CN104811393B - Multicast message duplication method, device and open flow controller - Google Patents

Abstract

The present invention provides a kind of multicast message duplication method, device and open flow controller, this method is applied to software defined network SDN, comprising: the link connection topological diagram of open flow switch OFSW in SDN is obtained from open flow controller OFC；Incidence relation according to link connection topological diagram and terminal user and multicast group determines OFSW for handling the flow table entry and group table clause of multicast group message；Determining flow table entry and group table clause are handed down to OFSW by OFC, wherein, OFSW carries out multicast replication processing to multicast group message according to flow table entry and group table clause, through the invention, it solves and carries out multicast message duplication by user on borde gateway BNG in the related technology, there are bandwidth wastes, the not high problem of resource utilization, and then reached reduction multicast replication as far as possible, bandwidth resources are saved, the effect of congestion is avoided.

Description

Multicast message duplication method, device and open flow controller

Technical field

The present invention relates to the communications fields, in particular to a kind of multicast message duplication method, device and open flows Controller (OpenFlow Controller, referred to as OFC).

Background technique

The current internet Internet(based on TCP/IP) by development in more than 40 years, huge success has been obtained, It is closely bound up with people, it has also become one of work, the essential infrastructure of studying and living.The internet of TCP/IP formula, The division of labor of " gateway/network equipment carries out simple process, and host side/side is given in complicated processing " at the beginning of because of its design and tissue Principle, form current internet architecture status: the application layer protocol of host computer side easily, can be repaired neatly Change and dispose, therefore application layer software is by leaps and bounds developed, therefore the function of application layer is greatly enriched；Therewith Form sharp contrast is network layer, although the design of network layer protocol is simple, scalability is not strong and is not easy to modify, Cause: on the one hand, many fatal loopholes that Internet level leaks out cruelly are difficult to be repaired and improved for a long time, such as network The forwarding strategy that management is difficult to dispose, network security problem is got worse, does one's best cannot provide the Service Quality of user demand Amount, multicast are difficult to dispose and apply；On the other hand, new agreement, new opplication are due to proposing that change requires and is difficult to network layer Accomplished, such as from IPv4 to IPv6 transition, ubiquitous mobility is increasingly presented in difficult, access device and heterogeneity is reliable to network Property and Differentiated Services ability is challenged, routing faces scalability problem, cloud computing and content point in the case of large scale network The application such as hair proposes new demand to forwarded efficiency, the father Vinton G.Cerf of TCP/IP also indicates that internet should be in net Doing better in terms of network safety and network reliability, (" safety and reliability is two most basic doors of Future Internet of marching toward Sill, otherwise this framework will be unable to survive ") etc..Therefore internet has currently formed that a kind of " application layer is flexible and changeable, all sorts of flowers are neat Put, network layer is stiff difficult to be become, is filled with flaws " difficult situation.Internet will solve current problem encountered and difficult situation, It needs to be inquired into, studied and reformed from the levels such as network architecture, control are profound, two Pius XIs could be met comprehensively The opportunity and huge challenge for recording new.

For how to solve current internet problem encountered and challenge, domestic and international research institution is from Generation Internet knot Structure level has carried out a large amount of positive explorations and research.The development that mainly experienced two stages, can divide the improvement of internet For two class modes, evolved is improved and revolutionary improvement.

For many years, for traditional IP quality of service guarantee, it is mobile support, high efficient and reliable and in terms of Exposing many problems, research field is all solved these problems targetedly and by the way of repairing using design respectively, once It was found that the weakness or mistake of the network of operation are just improved immediately, such as in conventional internet architecture, add new agreement With functional unit etc..The improved procedure of this " repairing -> finding the problem -> to modify again " is with existing internet TCP/IP system knot Based on structure, gradually evolution and development is carried out to existing network, is asked with the function and characteristic of adding new solve to face at present Topic, be a kind of Evolution(evolved) improved procedure.The advantage of this improved procedure is to be easy to dispose and implement, and has Has investment conducive to protecting in existing Internet Construction.But its defect is: (1) certain repairing is only in a small range Solve the problems, such as part；(2) existing improvement may introduce short-term yield and seeing for a long time then has destructive such as NAT, Huo Zheju Portion's income is to being integrated with destructiveness；(3) certain repairing may be not easy " be compatible with " future and continue to modify；(4) by repeatedly repairing It mends, internet becomes increasingly " thick and heavy ", and it is complicated, not flexible, have exceeded the simple architecture for designing Internet originally Ability to bear；(5) some intrinsic problems in conventional internet architecture are difficult to obtain the solution of essence.At present still Mainly gradually internet is improved in a manner of " evolution ".

Start within 2005, research field gradually forms another viewpoint, and only redesign network architecture could be from IP network problem encountered is fundamentally solved, and at present exactly internet architecture " Clean-Slate " (starting from scratch) Carry out the inning thoroughly changed comprehensively, give up existing internet architecture completely, design it is a kind of it is completely new, fusion is a variety of The next generation internet architecture of design object.This scheme, which is intended to fundamentally solve existing internet architecture, to be existed Various problems, be a kind of Revolution(revolutionary) improvement project.The advantage of this scheme is: (1) can get rid of Its constraint and frame are jumped out in the constraint of TPC/IP architecture, to solve internet for many years because leaving difficulty caused by architecture Topic；(2) internet again, comprehensively design, the problems of overall solution internet, overall arrangement internet Many new demands realization.But the defect of this scheme is: (1) since completely new network may not be able to be compatible with existing interconnection Net needs to replace the infrastructure of legacy network completely, therefore there is network deployment peace to slip over the problem of crossing；(2) how New architecture is established, and establishes that the problem of whether new architecture can solve current and future network faces is also deposited In great risk；(3) it needs to rebuild the experimental network for being suitble to completely new architecture, evolution cost is high.

