A kind of implementation method of movable virtual special net and system
Technical field
The present invention relates to the Virtual Private Network field, relate in particular to a kind of implementation method and system of movable virtual special net.
Background technology
VPN (Virtual Private Network, Virtual Private Network) technology is meant in the network that the common network service provider is provided and sets up dedicated network that user data is propagated in public network by the passage of a safety.VPN can allow the business user enjoy fail safe, priority, ease of manageability and the reliability the same with this user's dedicated network, has also reduced commercial expense simultaneously, and therefore, VPN becomes the development trend of enterprise network.
Wherein, MPLS (MultiProtocol Label Switching, multiprotocol label switching) VPN technologies are merit attention most in the various solutions of VPN a kind of, comprise three layers of MPLS VPN and two layers of MPLS VPN, described three layers of MPLS VPN have comprised BGP (Border GatewayProtocol again, Border Gateway Protocol)/MPLS VPN and two kinds of implementations of VR (Virtual Router, virtual router) VPN.MPLS offers label of each IP packet, and this label and IP data envelope are loaded on new MPLS packet, determines the transmission path and the priority of IP packet thus; And with the router of MPLS compatibility, can before the IP packet is transmitted by respective paths, only read the packet header label of this MPLS packet, need not go to read the information such as IP address in each IP packet again, so the exchange forwarding speed of packet is accelerated greatly.
In BGP/MPLS VPN, MBGP (MultiProtocol Border Gateway Protocol, the multi-protocols extending boundary gateway protocol) is used for distributing route at service provider backbone, MPLS is used for transmitting packet at backbone network, at PE (Provider Edge, provider's edge device) sets up different VRF (Virtual Route Forwarding for different VPN users on the router, virtual routing forwarding), and then formation MPLS transmits, the bearing capacity of the multi-protocols expansion by bgp protocol is transmitted member relation and the reachability information of VPN, realizes that between VPN user route is isolated and announcement, transmit Business Stream, allow the service provider to use public network that the VPN service of brand-new type is provided as the user.When original VPN network all functions are provided, strong QoS can also be provided (Quality Of Service, service quality) ability has reliability height, safe, characteristics such as extended capability is strong, control strategy is flexible and managerial ability is powerful.
Day by day ripe along with Next Generation Internet network standard I Pv6 (Internet Protocol Version 6, Internet protocol edition 6) had increasing manufacturer that support to the IPv6 network is provided.When each website among the VPN supported that all an IPv6 and an IPv6 interface and subinterface by PE are connected to service provisioning merchant backbone network, VPN herein promptly referred to IPv6 VPN.Wherein, BGP and extended attribute thereof are used to describe the route that IPv6 VPN website links to each other with other IPv6 VPN website by pe router.PE is by setting up reachability information and the forwarding information that VRF safeguards each IPv6VPN respectively.
In IPv6, allow each IPv6 VPN that its privately owned IPv6 address space is arranged, this means that a given address table is shown in the different system in the different VPN.This is to realize by a kind of new address family, i.e. VPN-IPv6 address family.Wherein, there are 24 bytes the VPN-IPv6 address, is made up of the RD (Route Distinguisher, route-distinguisher) of 8 bytes and the IPv6 address of 16 bytes.If two VPN use same IPv6 address prefix (representing different physical systems), PE translates into unique VPN-IPv6 address prefix to VPN by different RD, and then guarantee when two different VPN use identical address, this address will be loaded into two diverse routes, respectively corresponding each VPN.Wherein, the VPN-IPv6 address is always thought unique by BGP, and bgp extended community permission BGP carries routing iinformation and the MPLS label information from the address family of expansion.
Except encapsulating by the MPLS label switched path, the BGP/MPLS vpn solution also is extended to permission and comprises that by other tunneling technique gre tunneling, IP-in-IP tunnel and IPsec tunnel encapsulate.Equally, also allow to comprise that by MPLS LSP and other tunneling technique gre tunneling, IP-in-IP tunnel or IPsec tunnel realize the support to IPv6 VPN service.
