less than the number of links of the bandwidth r that is requested label switched path, factor alpha, ε, δ, θ is the p in the formula 1,1/| (b-r) |, n, the weight that b distributed, remain to be seized label switched path for the institute on this link that need seize, by formula 1 calculate the Y value one by one, the label switched path of being seized according to Y value selective sequential from big to small in every strip label switching path then, up to the bandwidth of the selected label switched path of seizing with more than or equal to r, for existing the remaining bandwidth of multistage link can not be on this selected optimal path more than or equal to the situation of r, account for enterprising the robbing of this multistage link successively according to the order that is from upstream to the downstream, all links that need seize on this selected optimal path have all been implemented preemptive priority, the head end router of optimal path that should be selected still regularly sends PATH message, this path each router of process when the PATH message that receives from upstream router, write down the routing information of EXPLICIT_ROUTE object in this PATH message.
In step b, the flow process of seizing and set up this label switched path for new label switched path foundation request enforcement is as follows:
B1. the head end router compares the bandwidth on demand r of this label switched path and the head end router remaining bandwidth to this section of next hop router link of head end router, if this remaining bandwidth is more than or equal to r, the head end router will be according to based on the RSVP of the traffic engineering regulation to PATH message, produce PATH message, send to the next hop router of head end router, i.e. second hop router, enable the path simultaneously and set up PATH message retransmission timer, after second hop router is received PATH message, the routing information of the EXPLICIT_ROUTE object of record PATH message, and the address of from the EXPLICIT_ROUTE object of PATH message, deleting this second hop router itself, check the bandwidth on demand r of this label switched path in the PATH message then, the remaining bandwidth of this section link of linking to each other to next hop router-Di three hop routers of second hop router of this bandwidth on demand r and this second hop router relatively, if this remaining bandwidth is more than or equal to r, second hop router will be according to based on the RSVP of the traffic engineering regulation to PATH message so, produce a new PATH message, then it is issued next hop router-Di three hop routers of this second hop router that indicates in the EXPLICIT_ROUTE object, the rest may be inferred, if the remaining bandwidth of all links in the selected optimal path is all more than or equal to r, PATH message will arrive the last-hop Router in the EXPLICIT_ROUTE object always, it is terminal router, terminal router comprises request and sets up the RESV message that label switched path uses label through returning to the head end router from each router of swimming over to the upstream down on this selected optimal path, receive RESV message and this message when the head end router and comprise when being requested to set up the label that label switched path uses that PATH message retransmission timer is set up in the forbidding path; If should select the remaining bandwidth of certain section link in the optimal path less than r, the top router of this link is just no longer sent out PATH message to the tail end router of this link, but along returning PATH ERR message to the head end router from the direction of swimming over to the upstream down, simultaneously, this top router is implemented to seize according to the label switched path of formula 1 on this link, the label switched path information that to be seized from this label switched path delete in the routing information table that writes down the all-router of process, upgrade simultaneously this label switched path the not bandwidth reserved information in the all-router traffic engineering database of process about link, seize finish after, the routing information of the label switched path that is requested to set up also from this label switched path that is requested to set up delete in the routing information table that writes down the all-router of process;
B2. this path value of setting up PATH message retransmission timer preferably is provided with to such an extent that to seize the time that is spent than the label switched path that carries out formula 1 on one section link long, like this, when a PATH message arrives at certain section needs and seizes the top router of link of link bandwidth, this top router carries out described link bandwidth and seizes on this link, when next PATH message arrives at this top router, the preemptive priority of this section link has been accused and has been finished, PATH message will arrive the router that next place need seize link bandwidth, can reduce the path like this and set up the number of times that PATH message retransmits, to set up PATH message retransmission timer overtime when the path, but when the head end router is not received the RESV message of the label that comprises the label switched path use of asking foundation, the head end router is attempted sending PATH message along this optimal path once more, enable the path simultaneously and set up PATH message retransmission timer, in setting up the timing of PATH message retransmission timer, the path waits for the RESV message that comprises the label that label switched path that request sets up uses once more, when receiving this message, requested label switched path is set up process and is finished, if timer expiry and do not receive this message, the head end router is attempted sending PATH message along this optimal path for the third time, enable the path simultaneously and set up PATH message retransmission timer, and the like, till the RESV message of the label of receiving the label switched path use that comprises request foundation, forbid the path and set up PATH message retransmission timer this moment, and the foundation declaration of the label switched path of this request finishes.
Because prior art when each section link is seized, is seized the LSP of existing lowest priority on this link earlier, and is not considered other factors after finding out optimized LSP and set up the path according to the CSPF algorithm.Characteristics of the present invention are: after finding out optimized LSP according to the CSPF algorithm and setting up the path, creatively proposed preemptive priority formula Y=α p+ ε/| (b-r) |+δ n+ θ b, and low priority LSP seizes on the link of needs being implemented to seize based on this formula.Like this, except the priority of LSP is seized in consideration, also to consider the bandwidth of the LSP that seized, LSP bandwidth that this quilt is seized and the required bandwidth of seizing matching degree, especially consider that these LSP that seized cross over the link condition that selected optimal L SP sets up the path, so just can realize that real preemption cost minimizes.
The present invention has considered especially that these LSP that seized cross over and has been positioned at this and seizes the node downstream, set up the shared same link situation in path with selected optimal L SP, promptly seize those shared these selected paths as far as possible and be positioned at the LSP of this node downstream than multilink, thereby reduce the LSP number that quilt is seized in the entire path, make total path preemption cost minimum.Because method of the present invention can be seized with selected optimal L SP and set up the shared more LSP of multilink in path, this method really can reach to minimize to be seized, and reduces the LSP heavy-route cost of being seized.
Another characteristics of the present invention are to adopt the remaining bandwidth of MPLS network link as route metric, use the CSPF algorithm and at first select to have in the link path of maximum remaining bandwidths for new LSP request, so just reduced the LSP number of being seized on the whole, making to minimize to seize becomes possibility.And, can reach load balancing with the remaining bandwidth of every link in the network preferably as the optimal path that path metric calculates, improve network resource utilization.