In order to solve the problems, such as that current internet exists, realize that quickly, flexibly deployment, opening can compile to new network protocol Journey network is suggested, and open programmable network, which refers to, allows network research person and more than device manufacturer, on network devices into Row programs and manages its network architecture or network protocol.Open programmable formula thinking is the representativeness of revolutionary improvement project One of achievement may be summarized to be substantially: by multiple original functional networks and deposit, whole, complicated MAN/WAN network or network Equipment carries out cutting by function, for example, being divided into data forwarding part and logic control part or system core part and use Family funtion part etc..Interface between each section is open and standard.Based on this opening and standardized interface, each It part can be with self evolution and improvement without notifying or influencing another part other parts, such whole network or the network equipment It will realize independent, smooth evolution and improvement.Open programmable formula thinking facing challenges are: (1) network hierarchy needs to have Certain reasonability, science and scalability；(2) interface between science, expansible layering is defined；(3) control plane is such as Fruit takes centralized management mode, then needs to consider interdomain connection, scalability (such as expanding to the whole world).

In terms of the research of open programmable network, Berkeley(Berkeley) the Scott Shenker et al. of university mentions SDN(Software Defined Networking out, software defined network) technology, the Stamford Stanford() university The technologies such as OpenFlow are the representative achievements of network opening Journal of Sex Research.The hierarchical model of SDN/OpenFlow technology includes basis Three facility layer, network control layer and application layer levels.The infrastructure layer of SDN/OpenFlow network is by each forwarding device It constitutes, structure is simpler, no complicated for the router, interchanger and all kinds of gateways in the relatively current network of forwarding device Control Plane(control plane), main work be carry out data flow forwarding.The capital equipment of network control layer is net Network operating system (or SDN/OpenFlow controller), network operating system simultaneously forward more by standardized interface Equipment is controlled, instead of the script Network Management System even current independently of the control plane in individual device, Ke Yishi Existing network management and end to end data flow rule issue (the more forwarding devices i.e. on forward-path issue stream rule), together When network operating system by API(Application Programming Interface, application programming interface) with answer With layer interaction.Application layer is made of different application, using the network management and control that can call directly control layer by api interface Function processed.

In SDN/OpenFlow network, (i.e. forwarding is set for network control layer equipment (such as controller) and infrastructure layer equipment It is standby) between interacted (such as OpenFlow agreement) by IP address-based communication protocol message, between the network terminal and net Data traffic between network terminal and application server, between application server and application server passes through stream between forwarding device Table is forwarded, and the flow table of every stream is generated by controller and be handed down to forwarding device, and forwarding device forwards no hit Inquiry and generation that controller carries out flow table are given on the data message of the currently stored flow table of equipment is unified, forwarding device needs Waiting controller to issue new flow table just can be with forwarding data packets.

Fig. 1 be in the related technology multicast user be based on BNG scene networking schematic diagram, as shown in Figure 1, user multicast generally by Wideband network gateway (Broadband Network Gateway, referred to as BNG) equipment realizes that direct or indirect every user is multiple System.Certain BNG can be assisted by Access Node Control Protocol (Access Node Control Protocol, referred to as ANCP) The view control interchangers such as SW1, SW2, issue corresponding multicast service strategy (for example, multicast access control, multicast replication control, Multicast bandwidth control and multicast charging policy), but process is complicated, the fineness of control and flexibility are also inadequate, so practical Deployment is seldom.ANCP is by GSMPv3 general switch management protocol (General Switch Management Protocol, letter Referred to as RFC3292) it expands, continue to use the section sets of GSMPv3: as based on Adjacency Protocol Message Neighbours' discovery Port Management is based on based on the Topology Discovery of Port Up/Down Event Message The topological arrangement of Message.In addition it can also be disposed on respective switch and open Internet Group Management Protocol (Internet Group Management Protocol, referred to as IGMP) Snooping function, by interchanger, perception user is added Multicast group simultaneously executes multicast replication work.But this mode is higher to the control plane functional requirement of interchanger, and is not suitable for deployment On SDN Autonomous Domain network internal forward node.Therefore, the method generally used on user access network is in borde gateway BNG is upper carries out multicast message duplication by user, as shown in Figure 1, if the N number of user of the BNG gateway down-hanging, what network side came Flux of multicast will do it N parts of duplications on BNG, this has very big bandwidth wave for the interchanger in user access network Take, as user4-userN N-3 part flux of multicast can always access switch SW3 link band is occupied between SW5 and BNG always It is wide.

Therefore, in the related art, multicast message duplication is carried out by user on borde gateway BNG, there are bandwidth waste, The not high problem of resource utilization.

Summary of the invention

The present invention provides a kind of multicast message duplication method, device and open flow controllers, at least to solve phase In the technology of pass, multicast message duplication is carried out by user on borde gateway BNG, there are bandwidth waste, resource utilization is not high Problem.

According to an aspect of the invention, there is provided a kind of multicast message duplication method, is applied to software defined network Network SDN, comprising: the link connection topological diagram of open flow switch OFSW in the SDN is obtained from open flow controller OFC；According to Determine the OFSW for handling multicast group according to the incidence relation of the link connection topological diagram and terminal user and multicast group The flow table entry and group table clause of message；Determining the flow table entry and described group of table clause are handed down to by the OFC The OFSW, wherein the OFSW carries out multicast to the multicast group message according to the flow table entry and described group of table clause Replication processes.

Preferably, the association according to the link connection topological diagram and the terminal user and the multicast group Relationship determines that flow table entry and group table clause of the OFSW for handling multicast group message include: in terminal user's access authentication By rear, the SDN network is passed through for terminal user calculating one and is connect from network side node interface NNI to Customer Premises Network The forward-path of mouth UNI；Along the forward-path, according to the incidence relation of the terminal user and the multicast group, Gather the flux of multicast for determining terminal user's set on OFSW by the terminal user that the multicast group is added in traversal one by one Enter corresponding out to set of interfaces to interface；Institute is generated according to going out described in each OFSW on the forward-path to set of interfaces State the corresponding described group of table clause of OFSW and the flow table entry, wherein include association in the action lists of the flow table entry To the movement of described group of table clause.