In addition, replace IPv4 (Internet Protocol Version 4 at IPv6, the Internet protocol edition 4) in the process, because the extensive use of existing IPv4 network, will there be very long a period of time in the IPv4 network, wherein IPv4 network environment IPv6 network environment is basic identical, and just the identification message of the VPN of IPv4 is different with the formation of the VPN identification message of IPv6.
Along with the continuous evolution and the Next Generation Internet network agreement of Internet technology are better supported network mobility, people will improve day by day to the mobility demand of agreement, equipment in the network, its ambulant support is also had higher utility, but existing BGP/MPLS VPN technologies can't solve its user network mobility is supported this problem.
Summary of the invention
The problem to be solved in the present invention provides a kind of implementation method and system of movable virtual special net, so that the defective that can't provide the user network mobility to support in the prior art to be provided.
For addressing the above problem, the invention provides a kind of implementation method of movable virtual special net, may further comprise the steps:
A, first subscriber equipment are issued routing iinformation by first provider's edge device to second provider's edge device;
B, second subscriber equipment obtain described routing iinformation by described second provider's edge device;
C, described second provider's edge device is inquired about second subscriber equipment whether new label is arranged, if having, represents that second subscriber equipment is a mobile subnetwork, then changes step e; Otherwise, represent that second subscriber equipment is a mobile host, then changes step D;
D, described mobile host carry out transfer of data by second provider's edge device and described first subscriber equipment;
E, described mobile subnetwork carry out transfer of data by the 3rd provider's edge device and described first subscriber equipment.
The forming process of new label comprises described in the step C:
C1, mobile subnetwork move to the 3rd provider edge device controlled area from the second provider edge device controlled area, receive the Agent Advertisement information that the 3rd provider's edge device sends;
C2, described mobile subnetwork are according to the address prefix information co-located care-of address in the Agent Advertisement information;
C3, the 3rd provider's edge device are that mobile subnetwork distributes a multi-protocols packet label, and bind with Care-of Address, form new label.
Also comprise after forming new label:
C4, mobile subnetwork send registration information by the 3rd provider's edge device to second provider's edge device;
C5, second provider's edge device send the binding response message by the 3rd provider's edge device to mobile subnetwork.
The described registration information of step C4 comprises Binding Update information and MP-BGP MP-BGP routing update information.
Also comprise between step C4 and the step C5:
Second provider's edge device judges by Binding Update information whether itself is the home agent of this mobile subnetwork, if then Care-of Address and the new label with second subscriber equipment writes route forwarding table, otherwise, abandon this Binding Update information.
Among the step C1, mobile subnetwork obtains Agent Advertisement information from the 3rd provider's edge device by Internet control message protocol.
The 3rd provider's edge device and second provider's edge device regularly send routing broadcast information, have the prefix addresses information on the link-local in this information, and second subscriber equipment is determined the position of second subscriber equipment according to this prefix addresses information.
The second subscriber equipment broadcast router solicited message receives that provider's edge device of this solicited message is all replied this information, makes second subscriber equipment know position own.
Step D further comprises:
D11, mobile host send packet to second customer edge;
D12, second customer edge are forwarded to second provider's edge device with this packet;
D13, second provider's edge device judge by the inquiry virtual routing forwarding whether the destination address in this packet mates, if coupling is then changeed step D14, otherwise, abandon this packet;
D14, second provider's edge device are inquired about next jumping and departure interface, and distribute top label;
D15, first provider's edge device receive this packet by backbone network;
D16, first subscriber equipment obtain this packet from first provider's edge device.
Step D16 further comprises:
D161, first provider's edge device inquire first subscriber equipment by bottom label;
D162, first subscriber equipment receive this packet by the IP layer.
Step e further comprises:
E11, first subscriber equipment send to second provider's edge device by backbone network with packet;
E12, the new label of second provider edge device utilization send to the 3rd provider's edge device with packet;
E13, the 3rd provider's edge device are transmitted this packet to mobile subnetwork by the inquiry virtual routing forwarding.