Based on prior art, the routing information of the LSP that network intermediate router and not knowing has been set up, i.e. the node situation of each bar LSP passing through network.The present invention is by changing the flow process of setting up of label switched path, allow intermediate router write down the routing information of EXPLICIT_ROUTE object in the PATH message, like this, just there is way to seize to be positioned at this node downstream, shares the more LSP of multilink, thereby reach the purpose that minimizes heavy-route LSP number with selected optimal L SP path.The present invention increases a path at the head end router and sets up PATH message retransmission timer, is successfully set up on the selected path that needs are seized to guarantee new request LSP.The time of PATH message retransmission timer can be adjusted according to network size and loading condition.Do not need existing RSVP-TE agreement is done any change, network has good backwards compatibility and adaptability.
Embodiment
The invention provides a kind of label exchange route setting method that in multi-protocol label switching network, minimizes the path preemption cost, comprising:
Step a. wipes out in the multi-protocol label switching network topology not bandwidth reserved less than the link of the bandwidth r of the label switched path that is requested to set up in the time will setting up the label switched path of a certain priority between head end router and terminal router.For existing Multiprotocol Label Switching Traffic Engineering, wherein bandwidth reserved (UB) is not meant that remaining bandwidth on this link adds that priority on this link is lower than the summation of bandwidth reserved of all label switched paths of setting up priority of this label switched path that is requested to set up; Traffic engineering (DS-TE) for the differential service sensitivity that is about to dispose, link not bandwidth reserved corresponding to the not bandwidth reserved of traffic engineering rank (TE-Class) under this label switched path that is requested to set up, the different bandwidth restricted model in the differential service sensitive flow engineering is depended in the calculating of this value, as maximum allocated bandwidth restricted model (MAM), Russia's cover baby's bandwidth constraint model (RDM), this calculating is prior art, so seldom give unnecessary details.In the rest network topology, the head end router uses the remaining bandwidth of every link as route metric, the operation of head end router is based on the shortest-path first algorithm (CSPF) of constraint, (this CSPF algorithm is described in background technology for this label switched path is sought optimal path, its any parameter that can utilize every link is as route metric, after moving this algorithm, can both find out optimal path, this algorithm is existing algorithm), if there are many such paths, select an optimal path of crossing over the minimum path of jumping figure as this label switched path.The number of links of the promptly middle process of jumping figure.Each link is called a jumping.
Step b. is for every link on the optimal path selected in step a, analyze its remaining bandwidth whether more than or equal to the bandwidth r (need not seize) of the label switched path that is requested to set up more than or equal to the r explanation, if the remaining bandwidth of all links is all more than or equal to the bandwidth r of the label switched path that is requested to set up on this optimal path, the head end router uses based on the RSVP of traffic engineering and sets up label switched path for the label switched path of this request, and upgrade this path what write down in each router of process is the not bandwidth reserved when judging the concerning of not bandwidth reserved and r when next time receiving the request of setting up new label switched path in step a (so that use) of that section link of starting point with this router, this path each router of process when receiving PATH message from upstream router (PATH message is a kind of message in the existing RSVP-TE agreement), write down the routing information of EXPLICIT_ROUTE object in this PATH message, i.e. each jumping (regulation carries this information in the routing information of the EXPLICIT_ROUTE of PATH message object) in this router downstream in this label switched path, when next round has new label switched path request comes and needs under the situation of bandwidth-hogging resource, the routing information of these records is used for the degree that label switched path that more every label switched path of having set up is requested to set up with this next round is shared same link; If the remaining bandwidth of one or more snippets link of existence can not be more than or equal to r on the optimal path that should select, need be on these links be requested to carry out in the low label switched path of the priority of label switched path than this seized, seizing by this link top router on the such link of certain bar carried out, concerning on it arbitrary wait to seize the label switched path, seize formula and be:
Y=α p+ ε/| (b-r) |+δ n+ θ b---formula 1
In the formula 1, p represents to wait to seize the priority of label switched path, b is a bandwidth of waiting to seize label switched path, 1/| (b-r) | the expression Bandwidth adaptation, n represents that label switched path to be seized is positioned at those links in this router downstream, link top and should selectes optimal path shared, and the link remaining bandwidth is less than the number of links of the bandwidth r that is requested label switched path, factor alpha, ε, δ, θ is the p in the formula 1,1/| (b-r) |, n, the weight that b distributed, remain to be seized label switched path for the institute on this link that need seize, by formula 1 calculate the Y value one by one, the label switched path of being seized according to Y value selective sequential from big to small in every strip label switching path then, up to the bandwidth of the selected label switched path of seizing with more than or equal to r, wherein, under the remaining bandwidth that has the multistage link on this selected optimal path can not the situation more than or equal to r, account for enterprising the robbing of this multistage link successively by the order that is from upstream to the downstream, all links that need seize on this selected optimal path have all been implemented preemptive priority, the head end router of optimal path that should be selected still regularly sends PATH message, this path each router of process when the PATH message that receives from upstream router, write down the routing information of EXPLICIT_ROUTE object in this PATH message.