Preferably, go out described in each OFSW on according to the forward-path corresponding to the set of interfaces generation OFSW Described group of table clause and the flow table entry after, further includes: receive the terminal user multicast change request；Foundation The multicast of the terminal user received changes request, along the forward-path, uses according to the updated terminal The incidence relation at family and the multicast group, updated on OFSW one by one described in enter to interface it is corresponding go out to set of interfaces；Foundation Each OFSW is updated described out to the corresponding flow table entry of the set of interfaces update OFSW on the forward-path With described group of table clause.

Preferably, corresponding to the set of interfaces generation OFSW according to going out described in each OFSW on the forward-path Described group of table clause and the flow table entry include: to judge whether the multicast resource of current OFSW on the forward-path is sufficient； If the determination result is NO, select the upper additional replication processes of a node in the current OFSW because of the OFSW group Broadcasting inadequate resource leads to the multicast message of current OFSW user's set of not reproducible forwarding, wherein the group additionally replicated Casting text passes through the current OFSW with unicast forwarding processing mode.

Preferably, it is selecting in the upper additional replication processes of a node of the current OFSW because of the current OFSW multicast Inadequate resource causes after the multicast message of current OFSW user's set of not reproducible forwarding, further includes: on described The group table clause that one node OFSW carries out multicast message replication processes and generates adds one to multiple buckets, wherein one Or the multicast group user set that can not handle of the multiple barrels of correspondence current OFSW is corresponding goes out to set of interfaces, it is one Or multiple buckets carry the operational motion for two layers of head of the multicast message to be revised as to unicast encapsulation.

According to another aspect of the present invention, a kind of multicast message replication processes device is provided, software defined network is applied to Network SDN, comprising: module is obtained, for obtaining the link of open flow switch OFSW in the SDN from open flow controller OFC Link topology；Determining module, for the incidence relation according to the link connection topological diagram and terminal user and multicast group Determine the OFSW for handling the flow table entry and group table clause of multicast group message；Module is issued, for it will determine described in Flow table entry and described group of table clause are handed down to the OFSW by the OFC, wherein the OFSW is according to the flow table entry Multicast replication processing is carried out to the multicast group message with described group of table clause.

Preferably, the determining module includes: computing unit, for being described after terminal user's access authentication passes through Terminal user calculates one and passes through the SDN network from network side node interface NNI to the forwarding road of Customer Premises Network interface UNI Diameter；Determination unit, for along the forward-path, according to the incidence relation of the terminal user and the multicast group, Gather the flux of multicast for determining terminal user's set on OFSW by the terminal user that the multicast group is added in traversal one by one Enter corresponding out to set of interfaces to interface；Generation unit, for according to described in each OFSW on the forward-path go out to Set of interfaces generates the corresponding described group of table clause of OFSW and the flow table entry, wherein the movement of the flow table entry It include the movement for being associated with described group of table clause in list.

Preferably, the determining module further include: receiving unit, the multicast for receiving the terminal user, which changes, asks It asks；First updating unit, for the multicast change request according to the terminal user received, along generation multicast change Terminal user corresponding to the forward-path, according to being associated with for the updated terminal user and the multicast group System, updated on OFSW one by one described in enter to interface it is corresponding go out to set of interfaces；Second updating unit, for turning according to described Send out that each OFSW on path is updated described to be gone out to set of interfaces to update the corresponding flow table entry of the OFSW and described Group table clause.

Preferably, the generation unit includes: judgment sub-unit, for judging the group of current OFSW on the forward-path Whether sufficient broadcast resource；Subelement is handled, for selecting in institute in the case where the judging result of the judgment sub-unit is no The upper additional replication processes of a node of current OFSW are stated because the current OFSW multicast resource deficiency leads to not reproducible forwarding The current OFSW user set multicast message, wherein the multicast message additionally replicated is logical with unicast forwarding processing mode Cross the OFSW.

Preferably, the generation unit further include: addition subelement, for carrying out a group casting for the upper node OFSW Literary replication processes and the group table clause that generates adds one to multiple buckets, wherein one or more of barrels of correspondence are described currently Multicast group user that OFSW can not be handled set is corresponding to be gone out to set of interfaces, and one or more of buckets are carried for will Two layers of head of the multicast message are revised as the operational motion of unicast encapsulation.

Also one side according to the present invention, provides a kind of open flow controller OFC, including dress described in any of the above embodiments It sets.

Through the invention, connected using the link for obtaining open flow switch OFSW in the SDN from open flow controller OFC Connect topological diagram；Incidence relation according to the link connection topological diagram and terminal user and multicast group determines that the OFSW is used In the flow table entry and group table clause of processing multicast group message；Determining the flow table entry and described group of table clause are passed through into institute It states OFC and is handed down to the OFSW, wherein the OFSW is according to the flow table entry and described group of table clause to the multicast group report Text carries out multicast replication processing, solves and carries out multicast message duplication by user on borde gateway BNG in the related technology, exists Bandwidth waste, the not high problem of resource utilization, and then reached reduction multicast replication as far as possible, bandwidth resources are saved, are kept away Exempt from the effect of congestion.

Detailed description of the invention

The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:

Fig. 1 is that multicast user is based on BNG scene networking schematic diagram in the related technology；

Fig. 2 is the flow chart of multicast message duplication method according to an embodiment of the present invention；

Fig. 3 is the structural block diagram of multicast message replication processes device according to an embodiment of the present invention；

Fig. 4 is the preferred structure block diagram of determining module 34 in multicast message replication processes device according to an embodiment of the present invention One；

Fig. 5 is the preferred structure block diagram of determining module 34 in multicast message replication processes device according to an embodiment of the present invention Two；

Fig. 6 is generation unit 46 in determining module 34 in multicast message replication processes device according to an embodiment of the present invention Preferred structure block diagram one；

Fig. 7 is generation unit 46 in determining module 34 in multicast message replication processes device according to an embodiment of the present invention Preferred structure block diagram two；

Fig. 8 is the structural block diagram of open flow controller according to an embodiment of the present invention；

Fig. 9 is the networking schematic diagram of the user multicast duplication under SDN framework according to an embodiment of the present invention；

Figure 10 is the flow chart of the user multicast clone method under SDN framework according to the preferred embodiment of the invention.

