WO2017096565A1 - 一种集中式身份网络路由协议cin - Google Patents

一种集中式身份网络路由协议cin Download PDF

Info

Publication number
WO2017096565A1
WO2017096565A1 PCT/CN2015/096860 CN2015096860W WO2017096565A1 WO 2017096565 A1 WO2017096565 A1 WO 2017096565A1 CN 2015096860 W CN2015096860 W CN 2015096860W WO 2017096565 A1 WO2017096565 A1 WO 2017096565A1
Authority
WO
WIPO (PCT)
Prior art keywords
router
identity
packet
host
identifier
Prior art date
Application number
PCT/CN2015/096860
Other languages
English (en)
French (fr)
Inventor
李挥
陈文生
陆军
尘福兴
Original Assignee
北京大学深圳研究生院
深圳市维金康智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京大学深圳研究生院, 深圳市维金康智能科技有限公司 filed Critical 北京大学深圳研究生院
Priority to PCT/CN2015/096860 priority Critical patent/WO2017096565A1/zh
Publication of WO2017096565A1 publication Critical patent/WO2017096565A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
    • H04L9/40—Network security protocols

Definitions

  • the present invention relates to the field of network protocols, and in particular, to a centralized identity network routing protocol CIN.
  • SDN Software Defined Network
  • Traditional switch functions are forwarded by the lowest level of traffic, and other advanced processing functions (such as network management control, Load balancing, etc.), the SDN strips off all the advanced processing functions of the switch except forwarding, and moves this part of the advanced processing functions to the "controller” to separate the control plane from the forwarding plane.
  • the communication rules are determined by communication between specific protocols.
  • the network operating system (NOS) running on the controller provides an application programming interface (API) to various SDN applications to meet different network requirements, such as custom routing. Scheduling rules, network isolation, traffic management, and quality of service management.
  • the overall upgrade of the network function can be realized through the writing and updating of the software, no need to configure for each hardware device, and the network configuration can be directly deployed through the form of network services and application programs, and the network can be customized. Easy to customize and accelerate network deployment cycle.
  • the dual semantics possessed by the IP address in the Internet system are the main factors affecting the size of the Internet and the mobility of the terminal.
  • the identity location separation technology emerges as the times require.
  • Identity routing is a technique for routing information directly using the identity of a communication terminal. Compared with a location identifier (such as a hierarchically-divided address in an IP network), the identity is flat.
  • the identity and location separation technology uses two identifiers to communicate. Each terminal has a globally unique identity, which is only used to mark the identity information of the communication terminal, and has no correlation with the location information, but its location identifier and terminal location. Location related. After the terminal moves, its identity remains unchanged.
  • the location identifier is provided by its route and is updated in real time to the identity and location mapping database.
  • the access router of A obtains the required mapping relationship.
  • the identity location mapping resolves and adds the location identifier to the data packet, and the data packet is forwarded by the location identifier between different routers. Simulating The cache is mapped to reduce the query latency of subsequent packets.
  • the existence of the identity identifier ensures the uniqueness of the terminal identifier, and the existence of the location identifier ensures the scalability of the routing table.
  • the real-time mapping and resolution of the two types of identifications ensures the continuity of the terminal communication, that is, mobility.
  • the Host Identity Protocol introduces a new protocol layer, the Host Identity Layer (HIL), between the network layer and the transport layer of the terminal protocol stack.
  • the HIP uses HI (Host Identity) to identify the identity of the terminal.
  • the IP address in the network layer is used as the location identifier only for routing and no longer plays the role of the host name.
  • the transport layer is no longer coupled to the network layer.
  • the host identification layer is logically located between the network layer and the transport layer.
  • the transport layer uses ⁇ HI, port> as the transport layer identifier instead of ⁇ IP address, port>, and the host identifier in the packet is completed by the host identification layer.
  • IP address translation The network layer is shielded from the transport layer, and any changes in the network layer (eg, changes in the host IP address during communication) do not affect the transport layer link unless the quality of service (QoS) changes.
  • QoS quality of service
  • HI represents a globally unique static name used to mark a host.
  • a host can have multiple HIs.
  • the storage and lookup of HI can be done by the host itself or by a third party such as DNSSEC, PGP or X.509.