In step b, the flow process of seizing and set up this label switched path for new label switched path foundation request enforcement is as follows:
The head end router compares the bandwidth on demand r of this label switched path and the head end router remaining bandwidth to this section of next hop router link of head end router, if this remaining bandwidth is more than or equal to r, the head end router will be according to based on the RSVP of the traffic engineering regulation to PATH message, produce PATH message, send to this next hop router-second hop router, enable the path simultaneously and set up PATH message retransmission timer, after second hop router is received PATH message, the routing information of the EXPLICIT_ROUTE object of record PATH message, and the address of deletion this second hop router itself (also is to delete itself address after router is received PATH message from the EXPLICIT_ROUTE object in the prior art from the EXPLICIT_ROUTE object of PATH message, only do not write down above-mentioned routing information, and there is comparison back of the present invention, so write down), check that then (for existing multi-protocol label switching traffic engineering, this bandwidth on demand r is the specific priority bandwidth of setting up of this label switched path for the bandwidth on demand r of this label switched path in the PATH message; Traffic engineering for the differential service sensitivity that is about to dispose, this bandwidth on demand r is the bandwidth on demand of certain TE-Class), the remaining bandwidth of this section link of linking to each other to next hop router-Di three hop routers of this second hop router of this bandwidth on demand r and this second hop router relatively, if this remaining bandwidth is more than or equal to r, this second hop router will be according to based on the RSVP of the traffic engineering regulation to PATH message so, produce a new PATH message, then it is issued the next hop router-the forth jump router that indicates in the EXPLICIT_ROUTE object, the rest may be inferred, if the remaining bandwidth of all links in the selected optimal path is all more than or equal to r, PATH message will arrive the last-hop Router in the EXPLICIT_ROUTE object always, it is terminal router, terminal router comprises the RESV message (RESV message also is the content in the existing protocol) that the label that label switched path uses is set up in request through returning to the head end router from each router of swimming over to the upstream down on this selected optimal path, when the head end router received that RESV message and this message comprise request and sets up the label that label switched path uses, PATH message retransmission timer was set up in the forbidding path; If should select the remaining bandwidth of certain link in the optimal path less than r, (the top router is different with the head end router for the top router of this link, the head end router is the starting point of whole label switched path, and the top router is the starting point of certain link) just no longer (the tail end router is different with terminal router to the tail end router of this link, terminal router is the terminal point of whole label switched path, and the tail end router is the terminal point of certain link) a PATH message, but along returning PATH ERR message (PATH ERR message also is existing protocol message) to the head end router from the direction of swimming over to the upstream down, simultaneously, this top router is implemented to seize according to the label switched path of formula 1 on this link, the label switched path information that to be seized from this label switched path delete in the routing information table that writes down the all-router of process, upgrade simultaneously this label switched path the not bandwidth reserved information in the all-router traffic engineering database (this database is existing) of process about link, seize finish after, the routing information of the label switched path that is requested to set up also from this label switched path that is requested to set up delete in the routing information table that writes down the all-router of process.As long as one section link needs LSP to seize because have wherein, the head end router just can not receive RESV message, and it still can send out PATH message, and intermediate router still can write down the routing information of the EXPLICIT_ROUTE object of PATH message when receiving PATH message next time.
The value that PATH message retransmission timer is set up in this path preferably is provided with to such an extent that to seize the used time than the label switched path that carries out formula 1 on a link long slightly, like this, when a PATH message arrives at certain section needs and seizes the top router of link of link bandwidth, this top router carries out described link bandwidth and seizes on this link, when next PATH message arrives at this top router, the preemptive priority of this section link has been accused and has been finished, PATH message will arrive the router that next place need seize link bandwidth, can reduce the path like this and set up the number of times that PATH message retransmits.To set up PATH message retransmission timer overtime when the path, but when the head end router is not received the RESV message that comprises the label that label switched path that request sets up uses (illustrating that finishing a LSP on the link seizes), the head end router is attempted sending PATH message along this optimal path once more, enable the path simultaneously and set up PATH message retransmission timer, in setting up the timing of PATH message retransmission timer, the path waits for the RESV message that comprises the label that label switched path that request sets up uses once more, when receiving this message, then requested label switched path is set up process and is finished; If timer expiry and do not receive this message, the head end router is attempted sending PATH message along this optimal path for the third time, enable the path simultaneously and set up PATH message retransmission timer, and the like, till the RESV message of the label of receiving the label switched path use that comprises request foundation, forbid the path and set up PATH message retransmission timer this moment, and the foundation declaration of the label switched path of this request finishes.
Now method of the present invention is described in detail as follows:
The first step: in the time will setting up the LSP of a certain priority between head end router and terminal router, at this moment the head end router will be analyzed whole network topology, for this requested LSP selects the route with minimum preemption cost.The MPLS network moves IGRP (IGP) expansion usually, as Open Shortest Path First traffic engineering (OSPF-TE) or intermediate system-intermediate system traffic engineering (IS-IS TE).For existing MPLS-TE, by the not bandwidth reserved (UB) of each priority on every link of they issues; For the DS-TE that be about to dispose, by the not bandwidth reserved (UB) of each TE-Class on every link of they issues, these bandwidth information be stored in router traffic engineering database (TED) and regularly or trigger and upgrade.Like this, the head end router stores has the topological diagram of whole network, comprises the bandwidth usage of every link.It at first wipes out the link of UB less than the bandwidth r of the LSP that is requested to set up from network topology.In the rest network topology, the remaining bandwidth that uses every link is as route metric, and operation CSPF is for this LSP seeks an optimal path.If there are many such paths, select an optimal path of crossing over the minimum path of jumping figure as this LSP.