Specific embodiment

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and in combination with Examples.It should be noted that not conflicting In the case of, the features in the embodiments and the embodiments of the present application can be combined with each other.

A kind of multicast message duplication method is provided in the present embodiment, is applied to software defined network SDN, Fig. 2 It is the flow chart of multicast message duplication method according to an embodiment of the present invention, as shown in Fig. 2, the process includes following step It is rapid:

Step S202 obtains the link connection topology of open flow switch OFSW in the SDN from open flow controller OFC Figure；

Step S204, the incidence relation according to the link connection topological diagram and terminal user and multicast group determine OFSW For handling the flow table entry and group table clause of multicast group message；

The determining flow table entry and group table clause are handed down to OFSW by OFC by step S206, wherein the OFSW according to Multicast replication processing is carried out to multicast group message according to above-mentioned flow table entry and group table clause.

Through the above steps, the incidence relation according to the link connection topological diagram of SDN and terminal user and multicast group is true Fixed flow table entry and group table clause for carrying out multicast message processing, i.e., carry out multicast according to the forward-path of terminal user and turn The flow table entry and group table clause for sending out data, relative in the related technology on borde gateway BNG by user's progress multicast message Duplication, there are bandwidth waste, the not high problem of resource utilization, and then reached reduction multicast replication as far as possible, save band Wide resource avoids the effect of congestion.

Incidence relation according to link connection topological diagram and terminal user and multicast group determines OFSW for handling multicast When the flow table entry and group table clause of group message, one for can be a network unit for completing the operation is embedded to be answered With for example, authentication processing is carried out to the terminal user, in terminal user after receiving the online triggering message of terminal user After access authentication passes through, one is calculated for the terminal user and passes through above-mentioned SDN network from network side node interface (NNI to user The forward-path of side network interface UNI；Along forward-path, according to the incidence relation of terminal user and multicast group, in OFSW one by one The upper flux of multicast by traversing the determining terminal user's set of terminal user's set that multicast group is added enters corresponding out to interface To set of interfaces；Corresponding group of table clause of OFSW and flow table item are generated according to going out to set of interfaces for each OFSW on forward-path Mesh, wherein it should be noted that when generating flow table entry, it can be according to multicast group destination address key assignments (multicast destination IP And/or multicast purpose MAC) and/or flux of multicast enter and generated to interface, include being associated with group in the action lists of the flow table entry The movement of table clause.

It include the movement for being associated with group table clause in the action lists of flow table entry.The SDN network topology provided by OFC Information is terminal user's calculating unicast forwarding route, and is each OFSW dimension along terminal user's unicast forwarding route on SDN Relevant group of table clause of user multicast is protected, when flux of multicast entrance, the user of designated multicast group is collected using on specified OFSW Conjunction is mapped to outgoing interface collection and closes, thus generation group table clause, by ensure that group for the binding of unicast forwarding and multicast forwarding The barrelage amount of table clause is minimum, to reduce the duplication of multicast message, effectively saves the bandwidth of multicast link.

Terminal user can be added or withdraw from multicast group in different situations, i.e. the multicast state of terminal user can be sent out Changing, and need to carry out corresponding update, example to the link for carrying out multicast according to the multicast state occurred when changing Such as, according on forward-path each OFSW go out to set of interfaces generate corresponding group of table clause of OFSW and flow table entry it Afterwards, the multicast change request of terminal user is received；According to the multicast change request of the terminal user received, along forwarding road Diameter, according to the incidence relation of updated terminal user and multicast group, updated on OFSW one by one into interface it is corresponding go out to Set of interfaces；It is updated out to the corresponding flow table entry of set of interfaces update OFSW and group according to each OFSW on forward-path Table clause, i.e., when the multicast relations of terminal user change, at corresponding group of terminal user of generation multicast relations change The each OFSW for broadcasting chain road carries out corresponding flow table entry and organizes the update of table clause (for example, convection current table clause and/or group table Entry at least one performs the following operation: creation, is deleted at modification), to go on smoothly multicast according to new multicast relations.

It should be noted that generating corresponding group of table of OFSW according to going out to set of interfaces for each OFSW on forward-path When entry and flow table entry, the situation of multicast resource deficiency may be met, for example, the entry of group table, or the bucket of group table.Cause This, can first judge whether the multicast resource of current OFSW on the forward-path is sufficient；If the determination result is NO, i.e., Judging result is to select in a upper node of current OFSW additionally at duplication in the insufficient situation of multicast resource on current OFSW Reason is because OFSW multicast resource deficiency leads to the multicast message of the current OFSW user set of not reproducible forwarding, wherein additional The multicast message of duplication passes through the current OFSW with unicast forwarding processing mode.

In addition, in selection in the upper additional replication processes of a node of current OFSW because current OFSW multicast resource deficiency is led After the multicast message of current OFSW user set for causing not reproducible forwarding, further includes: carry out group for a node OFSW on this The group table clause broadcasting literary replication processes and generating adds one to multiple buckets, wherein the one or more bucket is corresponding current Multicast group user that OFSW can not be handled set is corresponding to be gone out to set of interfaces, and one or more buckets are carried for by multicast Two layers of head of message are revised as the operational motion of unicast encapsulation.

A kind of multicast message replication processes device is additionally provided in the present embodiment, and the device is for realizing above-described embodiment And preferred embodiment, the descriptions that have already been made will not be repeated.As used below, term " module " may be implemented to make a reservation for The combination of the software and/or hardware of function.It is hard although device described in following embodiment is preferably realized with software The realization of the combination of part or software and hardware is also that may and be contemplated.