  • DNSSEC DNSSEC
  • PGP PGP
  • X.509 IP address
  • a dynamic DNS server can be used to provide a relatively static mapping of domain names to dynamic IP addresses.
  • the protocol designs a convergence server RS (Rendezvous Server) to meet the needs of querying the current IP address of the mobile node.
  • the aggregation server is a packet forwarder that can be used by the aggregation server to continue to send and receive data packets without interrupting communication when the mobile host moves to a new network location.
  • the aggregation server completes the forwarding of the data packet, and all the hosts that request the data packet forwarding are called the aggregation client.
  • To complete the forwarding of the data packet first add a resource record to the DNS server, store the aggregation server information for the aggregation client to query; secondly, the aggregation client registers its own HI and IP address and other information on the aggregation server.
  • the host When the host wants to send a data packet, it first queries the DNS server for the HI and IP address of the aggregation server, and sends the first data packet to the aggregation server, and then the aggregation server forwards the data packet to the destination address, and the subsequent data packet does not.
  • the communication is forwarded through the aggregation server, but the communication parties directly communicate, unless one of the two parties in the communication changes.
  • the aggregation server and the peer node are notified of the change of the address. Even if the network addresses of the two communicating parties change at the same time, the data can be forwarded through the aggregation server for normal communication.
  • the data link of the transport layer is not interrupted.
  • HIP mainly adopts two solutions: 1Using dynamic DNS, the DNS always keeps the latest IP address of the mobile node and the corresponding HI/HIT, but there is no doubt that the frequent movement of the host will cause the performance of the DNS server to drop sharply, and The unusable problem of DNS record buffering results in poor performance and scalability of the solution. In addition, the solution cannot solve the problem of simultaneous mobile communication. 2 Adopt RS, which requires each home network to be configured with RS, which is not conducive to Wide deployment and implementation of HIP. The entire network topology is unique at a certain time. Each router independently implements the discovery and routing calculation of the entire network topology, resulting in a certain degree of computational redundancy.
  • the present invention provides a centralized identity network routing protocol CIN, which solves the problem of complex infrastructure processing of identity-location mapping in the prior art.
  • the present invention provides a centralized identity network routing protocol CIN, including an application layer, a data link layer and a physical layer, and an identity/location layer, the identity/location layer being located between the application layer and the data link layer;
  • the identity/location layer decouples the identity attribute and the location attribute of the IP address; the identity/location mapping entry of the host needs to be placed in the entire network identity location mapping database ILMD and the identity location mapping cache ILMC.
  • each host and router has an independent and unique identifier, and the identity HID of the host is the identifier, and the location LOC of the host is the identifier RLID of the registered router; when the host sends a routing protocol packet, the host The host uses its own HID as the Source ID and the destination host HID as the Destination ID. If the destination host is in the same LAN, it is directly transmitted to the destination host. If it is not in the same LAN, it must be in the process of transmitting the routing protocol packet. The Source ID field and the Destination ID field are rewritten to identify the RLID for the source and destination routers, and the source and destination hosts are added to the extension field and routed between the routers via the RLID.
  • the identity/location storage resolution system is responsible for storing and parsing the identity-to-location mapping, specifically: placing a network-wide identity location mapping database ILMD on the control end, and placing an identity location mapping cache on the router side.
  • ILMC network-wide identity location mapping database
  • the access router will pass ⁇ HID, HLID> to ILMD, and ILMC will be used as a subset of ILMD to access the router.
  • the input packet quickly resolves the identity location mapping; if there is no hit, the router queries the console for the mapping and caches it in ILMC.
  • the CIN protocol adopts a control/forward separation policy.
  • the controller adopts a corresponding routing policy according to the application and the user requirements, and responds to the controller to query the request and deliver the routing entry.
  • the centralized identity network routing protocol CIN includes route calculation, and the controller acquires LSAs from each router to establish a network-wide LSDB, and the LSAs contain link information of one router to another router; W, W ij represents the link cost from router i to router j.
  • the Freud algorithm can be used to calculate the shortest path and the next hop of any two points.
  • Each application can customize the route as needed; each time the LSA changes, Recalculate the route.