Second step: found an optimal path for requested LSP in the first step, remaining issues is exactly how to set up a strip label switching path for this LSP along this optimal path, makes the preemption cost of setting up this strip label switching path for minimum simultaneously.The head end router compares the bandwidth on demand r of this label switched path and the head end router remaining bandwidth to this section of next hop router link of head end router, if this remaining bandwidth greater than or etc. r, the head end router will be according to based on the RSVP of the traffic engineering regulation to PATH message, produce PATH message, send to this next hop router, enable the path simultaneously and set up PATH message retransmission timer (software timer that the head end router increases newly among the present invention can be realized by software).Suppose that the head end router is R1, destination node is R5, and the optimal path that finds in the first step is R1-R2-R3-R4-R5, and this path is made of 4 sections links such as R1-R2, R2-R3, R3-R4, R4-R5.Suppose head end router R1 to the remaining bandwidth of first section link of next-hop node R2 more than or equal to requested bandwidth r, so PATH message that R1 sets up to the R2 transmit path, after next hop router R2 receives PATH message, the routing information of the EXPLICIT_ROUTE object in the PATH message is recorded in (this point is unlike the prior art in the LSP routing information table of this router, prior art is not preserved these routing informations), the routing information of EXPLICIT_ROUTE object be in the PATH message format should be meaningful, this routing information contains the address information of this request LSP through the all-router in this router downstream, R2 receives and the routing information of the EXPLICIT_ROUTE object that writes down is R2-R3-R4-R5 in this example, and from the EXPLICIT_ROUTE object this node address of deletion.Whether the remaining bandwidth of judging this section link between the R2-R3 then is more than or equal to r.If, on this section of R2-R3 link, do not need to implement preemptive priority, R2 will be according to the regulation of RSVP-TE agreement to PATH message so, produce a new PATH message, then this new PATH message is issued the next hop router of indicating in the routing information in the EXPLICIT_ROUTE object, i.e. R3.After R3 receives PATH message, the routing information of the EXPLICIT_ROUTE object in the PATH message is recorded in the LSP routing information table of R3, this routing information contains the information of this request LSP through the all-router in R3 downstream, be R3-R4-R5 in this example, and deletion oneself from the EXPLICIT_ROUTE object, whether the remaining bandwidth of judging link between the R3-R4 then is more than or equal to the bandwidth r of the LSP that is requested to set up.If the remaining bandwidth of link R3-R4 is more than or equal to the bandwidth r of the LSP that is requested to set up, on this link, do not need to seize, R3 will be according to the regulation of RSVP-TE agreement to PATH message so, produce a new PATH message, then this new PATH message is issued the next hop router R4 that indicates in the routing information in the EXPLICIT_ROUTE object, R4 also repeats the action of R2, R3, till reaching last router (being terminal router R5).Discuss above all is that the remaining bandwidth of hypothesis each section link on this selected optimal path is all more than or equal to the situation of the bandwidth r of the LSP that is requested to set up.If the remaining bandwidth of a certain section on this selected optimal path of discussion or a few sections links is less than the situation of the bandwidth r of the LSP that is requested to set up below.The remaining bandwidth of supposing this section of R2-R3 link is less than r, and R2 returns a PATH ERR message with regard to the hop router R1 that makes progress, and R2 implements preemptive priority to this section of R2-R3 link according to formula 1 then:
Y=α p+ ε/| (b-r) |+δ n+ θ b---formula 1
In the formula 1, p represents to be seized the priority (by formula 1 as can be known, priority is more little, and the p value is big more, and the Y value is also big more, shows that the priority of being seized LSP is the smaller the better) of LSP; B is seized the bandwidth of LSP, 1/| (b-r) | show that the bandwidth of being seized LSP is with the matching degree of required bandwidth r (b and r are approaching more, then mate well more, 1/| (b-r) | big more, the Y value is also big more); N represent those links that label switched path to be seized is positioned at this router downstream, link top with should selected optimal path share and the link remaining bandwidth less than the number of links of the bandwidth r that is requested label switched path.Suppose that the path that is requested to set up is LSP1, it is through R1-R2-R3-R4-R5, and judging on this section of R2-R3 link needs preemptive priority.Two LSP of LSP2 and LSP3 on this section of R2-R3 link, have been moved.Because according to above-mentioned method, when LSP2 and LSP3 when setting up, preserve LSP2 and LSP3 all downstream node information in the routing information table of router R2, also will pass through R2-R3-R4-R5-R6 such as LSP2 through R2 ..., and LSP3 also will pass through R2-R3-R6-R7-...The link that LSP2 also will pass through is shared R2-R3, R3-R4, these 3 sections links of R4-R5 with this selected optimal path R1-R2-R3-R4-R5, and the remaining bandwidth of supposing these 3 sections links is all less than r, so n=3; And for LSP3, also the link that will pass through is only shared this section of R2-R3 link with this selected optimal path, and the remaining bandwidth of this section link does not satisfy r (promptly less than r), so n=1.By the LSP that seized with should selected optimal path share and can not to satisfy the follow-up number of links of bandwidth on demand r big more, then n is big more, Y is also big more; The purpose that increases last θ b item in the formula is: when the LSP more than 2 have identical p, n and | during the b-r| value, the LSP of higher bandwidth has the bigger value of seizing Y, will pay the utmost attention to and be seized.Can avoid having the LSP more than two of the identical value of seizing Y like this, the LSP that is arranged in the less bandwidth of front is at first seized, because bandwidth is not enough, also need seize the situation of another LSP, thereby reduces preemption cost and heavy-route number.Factor alpha, ε, δ, θ are positioned at this number of links of seizing the node downstream for the priority of being seized LSP, matching degree, the LSP that seized of bandwidth and the bandwidth r of the LSP that is requested to set up that seized LSP share selected path, are seized the weight that the bandwidth of LSP is distributed, and these coefficients can be by operator's dynamic adjustments according to actual needs.For all LSP that moved on this link, by formula 1 calculate the Y value one by one, the LSP that is seized according to the Y value selective sequential from big to small of every LSP is up to the bandwidth of the selected LSP that seizes with more than or equal to r.To delete in the routing information of routing information from the router of all processes of selected these LSP that seize then (such as having selected to seize above-mentioned LSP2, the path of LSP2 is ... R2-R3-R4-R5-R6 ... just with the routing information of the selected label switching path LSP of seizing 2 from ... R2-R3-R4-R5-R6 ... the routing information table in delete, upgrade the not bandwidth reserved UB of all affected links in the network simultaneously.If the remaining bandwidth of all links in the optimal path that should select is all more than or equal to r, PATH message will arrive terminal router indicated in the LSP routing information of EXPLICIT_ROUTE object always, terminal router is a hop router loopback RESV message (RESV message also is one of message of existing protocol RSVP-TE regulation) upwards, and this RESV message comprises the label to next jumping that requested LSP uses.This receives that the router of RESV message changes the label value in the RESV message, and what equally this RESV message is transmitted back to a hop router this on goes up a hop router again, until arrival head end router.When head end router (R1 in this example) received that comprising requested LSP uses the RESV message of label, PATH message retransmission timer was set up in the forbidding path, is indicated as this requested LSP and has successfully set up a strip label switching path.If the remaining bandwidth of on optimal path that should be selected a certain section or a few sections links is less than the bandwidth r of requested LSP, because the low priority LSP that those nodes that need implement preemptive priority need certain hour to finish not satisfying the link that is requested bandwidth seizes, if the preemptive priority of multistage link has taken place on this label switched path, PATH message retransmission timer is set up in the path will be overtime.The head end router should be attempted setting up PATH message for this request LSP along set path transmit path once more, because the path is set up PATH message retransmission timer and is provided with longlyer slightly than the time of seizing of link, so when PATH message is delivered to last PATH message when arriving the router that needs to implement to seize again, this router has been finished the preemptive priority (PATH message retransmission timer is provided with too short meeting and causes PATH message frequently to send, and has increased the link protocol expense) to link.If should also have the remaining bandwidth of link can not satisfy r on the selected path, the top router of this link continues this link is implemented preemptive priority, when PATH message is built through after repeatedly retransmitting in the path of the LSP that is requested to set up, at this moment whole release busy bandwidth of being seized along each section link in path of low priority LSP, finally, all links along set path all have remaining bandwidth to admit the bandwidth r of the LSP that is requested to set up.Head end router transmit path is set up PATH message each time, all will trigger the low priority of one section link and seize.