Fig. 3 is the structural block diagram of multicast message replication processes device according to an embodiment of the present invention, as shown in figure 3, the dress Set and be applied to software defined network SDN, including obtain and module 32, determining module 34 and issue module 36, below to the device into Row explanation.

Module 32 is obtained, the link connection for obtaining open flow switch OFSW in SDN from open flow controller OFC is opened up Flutter figure；Determining module 34 is connected to above-mentioned acquisition module 32, for according to link connection topological diagram and terminal user and group The incidence relation for broadcasting group determines OFSW for handling the flow table entry and group table clause of multicast group message；Module 36 is issued, is connected To above-mentioned determining module 34, flow table entry and group table clause for will determine are handed down to OFSW by OFC, wherein OFSW according to Multicast replication processing is carried out to multicast group message according to flow table entry and group table clause.

Fig. 4 is the preferred structure block diagram of determining module 34 in multicast message replication processes device according to an embodiment of the present invention One, as shown in figure 4, the determining module 34 includes computing unit 42, determination unit 44 and generation unit 46, below to the determination mould Block 34 is illustrated.

Computing unit 42, for calculating one for terminal user and passing through SDN network after terminal user's access authentication passes through From network side node interface NNI to the forward-path of Customer Premises Network interface UNI；Determination unit 44 is connected to above-mentioned calculating list Member 42, for according to the incidence relation of terminal user and multicast group, passing through traversal addition group on OFSW one by one along forward-path The terminal user for broadcasting group gather determine the flux of multicast of terminal user's set enter to interface it is corresponding go out to set of interfaces；It generates single Member 46, is connected to above-mentioned determination unit 44, for generating OFSW pairs according to going out to set of interfaces for each OFSW on forward-path The group table clause and flow table entry answered, wherein include the movement for being associated with group table clause in the action lists of flow table entry.

Fig. 5 is the preferred structure block diagram of determining module 34 in multicast message replication processes device according to an embodiment of the present invention Two, as shown in figure 5, the determining module 34 is in addition to including all structures shown in Fig. 4, further includes: receiving unit 52, first updates Unit 54 and the second updating unit 56, are below illustrated the determining module 34.

Receiving unit 52 is connected to above-mentioned generation unit 46, and the multicast for receiving terminal user changes request；First Updating unit 54 is connected to above-mentioned receiving unit 52, for the multicast change request according to the terminal user received, along generation Forward-path corresponding to the terminal user of multicast change, according to the incidence relation of updated terminal user and multicast group, It is updated on OFSW one by one into corresponding out to set of interfaces to interface；Second updating unit 56 is connected to above-mentioned updating unit 54, For updated out to the corresponding flow table entry of set of interfaces generation OFSW and group table item according to each OFSW on forward-path Mesh.

Fig. 6 is generation unit 46 in determining module 34 in multicast message replication processes device according to an embodiment of the present invention Preferred structure block diagram one, as shown in fig. 6, the generation unit 46 includes: judgment sub-unit 62 and processing subelement 64, below to this First generation unit 46 is illustrated.

Whether judgment sub-unit 62, the multicast resource for judging current OFSW on forward-path are sufficient；Handle subelement 64, it is connected to above-mentioned judgment sub-unit 62, for selecting current in the case where the judging result of judgment sub-unit is no The upper additional replication processes of a node of OFSW are because current OFSW multicast resource deficiency leads to the current OFSW of not reproducible forwarding The multicast message of user's set, wherein the multicast message additionally replicated passes through OFSW with unicast forwarding processing mode.

Fig. 7 is generation unit 46 in determining module 34 in multicast message replication processes device according to an embodiment of the present invention Preferred structure block diagram two, as shown in fig. 7, the generation unit 46 in addition to including all structures shown in fig. 6, further includes adding sub- list Member 72, is below illustrated the addition subelement 72.

Subelement 72 is added, above-mentioned processing subelement 64 is connected to, it is multiple for carrying out multicast message for upper node OFSW The group table clause that system is handled and generated adds one to multiple buckets, wherein one or more buckets correspond to current OFSW and can not handle Multicast group user set it is corresponding go out to set of interfaces, one or more buckets are carried for by two layers of head of multicast message It is revised as the operational motion of unicast encapsulation.

Fig. 8 is the structural block diagram of open flow controller according to an embodiment of the present invention, as shown in figure 8, the opening flow control Device 80 includes the multicast message replication processes device 82 of any of the above-described.

Through the foregoing embodiment and preferred embodiment, above-mentioned multicast message replication processes can be related to the place of three aspects Reason, carry out the application (for example, BNG application) of multicast message duplication control processing, OFC and OFSW, the BNG using can and OFC Unification.The access authentication of processing user, multicast management are responsible in BNG application, and providing SDN network topology information according to OFC is user Unicast forwarding route is calculated, and is that each OFSW safeguards relevant group of table of user multicast along user's unicast forwarding route on SDN When flux of multicast entrance, user's compound mapping of designated multicast group is closed to outgoing interface collection using on specified OFSW for entry, To generation group table clause, the barrelage amount of guarantee group table clause is minimum, to reduce the duplication of multicast message, BNG application is generated Flow table entry and group table clause corresponding OFSW is handed down to by OFC.OFC is responsible for collecting SDN network topology, issues flow table item Mesh, group table clause etc., and the friendship of the online certification triggering message between terminal user and BNG application and multicast change message is provided Mutual channel.And OFSW carries out replication processes for the multicast message of match hit according to the flow table entry and group table clause that receive. Wherein, it should be pointed out that the protocol interface between OFC and OFSW is OpenFlow agreement.BNG is applied and OFC is deployed separately When, interface suggestion between the two is RESTful interface, such as hypertext transfer protocol (Hypertext Transfer Protocol, referred to as HTTP)；The two can be using internal privately owned interface when unifying.These are not limited.