  • the packet forwarding process by the centralized identity network routing protocol CIN is: (A) H1 sends a data packet to its ingress router Ingress Router, the source of the data packet / Destination ID field needs to be HID1 and HID2; (B) Each router maintains a recently used identity-location mapping cache ILMC.
  • the ingress router When the ingress router receives the H1 packet, it first queries the local ILMC table if the query hits Go to step (D), miss, go to step (C); (C) the ingress router obtains HID2 location information RLID2 by querying controller ILMD, further caches it to its ILMC; (D): ingress router Rewrite the packet with the source ID RLID1 and the destination ID RLID2; HID1 and HID2 are added to the extension field, and the packet is routed to another router; (E): When the ingress router passes to another router path When there is a routing table to another router in the node, the packet will be hopped to another router.
  • the router with the missing routing entry will send a request to the controller, and the controller calculates the route and distributes it to the phase.
  • the controller calculates the route and distributes it to the phase.
  • the router (F): When another router receives the packet, it checks the destination ID field for the router's own ID, and the other router analyzes the packet's destination host ID and forwards it to the destination host.
  • the packet forwarding process by the centralized identity network routing protocol CIN is: (A) the controller sends a message to notify the second router to delete the H2 entry in the RHT, and second Router cache ⁇ HID2, RLIDx> to ILMD; (B): The ingress router sends a packet from H1 to H2, the packet is sent to the second router due to the lagging update of the ILMC of the ingress router, and the second router local RHT query fails the query.
  • the local ILMC if the query fails or is still the second router itself, the second router queries the RLID corresponding to the controller H2 and caches it to its own ILMC; (C): the second router sends a notification message to the ingress router to notify the ingress router H2 of the latest identity. a location mapping relationship, at the same time, the second router rewrites the destination ID of the data packet and routes to the router corresponding to the query controller H2; (D): the ingress router updates the identity-location mapping relationship of H2.
  • each router only needs to maintain the state of the directly connected router and does not need to calculate and store the autonomous domain routing table.
  • the reserved address starting with "010" of the identity/location layer is used as its identity/location identifier.
  • the invention has the beneficial effects that the invention realizes the compatibility of the identity location separation network with the current IPv6 network by adding the Ipv6 extension header, and does not need to modify the host protocol stack; and proposes a simple and effective method for not requiring the aggregation server. Mobility issues.
  • Figure 1 is a conventional HIP network protocol stack.
  • FIG. 2 is a schematic diagram of a CIN protocol stack of the present invention.
  • FIG. 3 is a schematic diagram of a CIN packet format of the present invention.
  • FIG. 4 is a schematic diagram of a HID/RLID format of the present invention.
  • FIG. 5 is a schematic diagram of a packet transmission format change according to the present invention.
  • Figure 6 is a schematic diagram of the CIN control forwarding separation of the present invention.
  • FIG. 7 is a block diagram showing the overall design of the CIN of the present invention.
  • FIG. 8 is a block diagram of the data packet transmission of the present invention.
  • FIG. 9 is a diagram of a packet forwarding process when the host of the present invention moves.
  • a centralized identity network routing protocol CIN including an application layer, a data link layer, and a physical layer, and an identity/location layer, the identity/location layer being located between an application layer and a data link layer;
  • the location layer decouples the identity attribute and location attribute of the IP address; the identity/location mapping entry of the host needs to be placed in the entire network identity location mapping database ILMD and the identity location mapping cache ILMC.
  • Each host and router has an independent and unique identifier.
  • the identity HID of the host is the identifier.
  • the location LOC of the host is the identifier RLID of the registered router.
  • the host uses its own HID as the source. ID, the destination host HID is used as the Destination ID. If the destination host is in the same LAN, it is directly transmitted to the destination host. If it is not in the same LAN, the routing protocol packet is transmitted.
  • the Source ID field and the Destination ID field must be rewritten as the source router and destination router ID RLID, and the source and destination hosts are added to the extension field and routed between the routers through the RLID.
  • the identity/location storage mapping system is responsible for storing and parsing the identity-to-location mapping, specifically: placing a full-network identity location mapping database ILMD on the control end, and placing an identity location mapping cache ILMC on the router side; When one router moves to another access router, the access router will transmit ⁇ HID, HLID> to ILMD. At the same time, ILMC is used as a subset of ILMD to input the packet to the access router to quickly resolve the identity location mapping. If there is no hit, the router will query the console for the mapping and cache it in ILMC.