In the present invention, each weight coefficient α, the ε in the formula 1, δ, θ can adjust and configuration flexibly according to the actual needs of operator.Selected during as important criteria as priority p, operator can increase factor alpha, and the LSP of low priority may at first be selected as seizing object; Selected during when the matching degree of bandwidth b and r as important criteria, can increase coefficient ε, so, the LSP (bandwidth waste minimum) with best bandwidth adaptation preferentially is chosen as and seizes object; When being seized the LSP number as important criteria, can increase coefficient δ, like this, those LSP that cross over multistage link preferentially are chosen as and seize object, thereby reduce the LSP number removed, promptly minimize the LSP number of heavy-route.
Based on prior art, the routing information of the LSP that network intermediate router and not knowing has been set up, i.e. the router situation of each LSP passing through network.The present invention is by changing the flow process of setting up of label switched path, allow intermediate router write down the routing information of EXPLICIT_ROUTE target in the PATH message, like this, just there is way to seize and select optimal path and shares the more LSP of multilink, thereby minimize the LSP number of heavy-route.The time set of the timer that the present invention adopts can be adjusted according to network size and loading condition.Do not need existing RSVP-TE agreement is done any change, network has good backwards compatibility and adaptability.
Method of the present invention both had been applicable to the MPLS-TE of widespread deployment, also was applicable to the DS-TE network environment that is about to deployment.
Claims (2)
1. label exchange route setting method that minimizes the path preemption cost in multi-protocol label switching network comprises:
Step a. is in the time will setting up the label switched path of a certain priority between head end router and terminal router, wipe out in the multi-protocol label switching network topology not bandwidth reserved less than the link of the bandwidth r of the label switched path that is requested to set up, in the rest network topology, the head end router uses the remaining bandwidth of every link as route metric, operation is based on the shortest-path first algorithm (CSPF) of constraint, it is dijkstra's algorithm, for this label switched path is sought optimal path, if there are many such paths, select an optimal path of crossing over the minimum path of jumping figure as this label switched path;
Step b. is for every link on the optimal path selected in step a, analyze its remaining bandwidth whether more than or equal to the bandwidth r of the label switched path that is requested to set up, if the remaining bandwidth of all links is all more than or equal to the bandwidth r of the label switched path that is requested to set up on this optimal path, the head end router uses based on the RSVP of traffic engineering and sets up label switched path for the label switched path of this request, and upgrade this path what write down in each router of process is the not bandwidth reserved of that section link of starting point with this router, this path each router of process when the PATH message that receives from upstream router, write down the routing information of EXPLICIT_ROUTE object in this PATH message, i.e. each jumping in this router downstream in this label switched path, when next round has new label switched path request comes and needs under the situation of bandwidth-hogging resource, the routing information of these records is used for the degree that label switched path that more every label switched path of having set up is requested to set up with this next round is shared same link; If the remaining bandwidth of one or more snippets link of existence can not be more than or equal to r on the optimal path that should select, need contrast the low label switched path implementation of this priority that is requested label switched path on these links seizes, seizing by this link top router on the such link of certain bar carried out, concerning on it arbitrary wait to seize the label switched path, seize formula and be:
Y=α p+ ε/| (b-r) |+δ n+ θ b---formula 1
In the formula 1, p represents to wait to seize the priority of label switched path, b is a bandwidth of waiting to seize label switched path, 1/| (b-r) | the expression Bandwidth adaptation, n represents that label switched path to be seized is positioned at those links in this router downstream, link top and should selectes optimal path shared, and the link remaining bandwidth is less than the number of links of the bandwidth r that is requested label switched path, factor alpha, ε, δ, θ is the p in the formula 1,1/| (b-r) |, n, the weight that b distributed, remain to be seized label switched path for the institute on this link that need seize, by formula 1 calculate the Y value one by one, the label switched path of being seized according to Y value selective sequential from big to small in every strip label switching path then, up to the bandwidth of the selected label switched path of seizing with more than or equal to r, for existing the remaining bandwidth of multistage link can not be on this selected optimal path more than or equal to the situation of r, account for enterprising the robbing of this multistage link successively according to the order that is from upstream to the downstream, all links that need seize on this selected optimal path have all been implemented preemptive priority, the head end router of optimal path that should be selected still regularly sends PATH message, this path each router of process when the PATH message that receives from upstream router, write down the routing information of EXPLICIT_ROUTE object in this PATH message.