In the case where there is this unified controller architecture of SDN/OpenFlow, BNG is no longer single physical equipment, but A kind of virtual unit formed by an OF Autonomous Domain network, realizing the control plane function of BNG, (including certification, DynamicHost matches Set the protocol conversations such as agreement (Dynamic Host Configure Protocol, referred to as DHCP)/PPPoE processing) BNG Open flow controller (OpenFlow Controller, referred to as OFC) is resided in using (Application, referred to as APP) On, the mode that (Network Function Virtualization, referred to as NFV) can also be virtualized with network function is stayed It stays on individual server, then passes through the northbound interface of SDN and OFC interaction.

Based on above-mentioned consideration, the user multicast replication theme under a kind of SDN framework is provided in embodiments of the present invention, is led to Access network user's multicast replication based on SDN framework is crossed, is saved using the full-mesh network topological feature perceived and each forwarding The Multicast Replication Capability of point, as far as possible by link or isolation subnet (such as: virtual LAN (Virtual Local Area Network, referred to as VLAN)) multicast message is replicated, it is replicated so as to more flexible and convenient realization user multicast, effectively Crucial or backbone bandwidth resources on user access network are saved, congestion is avoided.

The user multicast replication theme under SDN framework is illustrated below.

Fig. 9 is the networking schematic diagram of the user multicast duplication under SDN framework according to an embodiment of the present invention, as shown in figure 9, OF Autonomous Domain network shows user access network in Fig. 9 schematic diagram, and OFC can have found all open flow switchs (OpenFlow Switch, OFSW) node and link connection topological view.Figure 10 is SDN according to the preferred embodiment of the invention The flow chart of user multicast clone method under framework, as shown in Figure 10, it should be noted that illustrate with more stressing below User multicast is replicated in the execution method on user access network, and for the user multicast duplication under SDN framework, which includes Following steps:

Step S1002, OFC passes through and all OFSW link setups form OF Autonomous Domain network, and utilizes link discovery mechanism (for example, Link Layer Discovery Protocol, Link Layer Discovery Protocol, referred to as LLDP) collects opening up for the network Connection is flutterred, the topological view being made of node set and link set is formed.

Step S1004, the online triggering message (such as PADI of DHCP Discovery or PPPoE) of user are connect by OF Autonomous Domain Enter the interface of message UNI(interchanger user oriented side: the interface of edge switch access user) where OFSW perception, and BNG is handed to using corresponding protocol processes (for example, as unknown flow rate or the preset stream entry of OFC by the path clustering of OFC The local specific protocol message of crawl, is encapsulated in OF-PacketIn message and gives OFC, and delivers after OFC parsing perception Give the processing of corresponding APP functional module).After the online certification of user passes through, the path calculation module of OFC is that the user calculates one It accesses UNI to interface of the first line of a couplet NNI(OF Autonomous Domain network towards public network side, can have multiple NNI, NNI is main here Express the entrance for flowing through the OF Autonomous Domain of some multicast source flux) where node two-way converting path, and give the path edge The OFSW on way issues corresponding stream entry.

Step S1006, after calculating the forward-path of the user, by each section in the user identifier and its forward-path Point and each link are associated, form the associated data table of user and forward node and link.It in this way can be according to user, section Point enters mark to interface, retrieves outgoing interface.

User and link association table clause example are as follows:

[UserID,NodeID,IngressInterface,EgressInterface]

It is as follows to retrieve example:

<UserID,NodeID,IngressInterface>--><NodeID,EgressInterface>

Step S1008, OFC perceive user by BNG control plane multicast protocol (role for serving as multicast router) and are added Or some multicast group is left, the associated data table of a user and multicast group is formed (such as < UserID, MulticastGroupID >), while the multicast replication for triggering OF Autonomous Domain calculates, it is each on the forward-path of the user for modifying in OF Autonomous Domain The upper ALL group table information (Xin Jia or update) relevant with the multicast group of OFSW.The group is added for example, retrieving by multicast group id The user's set for broadcasting group, traverse that the user is integrated into that the multicast traffics OFSW currently calculated on enter enters to interface correspondence The set gone out to interface, the group table clause of ALL type is then generated according to the outgoing interface set, is handed down to and currently calculates OFSW.Number of links due to meeting the constraint condition is generally much smaller than number of users, it is possible to reduce the duplication of multicast message Demand.If certain multicast replication calculates, the multicast resource on some node is (for example, group table clause or Bucket entry volume Degree has been expired) it is insufficient, then it selects to replicate in the upper node selection in the path by user, increases a Bucket, the Bucket newly Two layers of information of multicast message are changed to the mode of unicast (for example, the multicast purpose MAC of message is modified as user by middle setting movement MAC, multicast IP are constant), subsequent Bucket duplication transfers the message come and walks unicast process on path；User leaves the group When broadcasting group, the delete processing user multicast message replicates Bucket.

OFSW implements multicast message duplication by Bucket information present in ALL group table clause.

Step S1010, subsequent user is added or leaves some multicast group all can be according to above-mentioned rule triggering OFC to user institute The group table clause that each OFSW node in path corresponds to multicast group is calculated, and finds the group table item for having link change just to modify Mesh is simultaneously handed down to corresponding OFSW node.For example, the multicast group of node it is corresponding go out become to have to set of interfaces from null set it is non- When null set, ALL type group table clause is increased newly；The corresponding multicast group of node is out nonempty set to set of interfaces, set Increase when number of links has an increase and decrease or deletion group table clause in corresponding Bucket；The multicast group of node is corresponding out to set of interfaces The ALL type group table clause deleted when becoming null set from nonempty set.

Below with reference to preferred embodiment, the present invention will be described.