  • the CIN protocol adopts a control/forward separation policy. Specifically, the controller adopts a corresponding routing policy according to the application and user requirements, and responds to the controller to query the request and deliver the routing entry.
  • the centralized identity network routing protocol CIN includes route calculation, the controller acquires LSAs from each router to establish a network-wide LSDB, and the LSAs contain link information of one router to another router; establish a matrix W, and W ij represents a router i
  • the link cost to router j by default, the Freud algorithm can be used to calculate the shortest path and the next hop of any two points.
  • Each application can customize the route as needed; the route is recalculated each time the LSA changes.
  • the packet forwarding process through the centralized identity network routing protocol CIN is: (A) H1 sends a data packet to its ingress router Ingress Router, and the source/destination ID field of the data packet needs to be HID1. And HID2; (B) each router maintains a recently used identity-location mapping cache ILMC. When the ingress router receives the H1 packet, it first queries the local ILMC table. If the query hits, go to step (D).
  • step (C) the ingress router obtains HID2 location information RLID2 by querying controller ILMD, further caches it to its ILMC;
  • the packet When the router's routing table, the packet will be hopped to another router, otherwise the router with the missing routing entry will send a request to the controller, the controller calculates the route and distributes it to the corresponding router; (F): When another route The receiver receives the packet, it checks the destination ID field for the router's own ID, and the other router analyzes the packet's destination host ID and forwards it to the destination host.
  • the packet forwarding process by the centralized identity network routing protocol CIN is: (A) the controller sends a message to notify the second router to delete the H2 entry in the RHT, and the second router caches ⁇ HID2, RLIDx> To ILMD; (B): the ingress router sends a packet from H1 to H2, and the second router local RHT query fails to query the local ILMC due to the lagging update of the ingress router's ILMC, if the query fails or Or the second router itself, the second router queries the RLID corresponding to the controller H2 and caches it to the own ILMC; (C): the second router sends a notification message to the ingress router to notify the ingress router H2 of the latest identity-location mapping relationship, and The second router rewrites the destination ID of the data packet and routes to the router corresponding to the query controller H2; (D): the ingress router updates the identity-location mapping relationship of H2.
  • Each router only needs to maintain the state of the directly connected router and does not need to calculate and store the autonomous domain routing table.
  • the reserved address starting with "010" in the identity/location layer is used as its identity/location identifier.
  • the present invention proposes a new centralized identity network routing protocol CIN (Centralized Identifier Network), which can improve the above disadvantages of the HIP protocol: the invention adopts the method of adding the Ipv6 extension header to realize the identity location.
  • the separation network is compatible with the current IPv6 network, and does not need to modify the host network protocol stack; the present invention proposes a simple and effective method for dealing with the mobility problem without the aggregation server RS; in the present invention, each router only needs to maintain direct connection The status of the router does not need to calculate and store the autonomous domain routing table (to the controller), which greatly reduces the router's computing storage resources.
  • CIN Centralized Identifier Network
  • Figure 2 shows the CIN protocol stack.
  • Figure 3 shows the CIN packet format.
  • Our design makes the host protocol stack not need to be modified and is IPv6 compliant.
  • the invention decouples the identity attribute and the location attribute of the IP address, and each host and router has an independent and unique identifier, the identity of the host (HID) is the identifier, and the location of the host (LOC) is the identifier of the registered router ( RLID).
  • the host handles the CIN packet and is no different from the current processing IP packet, but In the process of transmission, the following work needs to be done: when the host sends a CIN packet, the host uses its own HID as the Source ID, and uses the destination host HID as the Destination ID. If the destination host is in the same LAN, it directly transmits to the destination. The host, if not in the same LAN, must rewrite the Source ID field and the Destination ID field to the source router and destination router identifier (RLID) during the transmission of the packet, and the source and destination hosts are added to the extension field. Routing between routers through RLID, the packet format changes as shown in Figure 5.