2. according to the process of claim 1 wherein in step b, the flow process of seizing and set up this label switched path for new label switched path foundation request enforcement is as follows:
B1. the head end router compares the bandwidth on demand r of this label switched path and the head end router remaining bandwidth to this section of next hop router link of head end router, if this remaining bandwidth is more than or equal to r, the head end router will be according to based on the RSVP of the traffic engineering regulation to PATH message, produce PATH message, send to the next hop router of head end router, i.e. second hop router, enable the path simultaneously and set up PATH message retransmission timer, after second hop router is received PATH message, the routing information of the EXPLICIT_ROUTE object of record PATH message, and the address of from the EXPLICIT_ROUTE object of PATH message, deleting this second hop router itself, check the bandwidth on demand r of this label switched path in the PATH message then, the remaining bandwidth of this section link of linking to each other to next hop router-Di three hop routers of second hop router of this bandwidth on demand r and this second hop router relatively, if this remaining bandwidth is more than or equal to r, second hop router will be according to based on the RSVP of the traffic engineering regulation to PATH message so, produce a new PATH message, then it is issued next hop router-Di three hop routers of this second hop router that indicates in the EXPLICIT_ROUTE object, the rest may be inferred, if the remaining bandwidth of all links in the selected optimal path is all more than or equal to r, PATH message will arrive the last-hop Router in the EXPLICIT_ROUTE object always, it is terminal router, terminal router comprises request and sets up the RESV message that label switched path uses label through returning to the head end router from each router of swimming over to the upstream down on this selected optimal path, receive RESV message and this message when the head end router and comprise when being requested to set up the label that label switched path uses that PATH message retransmission timer is set up in the forbidding path; If should select the remaining bandwidth of certain section link in the optimal path less than r, the top router of this link is just no longer sent out PATH message to the tail end router of this link, but along returning PATH ERR message to the head end router from the direction of swimming over to the upstream down, simultaneously, this top router is implemented to seize according to the label switched path of formula 1 on this link, the label switched path information that to be seized from this label switched path delete in the routing information table that writes down the all-router of process, upgrade simultaneously this label switched path the not bandwidth reserved information in the all-router traffic engineering database of process about link, seize finish after, the routing information of the label switched path that is requested to set up also from this label switched path that is requested to set up delete in the routing information table that writes down the all-router of process;
B2. this path value of setting up PATH message retransmission timer preferably is provided with to such an extent that to seize the time that is spent than the label switched path that carries out formula 1 on one section link long, like this, when a PATH message arrives at certain section needs and seizes the top router of link of link bandwidth, this top router carries out described link bandwidth and seizes on this link, when next PATH message arrives at this top router, the preemptive priority of this section link has been accused and has been finished, PATH message will arrive the router that next place need seize link bandwidth, can reduce the path like this and set up the number of times that PATH message retransmits, to set up PATH message retransmission timer overtime when the path, but when the head end router is not received the RESV message of the label that comprises the label switched path use of asking foundation, the head end router is attempted sending PATH message along this optimal path once more, enable the path simultaneously and set up PATH message retransmission timer, in setting up the timing of PATH message retransmission timer, the path waits for the RESV message that comprises the label that label switched path that request sets up uses once more, when receiving this message, requested label switched path is set up process and is finished, if timer expiry and do not receive this message, the head end router is attempted sending PATH message along this optimal path for the third time, enable the path simultaneously and set up PATH message retransmission timer, and the like, till the RESV message of the label of receiving the label switched path use that comprises request foundation, forbid the path and set up PATH message retransmission timer this moment, and the foundation declaration of the label switched path of this request finishes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610141436A CN100596102C (en) | 2006-09-29 | 2006-09-29 | Method for establishing label switched path of minimized path preemption cost |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610141436A CN100596102C (en) | 2006-09-29 | 2006-09-29 | Method for establishing label switched path of minimized path preemption cost |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101155131A true CN101155131A (en) | 2008-04-02 |
| CN100596102C CN100596102C (en) | 2010-03-24 |
Family
ID=39256577
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200610141436A Active CN100596102C (en) | 2006-09-29 | 2006-09-29 | Method for establishing label switched path of minimized path preemption cost |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100596102C (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102271294A (en) * | 2011-04-29 | 2011-12-07 | 东北大学 | Load balance-based unicast share multi-layer protection method in optical fiber network |