Step e further comprises:
E21, mobile subnetwork send to the 3rd provider's edge device with packet;
E22, the 3rd provider's edge device by the inquiry virtual routing forwarding judge this packet whether with the first subscriber equipment matching destination address, if coupling is then changeed step e 23, otherwise, change step e 25;
E23, the 3rd provider's edge device are forwarded to first provider's edge device with packet, and inform new label;
E24, first provider's edge device are transmitted packet to first subscriber equipment, and upgrade the information among the virtual routing forwarding VRF;
E25, the 3rd provider's edge device are forwarded to second provider's edge device with packet;
E26, second provider's edge device are inquired about the matching destination address clauses and subclauses that whether have among the virtual routing forwarding VRF with first subscriber equipment, if having, then packet are forwarded to first provider's edge device, and inform new label; Otherwise, abandon this packet.
Steps A further comprises:
A1, first provider's edge device obtain the routing iinformation of first subscriber equipment;
A2, first provider's edge device write virtual routing forwarding with the address prefix in the routing iinformation of first subscriber equipment;
A3, first provider's edge device distribute multi-protocols packet label bottom label for this routing iinformation.
14, as the implementation method of movable virtual special net as described in the claim 13, it is characterized in that when first subscriber equipment was main frame, steps A 1 further comprised:
A11, first customer edge obtain the address information of first main frame;
A12, first provider's edge device obtain the routing iinformation of first customer edge.
First customer edge is by manual configuration or the automatic address information of finding to obtain first main frame.
Step B further comprises:
B1, second provider's edge device obtain the routing iinformation of first subscriber equipment from first provider's edge device by MP-BGP;
B2, second provider's edge device compare this routing iinformation and virtual routing forwarding input policing, if coupling is then changeed step B3, otherwise abandoned this routing iinformation;
B3, second provider's edge device write the Virtual Private Network routing information base with this routing iinformation;
B4, second subscriber equipment obtain this routing iinformation from second provider's edge device.
Comprise in the MP-BGP information: the loop-back address of network address prefix, first provider's edge device is as the multi-protocols packet label that next is jumped, routing iinformation is assigned with of Border Gateway Protocol and the route target community of this routing iinformation.
When mobile subnetwork moves back to former this locality when acting on behalf of, then mobile subnetwork can be received the proxy announcement message that second provider's edge device sends, thereby judge it and returned link-local, mobile subnetwork is just nullified its present care-of address and new label to second provider's edge device.
First subscriber equipment and the communication process that is arranged in mobile subnetwork, when mobile subnetwork moves to the 4th provider edge device controlled area from the 3rd provider edge device controlled area, first subscriber equipment sends to the 3rd provider's edge device with packet, adopts the overtime re-transmission mode of transmission control protocol layer to carry out transfer of data.
Mobile subnetwork can be crossed provider's edge device of communicating by letter to all and its foundation and send a message of upgrading its Care-of Address and label;
First provider's edge device obtains upgrading after this message the content of virtual routing forwarding;
First provider's edge device mails to the 4th provider's edge device with the packet of first subscriber equipment;
The 4th provider's edge device sends this packet to mobile subnetwork.
The present invention also provides a kind of realization system of movable virtual special net, comprise: backbone network, first customer edge and mobile host, wherein comprise first provider's edge device, second provider's edge device in the backbone network, first customer edge and mobile host belong to same Virtual Private Network;
First subscriber equipment is issued routing iinformation by first provider's edge device in the backbone network to second provider's edge device; Second subscriber equipment obtains described routing iinformation by described second provider's edge device; Described mobile host carries out transfer of data by second provider's edge device and first provider's edge device and described first subscriber equipment.