  • RLID source router and destination router identifier
  • ILMD network-wide identity location mapping database
  • ILMC identity location mapping cache
  • the invention defines a centralized routing method for supporting mobility in an identity network, a centralized routing construction and maintenance of an identity controller, a notification node when the routing node updates the link state, and a centralized routing of the identity controller to perform a query request service.
  • the routing node simply maintains the link state of the direct-connected router. After the route controller implements the route calculation, it sends the routing entry mechanism to the routing node.
  • the host protocol stack is not modified, and the IPv6 extension field is added to implement the identity location separation.
  • the mobility support mechanism of the peer node is notified immediately upon moving.
  • the present invention employs a control/forward separation strategy.
  • the router does not have the topology discovery and route calculation function, but needs to query the controller for the next hop.
  • the controller can take the corresponding routing policy according to the application and user requirements, and respond to the controller to query the request and deliver the routing entry.
  • the controller obtains LSAs from each router to establish a network-wide LSDB.
  • the LSAs contain link information from one router to another. Therefore, a matrix W can be established, and W ij represents the link cost of the router i to the router j.
  • the Freud algorithm can be used to calculate the shortest path and the next hop of any two points, and each application can customize the route as needed. The route is recalculated each time the LSA changes.
  • SCN service bearer network
  • the format of the service bearer network is as follows:
  • Each router and the directly connected router periodically send ping packets (through the IdnRegisterPacket ping command, see section 3 of this section)) to check the connection status. If the ping packet is lost more than a certain number of times (can be set, the default is 5 times), then in the local adjacency list Set the neighbor state to NONACTIVE and delete the router in the IdnAdjPacket packet to update the controller LSDB information.
  • the logical network traffic is the total traffic of the data transmission packet of the local router to the logical network, and the unit is byte.
  • the logical network traffic of the controller is the sum of the traffic of all routers belonging to the logical network under the logical network.
  • the router cannot query the FIB entry (according to the destination host ID and LogicCode), and queries the controller for the routing table through the IDSQUERY command of IdnRegisterPacket (see 3).
  • the controller first queries the destination host ID directly through the ILMD table, and then finds the shortest path according to ⁇ LogicCode, source routing ID, destination routing ID ⁇ , and the result is obtained. Issued to each relevant router.
  • the commands are:
  • Commands 1 to 3 implement the identity registration of the host to the router, that is, the host adds an HRT entry to the router.
  • the host When the host registers with the router, when the host updates the HRT table, it will send a command 6 or 7 to update the mapping between the identity host ID and the direct router ID in the ILMD.
  • the logical network home extension header must be added to the extension header, otherwise it cannot be routed;
  • Each router can be customized to serve multiple logical networks, and packets generated by a certain logical network can only be routed within the logical network.
  • the data transmission packet starting from the host must be added with the logical network attribution number to be transmitted, otherwise it will be discarded.
  • the packets transmitted by process 1, 3 are IdnIngressPacket, and the packets transmitted by process 2 are IdnDataPacket.
  • Initial state first register for H1 to R1, R1 puts HID1 into its RHT table, and uploads ⁇ HID1, RLID1> to ILMD; the same process occurs at H2 and R2;
  • Step 1 H1 sends a data packet to its ingress router (R1), that is, R1, the source/destination ID field of the data packet needs to be HID1 and HID2;
  • Step 2 Each router maintains a recently used Identity-Location Map Cache (ILMC).
  • ILMC Identity-Location Map Cache
  • Step 3 R1 obtains the location information (RLID2) of HID2 by querying the controller ILMD. It further caches it to its ILMC;
  • Step 4 R1 rewrites the source ID of the data packet to RLID1, and the destination ID is RLID2; HID1 and HID2 are added to the extension field. The packet will be routed to R2.
  • Step 5 When there is a routing table to R2 in the node through which the R1 to R2 path passes, the packet will be hopped to R2. Otherwise, the router with the missing routing entry sends a request to the controller, which calculates the route and distributes it to the appropriate router.
  • Step 6 When R2 receives the packet, it checks the destination ID field as the router's own ID, and R2 analyzes the destination host ID of the packet and forwards it to the destination host.