| CN102333250A (en) * | 2011-10-26 | 2012-01-25 | 西安电子科技大学 | Low-congestion communication method and router for realizing shared path transmission of optical network on chip |
| CN102487352A (en) * | 2010-12-02 | 2012-06-06 | 中兴通讯股份有限公司 | Service distributing method and device |
| CN102523156A (en) * | 2011-12-09 | 2012-06-27 | 北京邮电大学 | Routing method based on hop count constraint |
| CN102904808A (en) * | 2011-07-25 | 2013-01-30 | 中兴通讯股份有限公司 | Method and system for establishing label switch path of resource reservation-protocol traffic engineering |
| CN101741715B (en) * | 2008-11-20 | 2013-03-20 | 华为技术有限公司 | Method for sending message, access routing node device and data cache system |
| WO2013037319A1 (en) * | 2011-09-16 | 2013-03-21 | 华为技术有限公司 | Bandwidth resource management method, device and system for label switched path |
| CN103139849A (en) * | 2011-12-05 | 2013-06-05 | 中国移动通信集团公司 | Quality of service (QoS) business execution method in multi-network cooperation and advanced function (AF), policy and charging rules function (PCRF) |
| CN103346979A (en) * | 2013-06-21 | 2013-10-09 | 杭州华三通信技术有限公司 | Flow distribution method and equipment in SPBM network |
| CN103828311A (en) * | 2013-12-16 | 2014-05-28 | 华为技术有限公司 | Rerouting sequence programming method and system |
| CN104009915A (en) * | 2014-06-09 | 2014-08-27 | 北京邮电大学 | Novel routing algorithm for simplifying complex network through bandwidth allocation |
| CN104270313A (en) * | 2014-09-29 | 2015-01-07 | 清华大学 | Method for adjusting utilization rate of network link |
| CN102273146B (en) * | 2008-04-28 | 2015-03-11 | 华为技术有限公司 | Transparent bypass and associated mechanisms |
| CN101610205B (en) * | 2008-06-20 | 2015-04-08 | 华为技术有限公司 | Method for obtaining path, transmitting equipment, receiving equipment and network system |
| WO2015149676A1 (en) * | 2014-04-03 | 2015-10-08 | Hangzhou H3C Technologies Co., Ltd. | Reserving bandwidth in software defined network |
| WO2016095574A1 (en) * | 2014-12-16 | 2016-06-23 | 上海斐讯数据通信技术有限公司 | Sdn-based mpls network control system and method |
| CN107454005A (en) * | 2016-05-31 | 2017-12-08 | 鸿富锦精密工业（深圳）有限公司 | Data network loads dynamic adjusting device and method |
| CN108023830A (en) * | 2017-11-21 | 2018-05-11 | 新华三技术有限公司 | flow transmission method and device |
| CN108199961A (en) * | 2018-01-17 | 2018-06-22 | 新华三技术有限公司 | A kind of flow transmission control method and device |
| CN108337189A (en) * | 2018-01-29 | 2018-07-27 | 新华三技术有限公司 | Bandwidth scheduling method and device |
| WO2018161754A1 (en) * | 2017-03-07 | 2018-09-13 | 中兴通讯股份有限公司 | Method and apparatus for recovering tunnel bandwidth resource |
| CN108600102A (en) * | 2018-03-29 | 2018-09-28 | 北京交通大学 | A kind of flexible data Transmission system in the contract network based on wisdom |
| US10574578B2 (en) | 2016-05-31 | 2020-02-25 | Nanning Fugui Precision Industrial Co., Ltd. | Dynamic network load regulating device and method |
| CN111694902A (en) * | 2019-03-11 | 2020-09-22 | 阿里巴巴集团控股有限公司 | Difference path identification method and device, electronic equipment and computer storage medium |
| CN113300960A (en) * | 2021-07-27 | 2021-08-24 | 南京中网卫星通信股份有限公司 | Delay deterministic transmission method based on routing scheduling and joint optimization |
| CN113347679A (en) * | 2021-05-31 | 2021-09-03 | 浙江大华技术股份有限公司 | Data transmission method and device, storage medium and electronic device |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100392649B1 (en) * | 2001-06-21 | 2003-07-25 | 에스케이 텔레콤주식회사 | Method for determining paths of traffic for Protection Switching and/or Fast Reroute in MPLS based data telecommunication network |
| KR100384045B1 (en) * | 2001-09-26 | 2003-05-14 | 한국과학기술원 | Explicit Routing Algorithms for MPLS Routers in IP Network |
| CN1287563C (en) * | 2002-06-24 | 2006-11-29 | 华为技术有限公司 | Method for preempting multi-procotol label switching flow controlling band width resource |
| US7643425B2 (en) * | 2004-07-23 | 2010-01-05 | Ericsson Ab | LSP path selection |
| CN1756233B (en) * | 2004-09-30 | 2010-07-28 | 富士通株式会社 | Route selecting method and apparatus in telecom network |
-
2006
- 2006-09-29 CN CN200610141436A patent/CN100596102C/en active Active
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102273146B (en) * | 2008-04-28 | 2015-03-11 | 华为技术有限公司 | Transparent bypass and associated mechanisms |
| CN101610205B (en) * | 2008-06-20 | 2015-04-08 | 华为技术有限公司 | Method for obtaining path, transmitting equipment, receiving equipment and network system |
| CN101741715B (en) * | 2008-11-20 | 2013-03-20 | 华为技术有限公司 | Method for sending message, access routing node device and data cache system |
| CN102487352A (en) * | 2010-12-02 | 2012-06-06 | 中兴通讯股份有限公司 | Service distributing method and device |
| CN102487352B (en) * | 2010-12-02 | 2014-09-10 | 中兴通讯股份有限公司 | Service distributing method and device |
| CN102271294B (en) * | 2011-04-29 | 2013-07-24 | 东北大学 | Load balance-based unicast share multi-layer protection method in optical fiber network |
| CN102271294A (en) * | 2011-04-29 | 2011-12-07 | 东北大学 | Load balance-based unicast share multi-layer protection method in optical fiber network |
| CN102904808A (en) * | 2011-07-25 | 2013-01-30 | 中兴通讯股份有限公司 | Method and system for establishing label switch path of resource reservation-protocol traffic engineering |
| CN102904808B (en) * | 2011-07-25 | 2017-08-25 | 中兴通讯股份有限公司 | Across the method for building up and system of resource reservation protocol flow engineering label switched path |
| WO2013037319A1 (en) * | 2011-09-16 | 2013-03-21 | 华为技术有限公司 | Bandwidth resource management method, device and system for label switched path |
| CN102333250B (en) * | 2011-10-26 | 2014-03-12 | 西安电子科技大学 | Low-congestion communication method and router for realizing shared path transmission of optical network on chip |
| CN102333250A (en) * | 2011-10-26 | 2012-01-25 | 西安电子科技大学 | Low-congestion communication method and router for realizing shared path transmission of optical network on chip |
| CN103139849A (en) * | 2011-12-05 | 2013-06-05 | 中国移动通信集团公司 | Quality of service (QoS) business execution method in multi-network cooperation and advanced function (AF), policy and charging rules function (PCRF) |
| CN102523156B (en) * | 2011-12-09 | 2014-11-26 | 北京邮电大学 | Routing method based on hop count constraint |