Preferred embodiment one

Based on shown in Fig. 9, changed with networking example above to describe group table caused by user joins and departs:

1, after User1 certification access, the associated data table clause of user and forward node and link is formed (because of user group The main path for considering NNI to UNI is broadcast, so below only by taking the data table entries in the path as an example), comprising:

[User1ID, OFSW1ID, Intf1-2, Intf1-1(UNI)]

[User1ID, OFSW4ID, Intf4-3, Intf4-1]

[User1ID, OFSW6ID, Intf6-3(NNI), Intf6-1]

2, multicast group MulticastGroup1 is added in User1, forms the associated data table clause of user and multicast group: [User1ID,MulticastGroup1ID]。

3, the group table clause of User1 forward-path in OF Autonomous Domain is calculated for multicast group MulticastGroup1.

A, user User1 has been found by MulticastGroup1ID.

B, along the change of the forward-path node-by-node algorithm group table clause of User1 (corresponding MulticastGroup1).

C, it on OFSW6, with the Intf6-1 of<User1ID, OFSW6ID, Intf6-3(NNI)>find OFSW6, increases newly ALL type group table clause, this group of entry have a Bucket, there is the movement of output Intf6-1 in the Bucket, if should Interface is the sub-interface for configuring VLAN, then the semanteme for having set expression to export in Bucket from the sub-interface is (for example, setting VLAN tag, actual physical interfaces are output to etc.).Similar operation is done on OFSW4 and OFSW1 calculates corresponding group of table item Mesh.

D, OFC is triggered to each OFSW and is issued corresponding flow table entry (flow table of identification MulticastGroup1, mounting Corresponding group of table clause ID) and group table clause, the flux of multicast of user is got through, is also allowed in step C certainly after node-by-node algorithm It issues point by point.

4, after User2 certification access, the associated data table clause of user and forward node and link is formed (because of user group The main path for considering NNI to UNI is broadcast, so below only by taking the data table entries in the path as an example).

[User2ID, OFSW2ID, Intf2-3, Intf2-1(UNI)]

[User2ID, OFSW4ID, Intf4-3, Intf4-2]

[User2ID, OFSW6ID, Intf6-3(NNI), Intf6-1]

5, multicast group MulticastGroup1 is also added in User2, forms the associated data table clause of user and multicast group: [User2ID,MulticastGroup1ID]。

6, the group table clause of User2 forward-path in OF Autonomous Domain is calculated for multicast group MulticastGroup1.

A, user User1 and User2 have been found by MulticastGroup1ID.

B, along the change of the forward-path node-by-node algorithm group table clause of User2 (corresponding MulticastGroup1).

C, on OFSW6, with the Intf6-1 of<User1ID, OFSW6ID, Intf6-3(NNI)>find OFSW6, with< User2ID, OFSW6ID, Intf6-3(NNI) > Intf6-1 of OFSW6 is found, illustrate there there is no corresponding group of table clause on OFSW6 Change.

D, the interface of the link information data OFSW6Intf6-1 connection OFSW4 attempted according to the topology that OFC has found It is Intf4-3.So on OFSW4, with<User1ID, the port Intf4-2 of OFSW4ID, Intf4-3>find OFSW4, with< The port Intf4-1 of User1ID, OFSW4ID, Intf4-3 > find OFSW4, have altered, then update existing group of table clause, increase Add a Bucket, there is the movement of output Intf4-2 in the Bucket, if the interface is the sub-interface for configuring VLAN, The semanteme for having set expression export in Bucket from the sub-interface is (for example, being arranged VLAN tag, being output to actual physics and connect Mouthful etc.).

E, on OFSW2, because of the not information of User1, do not have to look into User1, with < User2ID, OFSW2ID, The Intf2-1(UNI of Intf2-3 > find OFSW2), it increases ALL type group table clause newly, has a Bucket, have in the Bucket The movement of output Intf2-1, if the interface is the sub-interface for configuring VLAN, have in Bucket set expression from The semanteme (for example, setting VLAN tag, be output to actual physical interfaces etc.) of sub-interface output.

F, OFC, which is triggered, issues corresponding flow table entry to each OFSW for having group table clause to change and (such as is known with multicast MAC The stream of other MulticastGroup1 mounts corresponding group of table clause ID) and group table clause, the flux of multicast of user is got through, certainly Also allow to issue point by point after node-by-node algorithm in step C.

7, and so on, it will be appreciated that the meter of User3, User4 and UserN addition multicast group MulticastGroup1 Calculation mode and process.

8, User1 leaves MulticastGroup1, and after OFC perceives the event, path pair is each where prolonging User1 It is corresponding to MulticastGroup1 out to interface link collection to check that MulticastGroup1 user gathers variation for it for node The influence of conjunction.For OFSW6, since going out to interface for User2 is also Intf6-1, leaving for User1 is had no effect on MulticastGroup1 is corresponding out to set of interfaces, so MulticastGroup1 corresponding group of table clause on OFSW6 Do not change；For OFSW4, leaving for User1 causes MulticastGroup1 is corresponding to lack Intf4- to set of interfaces out 1, so MulticastGroup1 corresponding group of table clause on OFSW4 should delete an outgoing interface and correspond to Intf4-1's Bucket, OFC issue corresponding group of table modification message and update corresponding group of table clause to OFSW4；For OFSW1, User1 from Convince by patient analysis and cause MulticastGroup1 is corresponding to go out to set of interfaces to become null set, OFC issue corresponding group of table modification message to OFSW1 deletes corresponding group of table clause.

9, and so on, it will be appreciated that User2, User3, User4 and UserN leave multicast group The calculation and process of MulticastGroup1.

Obviously, those skilled in the art should be understood that each module of the above invention or each step can be with general Computing device realize that they can be concentrated on a single computing device, or be distributed in multiple computing devices and formed Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored It is performed by computing device in the storage device, and in some cases, it can be to be different from shown in sequence execution herein Out or description the step of, perhaps they are fabricated to each integrated circuit modules or by them multiple modules or Step is fabricated to single integrated circuit module to realize.In this way, the present invention is not limited to any specific hardware and softwares to combine.