  • H2 is moved from R2 to Rx, Rx adds HID2 to the RHT table and uploads to ILMD;
  • Step 1 The controller sends a message to notify R2 to delete the H2 entry in the RHT, and the R2 cache ⁇ HID2, RLIDx> to the ILMD;
  • Step 2 R1 sends a packet from H1 to H2, and the packet is sent to R2 due to the lagging update of the ILMC of R1 (the RLID of H2 is still buffered as R2).
  • the R2 local RHT query fails to query the local ILMC. If the query fails or is still R2 itself, R2 queries the RLID corresponding to the controller H2 (ie, RLIDx) and caches it to its own ILMC.
  • Step 3 R2 sends a notification message to R1 to notify R1H2 of the latest identity-location mapping relationship. At the same time, R2 rewrites the destination ID of the packet and routes it to Rx;
  • Step 4 R1 updates the identity-location mapping relationship of H2. All packets passing through R1 to H2 are then routed to Rx.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明涉及网络协议领域，其公开了一种集中式身份网络路由协议CIN：包括应用层、数据链路层和物理层，还包括身份/位置层，所述身份/位置层位于应用层和数据链路层之间；所述身份/位置层解耦IP地址的身份属性和位置属性；所述协议中主机的身份/位置映射项需放入全网身份位置映射数据库ILMD和身份位置映射缓存ILMC。本发明的有益效果是：本发明采用增加Ipv6扩展首部的方式实现身份位置分离网络与当前IPv6网络的兼容，且不需要修改主机协议栈；提出一种不需要汇聚服务器，简单且有效的方法应对移动性问题。

Description

一种集中式身份网络路由协议CIN 【技术领域】
本发明涉及网络协议领域，尤其涉及一种集中式身份网络路由协议CIN。
【背景技术】
软件定义网络SDN(Software Defined Network)——其核心为将传统交换机(路由器)设备进行“拆分”：传统的交换机功能由最底层的流量转发，以及更高级的其他处理功能(如网管控制、负载均衡等)这两部分组成，而SDN剥离了交换机除转发之外的所有高级处理功能，并且将这部分高级处理功能移到“控制器”中，实现控制面与转发面分离。两者之间通过特定协议进行通信确定流的转发规则，运行于控制器之上的网络操作系统(NOS)提供应用编程接口(API)给各类SDN应用以满足不同的网络需求，例如定制路由调度规则、网络隔离、流量管理、服务质量管理等业务。在SDN构架下，通过软件的编写、更新即可实现网络功能的整体升级，无需再针对每一个硬件设备进行配置，通过网络服务和应用程序的形式可直接部署实现网络配置，实现网络的可定制、易定制，同时加速网络部署周期。
互联网体系中的IP地址所拥有的双重语义(既代表节点的身份标识，也代表节点的位置标识。)是影响因特网规模和终端移动性的主要因素，身份位置分离技术应运而生。身份标识路由寻址是一种直接使用通信终端的身份标识进行路由寻址的技术，与位置标识(如：IP网络中的层次划分的地址)相比身份标识具有扁平性的特点。身份与位置分离技术采用两种标识进行通信，每一个终端都有一个全球唯一的身份标识，仅用来标记通信终端的身份信息，与其所在的位置信息没有相关性，而其位置标识与终端所在位置有关。终端移动后，其身份标识保持不变，位置标识由其所在路由提供，并实时更新到身份与位置映射关系数据库，终端A向终端B通信时，A的接入路由器获取所需的映射关系完成身份位置映射解析并将位置标识添加到数据包中，在不同路由器间通过位置标识进行数据包的转发。同时路由 器缓存映射关系以减少后续数据包的查询时延。身份标识的存在保证了终端标识的唯一性，位置标识的存在保证了路由表的可扩展性。两种标识实时映射与解析则保证了终端通信的连续性，也即移动性。