| CN102523156A (en) * | 2011-12-09 | 2012-06-27 | 北京邮电大学 | Routing method based on hop count constraint |
| CN103346979A (en) * | 2013-06-21 | 2013-10-09 | 杭州华三通信技术有限公司 | Flow distribution method and equipment in SPBM network |
| CN103346979B (en) * | 2013-06-21 | 2017-02-08 | 杭州华三通信技术有限公司 | Flow distribution method and equipment in SPBM network |
| CN103828311B (en) * | 2013-12-16 | 2016-09-28 | 华为技术有限公司 | A kind of heavy-route sequential program(me) method and system |
| CN103828311A (en) * | 2013-12-16 | 2014-05-28 | 华为技术有限公司 | Rerouting sequence programming method and system |
| US10015079B2 (en) | 2013-12-16 | 2018-07-03 | Huawei Technologies Co., Ltd. | Rerouting sequence planning method and system |
| WO2015149676A1 (en) * | 2014-04-03 | 2015-10-08 | Hangzhou H3C Technologies Co., Ltd. | Reserving bandwidth in software defined network |
| CN104009915A (en) * | 2014-06-09 | 2014-08-27 | 北京邮电大学 | Novel routing algorithm for simplifying complex network through bandwidth allocation |
| CN104009915B (en) * | 2014-06-09 | 2017-12-01 | 北京邮电大学 | A kind of method for routing for simplifying network using bandwidth allocation |
| CN104270313A (en) * | 2014-09-29 | 2015-01-07 | 清华大学 | Method for adjusting utilization rate of network link |
| CN104270313B (en) * | 2014-09-29 | 2017-05-17 | 清华大学 | Method for adjusting utilization rate of network link |
| WO2016095574A1 (en) * | 2014-12-16 | 2016-06-23 | 上海斐讯数据通信技术有限公司 | Sdn-based mpls network control system and method |
| CN107454005A (en) * | 2016-05-31 | 2017-12-08 | 鸿富锦精密工业（深圳）有限公司 | Data network loads dynamic adjusting device and method |
| CN107454005B (en) * | 2016-05-31 | 2020-04-24 | 南宁富桂精密工业有限公司 | Data network load dynamic adjusting device and method |
| US10574578B2 (en) | 2016-05-31 | 2020-02-25 | Nanning Fugui Precision Industrial Co., Ltd. | Dynamic network load regulating device and method |
| WO2018161754A1 (en) * | 2017-03-07 | 2018-09-13 | 中兴通讯股份有限公司 | Method and apparatus for recovering tunnel bandwidth resource |
| CN108574646A (en) * | 2017-03-07 | 2018-09-25 | 中兴通讯股份有限公司 | A kind of tunnel bandwidth recovery method as resource and device |
| CN108023830A (en) * | 2017-11-21 | 2018-05-11 | 新华三技术有限公司 | flow transmission method and device |
| CN108199961B (en) * | 2018-01-17 | 2021-09-21 | 新华三技术有限公司 | Flow forwarding control method and device |
| CN108199961A (en) * | 2018-01-17 | 2018-06-22 | 新华三技术有限公司 | A kind of flow transmission control method and device |
| CN108337189A (en) * | 2018-01-29 | 2018-07-27 | 新华三技术有限公司 | Bandwidth scheduling method and device |
| CN108337189B (en) * | 2018-01-29 | 2021-11-05 | 新华三技术有限公司 | Bandwidth scheduling method and device |
| CN108600102A (en) * | 2018-03-29 | 2018-09-28 | 北京交通大学 | A kind of flexible data Transmission system in the contract network based on wisdom |
| CN108600102B (en) * | 2018-03-29 | 2020-07-17 | 北京交通大学 | Flexible data transmission system based on intelligent cooperative network |
| CN111694902A (en) * | 2019-03-11 | 2020-09-22 | 阿里巴巴集团控股有限公司 | Difference path identification method and device, electronic equipment and computer storage medium |
| CN111694902B (en) * | 2019-03-11 | 2023-04-07 | 阿里巴巴集团控股有限公司 | Difference path identification method and device, electronic equipment and computer storage medium |
| CN113347679A (en) * | 2021-05-31 | 2021-09-03 | 浙江大华技术股份有限公司 | Data transmission method and device, storage medium and electronic device |
| CN113300960A (en) * | 2021-07-27 | 2021-08-24 | 南京中网卫星通信股份有限公司 | Delay deterministic transmission method based on routing scheduling and joint optimization |
| CN113300960B (en) * | 2021-07-27 | 2021-11-23 | 南京中网卫星通信股份有限公司 | Delay deterministic transmission method based on routing scheduling and joint optimization |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100596102C (en) | 2010-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100596102C (en) | Method for establishing label switched path of minimized path preemption cost | |
| EP2911348B1 (en) | Control device discovery in networks having separate control and forwarding devices | |
| US8325706B2 (en) | Hierarchical segmented label switched paths | |
| EP1395003B1 (en) | Constraint-based shortest path first method for dynamically switched optical transport networks | |
| US9584404B2 (en) | Rapid alternate paths for network destinations | |
| US6925061B2 (en) | Multi-constraint routing system and method | |
| US7406032B2 (en) | Bandwidth management for MPLS fast rerouting | |
| US8218445B2 (en) | Smart ethernet edge networking system | |
| CN100512209C (en) | Method for solving resource conflict in distribution route counting | |
| US20040042473A1 (en) | Method for providing QoS (quality of service) - guaranteeing multi-path and method for providing disjoint path using the same | |
| CN101741709A (en) | Method and system for establishing label switched path and network node | |
| JP4421978B2 (en) | Delay guarantee path setting system | |
| CN101127689B (en) | Method for minimizing priority seizing cost | |
| WO2003058868A2 (en) | Dynamic route selection for label switched paths in communication networks | |
| CN103493443B (en) | Subpath maintenance unit (SPME) is used to share grid protection for multiprotocol label switching (MPLS) | |
| US20120124238A1 (en) | Prioritization of routing information updates | |
| CN100484046C (en) | Soft preemption feedback | |
| CN105634941A (en) | Cross-domain path calculation method and device | |
| US7168044B1 (en) | Apparatus and method for automatic network connection provisioning | |
| CN114363250A (en) | Re-computation of multipaths in a network supporting segment routing | |
| US7443832B2 (en) | Device for determining switching paths in a label switched communication network in the presence of selection attributes | |
| CN102487352B (en) | Service distributing method and device | |
| CN111245724A (en) | SDN load balancing routing method based on virtual switch deployment | |
| Cho et al. | Multi-path constraint-based routing algorithms for MPLS traffic engineering | |
| CN102685004A (en) | Method for implementing traffic engineering in GMPLS/OBS (generalized multi-protocol label switching/optical burst switching) network |
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 |