Comprise: backbone network, first customer edge and mobile subnetwork, wherein comprise first provider's edge device, second provider's edge device and the 3rd in the backbone network for merchant's edge device, described equipment belongs to same Virtual Private Network;
When mobile subnetwork is moved to the 3rd provider edge device control regional by the control area of second provider's edge device, will receive the Agent Advertisement information that the 3rd provider's edge device sends; Mobile subnetwork is according to the address prefix information co-located care-of address in the Agent Advertisement information; The 3rd provider's edge device is that mobile subnetwork distributes a multi-protocols packet label, and binds with Care-of Address, forms new label;
When first subscriber equipment when mobile subnetwork sends packet,
First subscriber equipment sends to second provider's edge device by first provider's edge device in the backbone network with packet; The new label of second provider edge device utilization sends to the 3rd provider's edge device with packet; The 3rd provider's edge device is transmitted this packet to mobile subnetwork by the inquiry virtual routing forwarding;
When mobile subnetwork when first subscriber equipment sends packet,
Mobile subnetwork sends to the 3rd provider's edge device with packet; The 3rd provider's edge device judge this packet whether with the first subscriber equipment matching destination address, if coupling, then the 3rd provider's edge device is forwarded to first provider's edge device with packet, and inform new label, first provider's edge device is transmitted packet to first subscriber equipment, and upgrades the information in the virtual routing forwarding; Otherwise, the 3rd provider's edge device is forwarded to second provider's edge device with packet, whether matching destination address clauses and subclauses with first subscriber equipment are arranged in second provider edge device inquiry virtual routing forwarding, if have, then packet is forwarded to first provider's edge device, and informs new label; Otherwise, abandon this packet.
Also comprise the 4th provider's edge device, first subscriber equipment and the communication process that is arranged in mobile subnetwork, when mobile subnetwork moved to the 4th provider edge device controlled area from the 3rd provider edge device controlled area, mobile subnetwork can be crossed provider's edge device of communicating by letter to all and its foundation and send a message of upgrading its Care-of Address and label; First provider's edge device obtains upgrading after this message the content of virtual routing forwarding; First provider's edge device mails to the 4th provider's edge device with the packet of first subscriber equipment; The 4th provider's edge device sends this packet to mobile subnetwork.
Compared with prior art, the present invention has the following advantages:
The present invention is directed to the network configuration that MPLS backbone network platform and user network are formed, relevant Virtual Private Network equipment and corresponding data packet transmission flow process are improved, proposed the solution that a kind of MPLS of making backbone network and mobile subnetwork realize that the BGP/MPLS VPN mobility of combination is supported, having solved can not the ambulant defective of virtual support private network in the prior art.
The present invention realizes that the subnet locomotive function can be applied to intra-company and hold a meeting under mobile environment such as train, steamer, All hosts in the whole like this mobile environment has just constituted a mobile website at this moment, promptly formed an in-company mobile subnetwork, wherein just having multiple host need keep getting in touch at any time at other website of longer following period of time and company.
The present invention does not need upgrading hardware, only needs the software of PE equipment is improved simple, the easy row of configuration mode.In addition, structure of the present invention meets popular mobile subnetwork system at present, is with good expansibility and marketing.
Description of drawings
Fig. 1 is the implementation method flow chart of movable virtual special net in a kind of specific embodiment of the present invention;
Fig. 2 is the realization system construction drawing of movable virtual special net in a kind of specific embodiment of the present invention;
Fig. 3 is a VPN routing iinformation issue flow chart in a kind of specific embodiment of the present invention;
Fig. 4 is the data transport stream journey figure of a kind of specific embodiment in Fig. 2 system;
Fig. 5 is the realization system construction drawing of movable virtual special net in another specific embodiment of invention;
Fig. 6 is that the Care-of Address and the new label of a specific embodiment in Fig. 5 system forms flow chart;
Fig. 7 is that the fixed host computer of a specific embodiment in Fig. 5 system sends data flowchart to mobile subnetwork;
Fig. 8 is that the mobile subnetwork of a specific embodiment in Fig. 5 system sends data flowchart to fixed host computer.
Embodiment
Following the present invention will be in conjunction with the accompanying drawings, and optimum implementation of the present invention is described in detail.