主机标识协议(HIP：Host Identity Protocol)在终端协议栈的网络层和传输层之间引入了新的协议层——主机标识层(Host Identity Layer,HIL)。在该协议层中，HIP用HI(Host Identity)来标识终端的身份，网络层中的IP地址作为位置标识只用来进行路由寻址，不再扮演主机名称的角色。引入主机标识协议层后,传输层不再与网络层耦合。主机标识层在逻辑上位于网络层与传输层之间,传输层使用<HI,端口>作为传输层标识符而不是用<IP地址,端口>,由主机标识层完成数据包中的主机标识符和IP地址转换。网络层对于传输层是屏蔽的,网络层的任何变化(例如,在通信过程中主机IP地址的变化)不会影响传输层链路,除非服务质量(QoS)发生变化。
在HIP名字空间中，HI代表一个全球唯一的静态的名字用来标志一个主机。一个主机可以有多个HI，HI的存储和查找可以由主机自己来完成，也可以由第三方如DNSSEC，PGP或X.509来执行。在移动环境下，由于节点具有移动性，它的IP地址必须保持更新。这种情况可以使用动态DNS服务器来提供相对静态的域名与动态IP地址的映射。但是在HIP环境中考虑到两个通信节点同时移动的情况，该协议设计了汇聚服务器RS(Rendezvous Server)来满足查询移动节点当前IP地址的需要。汇聚服务器是一个包转发器,用于移动主机移动到新的网络位置后通过汇聚服务器能够继续发送和接收数据包而不中断通信。汇聚服务器完成数据包的转发工作,而所有请求数据包转发的主机都叫汇聚客户机。为了完成数据包的转发,首先在DNS服务器上增加一条资源记录,存储汇聚服务器信息以供汇聚客户机查询；其次,汇聚客户机在汇聚服务器上注册自己的HI及IP地址等信息。主机要发送数据包时,先从DNS服务器查询汇聚服务器的HI及IP地址,并将第一个数据包发送到汇聚服务器,再由汇聚服务器将此数据包转发给目标地址,后续的数据包不再通过汇聚服务器转发而是通信双方直接进行通信,除非通信双方中有一方地址发生变化.在通 信过程中,如果有一方的网络地址发生变化时,将自己的地址变化同时通知汇聚服务器和对等节点.即使通信双方的网络地址同时发生变化时,可以通过汇聚服务器的数据包转发进行正常通信而不会中断传输层的数据链路。
由于需要在网络层和传输层插入一层，所以我们需要修改TCP/IP协议栈。一个非常重要的问题就是解决身份-位置映射的基础设施问题。HIP主要采用了两种解决方案：①利用动态DNS，使得DNS始终保持移动节点最新的IP地址以及对应的HI/HIT，但毫无疑问，主机频繁的移动会造成DNS服务器的性能急剧下降，以及DNS记录缓冲的不可使用问题，导致该方案较差的性能和可扩展性，另外该方案无法解决通信双方同时移动的问题；②采用RS，这就要求每个家乡网络都必须配置RS，不利于HIP的广泛部署和实施。全网拓扑在某一个时刻是唯一的，每个路由器独立实现全网拓扑的发现和路由计算，造成一定程度的计算冗余。
【发明内容】
为了解决现有技术中的问题，本发明提供了一种集中式身份网络路由协议CIN，解决现有技术中身份-位置映射的基础设施处理复杂的问题。
本发明提供了一种集中式身份网络路由协议CIN，包括应用层、数据链路层和物理层，还包括身份/位置层，所述身份/位置层位于应用层和数据链路层之间；所述身份/位置层解耦IP地址的身份属性和位置属性；主机的身份/位置映射项需放入全网身份位置映射数据库ILMD和身份位置映射缓存ILMC。
作为本发明的进一步改进：每台主机和路由器都有独立唯一的标识，主机的身份HID即为该标识，主机的位置LOC则为其注册路由器的标识RLID；当主机发送一个路由协议包，该主机用自己的HID作为Source ID，用目的主机HID作为Destination ID，如果目的主机在同一个局域网内，则直接传输到目的主机，如果不在同一个局域网内，在传输路由协议包的过程中必须把Source ID域和Destination ID域重写为源路由器和目的路由器标识RLID，而源主机和目的主机被添加到扩展字段中，在路由器之间通过RLID进行路由。
作为本发明的进一步改进：所述身份/位置存储解析系统负责对身份到位置映射存储和解析，具体为：在控制端放置了全网身份位置映射数据库ILMD，在路由器端放置了身份位置映射缓存ILMC；当一台主机从一台路由器移动到另一台接入路由器时IR，接入路由器会将<HID,HLID>传到ILMD，同时，ILMC作为ILMD的子集，用来给接入路由器输入包快速解析身份位置映射；如果没有命中，路由器会查询控制端该映射并缓存于ILMC。