These are only the preferred embodiment of the present invention, is not intended to restrict the invention, for those skilled in the art For member, the invention may be variously modified and varied.All within the spirits and principles of the present invention, it is made it is any modification, Equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.

Claims (9)

1. a kind of multicast message duplication method, which is characterized in that be applied to software defined network SDN, comprising:

The link connection topological diagram of open flow switch OFSW in the SDN is obtained from open flow controller OFC；

Incidence relation according to the link connection topological diagram and terminal user and multicast group determines the OFSW for handling The flow table entry and group table clause of multicast group message；

Determining the flow table entry and described group of table clause are handed down to the OFSW by the OFC, wherein the OFSW Multicast replication processing is carried out to the multicast group message according to the flow table entry and described group of table clause；

Wherein, true according to the link connection topological diagram and the terminal user and the incidence relation of the multicast group The fixed OFSW is used to handle the flow table entry of multicast group message and group table clause includes: to pass through in terminal user's access authentication Afterwards, one is calculated for the terminal user pass through the SDN network from network side node interface NNI to Customer Premises Network interface The forward-path of UNI；Along the forward-path, according to the incidence relation of the terminal user and the multicast group, by Gathered on a OFSW by the terminal user that the multicast group is added in traversal and determines that the flux of multicast of terminal user's set enters It is corresponding out to set of interfaces to interface；According to described in each OFSW on the forward-path go out to set of interfaces generate described in The corresponding described group of table clause of OFSW and the flow table entry, wherein include being associated in the action lists of the flow table entry The movement of described group of table clause.

2. the method according to claim 1, wherein going out described in each OFSW on according to the forward-path After generating the corresponding described group of table clause of OFSW and the flow table entry to set of interfaces, further includes:

Receive the multicast change request of the terminal user；

According to the multicast change request of the terminal user received, along the forward-path, according to updated institute The incidence relation for stating terminal user Yu the multicast group, updated on OFSW one by one described in enter to interface it is corresponding go out to sets of interfaces It closes；

It is updated described corresponding described to the set of interfaces update OFSW out according to each OFSW on the forward-path Flow table entry and described group of table clause.

3. the method according to claim 1, wherein according to described in each OFSW on the forward-path go out to Set of interfaces generates the corresponding described group of table clause of OFSW and the flow table entry includes:

Judge whether the multicast resource of current OFSW on the forward-path is sufficient；

If the determination result is NO, the upper additional replication processes of a node in the current OFSW are selected because described OFSW multicast resource deficiency leads to the multicast message of current OFSW user's set of not reproducible forwarding, wherein additional multiple The multicast message of system passes through the current OFSW with unicast forwarding processing mode.

4. according to the method described in claim 3, it is characterized in that, additionally multiple in a upper node of the selection in the current OFSW System processing is because the current OFSW multicast resource deficiency leads to the group of current OFSW user's set of not reproducible forwarding After casting text, further includes:

The group table clause for carrying out multicast message replication processes for the upper node OFSW and generating adds one or more buckets, In, the multicast group user set that the one or more of barrels of correspondence current OFSW can not be handled is corresponding to be gone out to sets of interfaces It closes, one or more of buckets carry the operational motion for two layers of head of the multicast message to be revised as to unicast encapsulation.

5. a kind of multicast message replication processes device, which is characterized in that be applied to software defined network SDN, comprising:

Module is obtained, for obtaining the link connection topology of open flow switch OFSW in the SDN from open flow controller OFC Figure；

Determining module determines institute for the incidence relation according to the link connection topological diagram and terminal user and multicast group OFSW is stated for handling the flow table entry and group table clause of multicast group message；

Module is issued, for the flow table entry determined and described group of table clause to be handed down to the OFSW by the OFC, Wherein, the OFSW carries out multicast replication processing to the multicast group message according to the flow table entry and described group of table clause；

Wherein, the determining module includes: computing unit, for being used after terminal user's access authentication passes through for the terminal Family calculates one and passes through the SDN network from network side node interface NNI to the forward-path of Customer Premises Network interface UNI；Really Order member, is used for along the forward-path, according to the incidence relation of the terminal user and the multicast group, one by one Gathered on OFSW by the terminal user that the multicast group is added in traversal determine the flux of multicast of terminal user's set enter to Interface is corresponding out to set of interfaces；Generation unit, for going out according to described in each OFSW on the forward-path to interface Set generates the corresponding described group of table clause of OFSW and the flow table entry, wherein the action lists of the flow table entry In include being associated with the movement of described group of table clause.

6. device according to claim 5, which is characterized in that the determining module further include:

Receiving unit, the multicast for receiving the terminal user change request；

First updating unit becomes for the multicast change request according to the terminal user received along multicast occurs The forward-path corresponding to terminal user more, according to being associated with for the updated terminal user and the multicast group System, updated on OFSW one by one described in enter to interface it is corresponding go out to set of interfaces；

Second updating unit, for updated described out to set of interfaces update institute according to each OFSW on the forward-path State the corresponding flow table entry of OFSW and described group of table clause.

7. device according to claim 5, which is characterized in that the generation unit includes:

Whether judgment sub-unit, the multicast resource for judging current OFSW on the forward-path are sufficient；

Subelement is handled, for selecting in the current OFSW in the case where the judging result of the judgment sub-unit is no The upper additional replication processes of a node because the current OFSW multicast resource deficiency leads to the described current of not reproducible forwarding The multicast message of OFSW user's set, wherein the multicast message additionally replicated passes through the OFSW with unicast forwarding processing mode.

8. device according to claim 7, which is characterized in that the generation unit further include:

Subelement is added, the group table clause addition generated for carrying out multicast message replication processes for the upper node OFSW One or more buckets, wherein the multicast group user that the one or more of barrels of correspondence current OFSW can not be handled gathers institute It is corresponding go out to set of interfaces, one or more of buckets are carried for two layers of head of the multicast message to be revised as unicast The operational motion of encapsulation.

9. a kind of open flow controller OFC, which is characterized in that including device described in any one of claim 5 to 8.