The flow process of the implementation method of movable virtual special net of the present invention may further comprise the steps as shown in Figure 1:
Step s101, first subscriber equipment is issued routing iinformation by first provider's edge device to second provider's edge device.
Step s102, second subscriber equipment obtains described routing iinformation by described second provider's edge device.
Step s103, whether described second provider's edge device is inquired about second subscriber equipment new label, if having, represents that second subscriber equipment is a mobile subnetwork, then changes step s105; Otherwise, represent that second subscriber equipment is a mobile host, then changes step s104.The forming process of new label comprises: mobile subnetwork moves to the 3rd provider edge device controlled area from the second provider edge device controlled area, receives the Agent Advertisement information that the 3rd provider's edge device sends; Described mobile subnetwork is according to the address prefix information co-located care-of address in the Agent Advertisement information; The 3rd provider's edge device is that mobile subnetwork distributes a multi-protocols packet label, and binds with Care-of Address, forms new label.
Step s104, described mobile host carries out transfer of data by second provider's edge device and described first subscriber equipment.
Step s105, described mobile subnetwork carries out transfer of data by the 3rd provider's edge device and described first subscriber equipment.
The present invention is equally applicable to the IPv4 network environment, just the identification message of the VPN of IPv4 is different with the formation of the VPN identification message of IPv6, in mobile IPv 4, mobile node is told home agent by the log-on message that carries in the UDP/IP bag with its Care-of Address, and the mobile node in the mobile IP v 6 notifies other nodes its Care-of Address with destination address option (Destination Options), make mobile node obtain the agreement difference that Agent Advertisement uses from the 3rd provider's edge device, similar on the flow process.Therefore, below only to describe under the IPv6 network environment.
Because for BGP/MPLS VPN, generally do not have the situation that has mobile device in its MPLS backbone network, thereby its mobility support mainly concentrates on two kinds of following situations: the one, realize to the mobility support under the IPv6 network environment that is: CE fixes, main frame moves in that each website of BGP/MPLS VPN is inner; The 2nd, main frame and each CE under each website all possess the mobility support under the IPv6 network environment, make that the whole website that links to each other with CE is that subnet has been realized locomotive function.Therefore specific embodiments of the invention also move from main frame and move two aspects with subnet and be described.
Inner realization of each website of BGP/MPLS VPN the mobility under the IPv6 network environment is realized system as shown in Figure 2, comprise: MPLS backbone network, website 1, website 2, website 3, website 4, CE1, CE2, CE3 and CE4, wherein, main frame 1 or CE1 are that first subscriber equipment, main frame 2 are second subscriber equipment; Backbone network comprises: PE1, PE2, P1, P2 and P3 have two virtual routing forwarding VRFA and VRFB respectively among PE1 and the PE2; Website 1 and website 2 belong to Virtual Private Network A, website 3 and website 4 and belong to Virtual Private Network B; Sub-district a and sub-district b are arranged among the CE2, and all P, PE and ce router are all fixed, and the main frame 2 in the website can move between sub-district a and sub-district b.Above-mentioned whole network is the territory of an IPv6, and wherein all devices is all supported IPv6.
At first carry out the issue of VPN routing iinformation,
Step s101 and step s102 further comprise as shown in Figure 3 among Fig. 1: CE1 by manual configuration or automatically mode such as discovery obtain the address information of main frame 1, then again oneself IPv6 route prefix is broadcast to PE1; PE1 obtains the routing iinformation of CE1 by route learning mechanism; PE1 writes virtual routing forwarding with the address prefix in the routing iinformation of main frame 1; Before this route of broadcasting, PE1 distributes multi-protocols packet label bottom label for this routing iinformation.PE2 obtains the routing iinformation of main frame 1 by MP-BGP broadcasting from PE1, and this broadcasting comprises: the VPN-IPv6 address prefix of this route, PE1 return the MPLS label that the address is assigned with as BGP next hop, this route, the route target community (being the route output policy) of this route VRF configuration; PE2 compares this routing iinformation and virtual routing forwarding input policing, if coupling, then PE2 writes Virtual Private Network IPv6 routing information base VPN-IPv6.RIB with this routing iinformation, and main frame 2 obtains this routing iinformation from PE2; Otherwise abandon this routing iinformation.
Step s104 further comprises as shown in Figure 4:
Will be as main frame 2 to main frame 1 issue vpn service, then earlier packet is sent to CE2, CE2 carries out the longest coupling routing inquiry, again this packet is forwarded to direct-connected PE2, PE2 inquires about corresponding VRF based on sub-interface, as destination address the coupling clauses and subclauses are arranged, then its next jumping of routing inquiry and departure sub-interface; PE2 distributes top label, arrives PE1 by the MPLS backbone network; Inquire corresponding C E1 at the PE1 place by bottom label, transmit by the IP layer then and wrap to CE1, table of query and routing is forwarded to main frame 1 again.
The mobility that whole website all is implemented under the IPv6 network environment in BGP/MPLS VPN realizes system as shown in Figure 5, comprise: MPLS backbone network, website 1, website 2, CE1 and CE2, wherein, main frame 1 or CE1 are that first subscriber equipment, CE2 are second subscriber equipment; Backbone network comprises: PE1, PE2, PE3, PE4, P1, P2 and P3 have virtual routing forwarding VRFA among PE1, PE2, PE3 and the PE4; Website 1 and website 2 belong to Virtual Private Network A; Above-mentioned whole network is the territory of an IPv6, and wherein all devices is all supported IPv6; All P, pe router are fixed, and All hosts in the website 2 and the CE2 corresponding with it all move, and make the whole website that links to each other with CE realize locomotive function.
The issue flow process of the VPN routing iinformation of this kind situation is identical with the issue flow process of mobile host, repeats no more here.
Suppose that whole website 2 is in mobile status.Each mobile host in the website 2 need not finished when leaving this locality or distinguishing more and go for couple in router AR respectively and to the work of home agent HA registration and so on this moment, remained when the host work that fix to get final product.The selected CE2 of present embodiment promptly is equivalent to MR (Mobile Router, mobile router) as the unified agent router of mobile site 2, finishes a series of activities that is similar to single MIP.
When website 2 when local (district) moves to other places (two districts), a fixing router to be arranged as HA (Home Agent, home agent) in this locality.Present embodiment selects PE2 as HA, selects PE3 as couple in router AR (Access Router, couple in router) simultaneously, and mobile subnetwork arrives the idiographic flow that communicates behind the other places as shown in Figure 6:
After website 2 moves to two districts by a district, CE2 receives the Agent Advertisement information that PE3 regularly sends, under the IPv6 network environment, CE2 finishes the router function of search by ICMPv6 router solicitation in the neighbor discovery mechanisms and router broadcast message, obtains Agent Advertisement information.PE3 and PE2 can regularly send routing broadcast message, wherein have the Prefix message on the link-local; Certainly, also can send router request message, receive that each router of this message all should be used the router broadcast acknowledgement messaging by CE2.CE2 learns that the website of oneself has moved on the foreign link after receiving its proxy announcement message, will be disposed automatically by the address according to this new address prefix information and obtain a Care-of Address.Obtain Care-of Address dual mode is arranged: passive address configuration automatically disposes automatically with the active address.This moment, PE3 can distribute a new MPLS label for CE2, and bound with its Care-of Address.
CE2 sends login request message by PE3 to its home agent PE2.This moment, CE2 need be with a mark<R in the Binding Update that is sent to the PE2 transmission 〉, to represent mobile router but not common mobile node.Also to comprise mobile network's prefix in the same Binding Update, can use for reference a kind of new mobility Header option that proposes among the NEMO and come this prefix message of load.Specific as follows: PE3 can find the CE2 address of affiliated PE2 originally by inquiring about corresponding VRFA after distributing new label for CE2.Then PE3 just is sent to PE2 with binding update messages and MP-BGP routing update message (comprising mark<R 〉, the Care-of Address of CE2 and new label etc.).
PE2 sends the binding response message by PE3 to CE2.Receive the binding update messages of CE2 as PE2 after, PE2 just can confirm its Binding Update, thereby and obtain Care-of Address and the new label of CE2 and write among the VRFA and the corresponding part of CE2, utilize MP-BGP to pass through the MPLS backbone network then and send the binding response message to CE2.After PE3 receives this response message, just the route of CE2 and new label binding will be led among its VRFA.
Owing to be in the IPv6 network environment, thereby CE2 has adopted the notice mode to come to notify its Care-of Address and new label (if in the IPv4 network environment, CE2 adopts the announcement mode to come to notify its Care-of Address and new label to other website of PE2 and VPNA simultaneously) to other website of PE2 and VPNA simultaneously.Wherein, notice has comprised Binding Update, has bound and reply and three kinds of type of messages of bind request, and they all are extension header that is placed on IPv6---in the optional header in destination.By adopting the destination address option in this IP header that the Care-of Address of CE2 is notified to PE2.Wherein bind request is: when the life span territory in the binding update messages that other website of VPNA had before been received wants expired, will send this message, come requirement CE2 to send to its new Binding Update, to obtain available Care-of Address and new label.
Suppose main frame 1 in the website 1 main frame issue vpn service in will the mobile subnetwork under CE2, idiographic flow as shown in Figure 7, main frame 1 is forwarded to PE2 by the MPLS backbone network with packet; PE2 inquires about its corresponding VRFA, find that CE2 has arrived the other places, so in table, inquired the new label of CE2 again, then same mode according to two-layer label has found PE3, PE3 is sent to CE2 by searching corresponding VRFA with bag, and CE2 inquires about corresponding routing table again and sends it to destination host.
When the mobile host among the CE2 will be when the main frame in the website 11 sends data, idiographic flow as shown in Figure 8, mobile host sends to CE2 by table of query and routing with packet, CE2 sends to PE3 by table of query and routing with this packet, PE3 inquires about corresponding VRFA, as destination address the coupling clauses and subclauses are arranged, then its next jumping of routing inquiry and departure sub-interface.Transmit bag to the destination by the MPLS backbone network again.In this process, CE2 can announce new label to PE1 by MP-BGP, after PE1 receives this message, corresponding VRFA is upgraded.Afterwards with website 2 in the communicate by letter link that just is sent to the representative of this label of mobile host, thereby this scheme has also been avoided the triangle routing issue that exists in NEMO.
When website 2 moved back to a district, then CE2 can receive the proxy announcement message that PE2 sends, and had returned home link thereby judge it, so CE2 just nullifies its present care-of address and new label to PE2, so just can work as stationary nodes again.
As at main frame 1 and the communication process that is arranged in the mobile node of website 2, website 2 has moved to another nonlocal three districts again from two districts, and this moment, main frame 1 just mail to a packet PE3, so at this moment, the mechanism of a kind of being called " overtime re-transmission " that can be by the TCP layer, promptly so long as in the scope of a<time<b (a, b can artificially limit), just think packet loss rather than think that network is obstructed and abandon the transmission of this bag is handled.And in this process, all when two districts and CE2 set up nodes in communication information and all can be recorded in it transmits, CE2 can send a message of upgrading its Care-of Address and label to these nodes, like this, PE1 will upgrade content corresponding among the VRFA after obtaining this message, then the bag with main frame 1 mails to PE4, thereby makes mobile host receive packet, has avoided packet loss that may bring owing to handover or the problem that manually retransmits.Wherein, CE2 is after confirming switching, should be immediately send binding update messages in two districts to main frame 1 and CN that all are got in touch with it, by the time after the business recovery of each CN and CE2, send Binding Update to PE2 again, embody the existing business of preferential assurance like this, then considered the service strategy of the foundation of new business.
The above only is a preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.