作为本发明的进一步改进：CIN协议采用控制/转发分离策略，具体为：控制器根据应用和用户需求采取相应路由策略，响应控制器查询请求并下发路由表项。
作为本发明的进一步改进：所述的集中式身份网络路由协议CIN包括路由计算，控制器从各个路由器获取LSAs建立全网LSDB，LSAs包含了一个路由器到另一个路由器的链路信息；建立一个矩阵W，Wij表示路由器i到路由器j的链路开销，默认运用弗洛伊德算法即可算出任何两点的最短路径及下一跳，各应用可根据需要定制路由；每次LSA改变时都会重新计算路由。
作为本发明的进一步改进：各个主机均处于稳定位置时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)H1发送数据包给它的入口路由器Ingress Router，数据包的源/目的ID域需为HID1和HID2；(B)每个路由器维持一个最近使用的身份-位置映射缓存ILMC,当入口路由器收到H1的数据包，它先在本地ILMC表中查询，如果查询命中，转到步骤(D)，不命中，转到步骤(C)；(C)入口路由器通过查询控制器ILMD获得HID2的位置信息RLID2，进一步将它缓存到它的ILMC；(D)：入口路由器重写该数据包的源ID为RLID1，目的ID为RLID2；HID1和HID2被添加到扩展字段中，数据包会被路由到另一路由器；(E)：当入口路由器到另一路由器路径经过的节点中都有到另一路由器的路由表时，包会逐跳传到另一路由器，否则缺失路由表项的路由器会发送请求到控制器，控制器计算路由并分发到相应路由器；(F)：当另一路由器接收到包，它检查目的ID域为路由器自己的ID，另一路由器分析包的目的主机ID并转发到目的主机。
作为本发明的进一步改进：当主机处于移动状态时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)控制器发送消息通知第二路由器删除RHT中H2表项，第二路由器缓存<HID2,RLIDx>到ILMD；(B)：入口路由器发送从H1到H2的数据包，由于入口路由器的ILMC的滞后更新发送该数据包到第二路由器，第二路由器本地RHT查询失败查询本地ILMC，如果查询失败或者还是第二路由器本身，则第二路由器查询控制器H2对应的RLID并缓存到自身ILMC；(C)：第二路由器发送通知消息到入口路由器通知入口路由器H2的最新身份-位置映射关系，同时，第二路由器重写该数据包的目的ID并路由到查询控制器H2对应的路由器；(D)：入口路由器更新H2的身份-位置映射关系。
作为本发明的进一步改进：每个路由器只需维护直连路由器的状态且不需计算、存储自治域路由表。
作为本发明的进一步改进：所述身份/位置层以“010”开头的保留地址用来当做其身份/位置标识。
本发明的有益效果是：本发明采用增加Ipv6扩展首部的方式实现身份位置分离网络与当前IPv6网络的兼容，且不需要修改主机协议栈；提出一种不需要汇聚服务器，简单且有效的方法应对移动性问题。
【附图说明】
图1是传统HIP网络协议栈。
图2是本发明CIN协议栈示意图。
图3是本发明CIN包格式示意图。
图4是本发明HID/RLID格式示意图。
图5是本发明数据包传输格式变化示意图。
图6是本发明CIN控制转发分离示意图。
图7是本发明CIN总体设计框图。
图8是本发明数据包传输框图。
图9是本发明目的主机移动时数据包转发过程图。
【具体实施方式】
下面结合附图说明及具体实施方式对本发明进一步说明。
缩略语和关键术语定义
SDN    Software Defined Network         软件定义网络
HIP    Host Identity Protocol           主机标识协议
HID    Host IDentifier                  主机标识
RLID Router Location Identifier      路由器位置标识
LSDB Link State DataBase             链路状态数据库
ILMD Identifier-to-Locator Mapping Database  身份-位置映射数据库
ILMC Identifier-to-Locator Mapping Cache     身份-位置映射缓存
RHT    Registered Host Table                  注册主机表
PCNL Physically Connected Neighbors List   直连邻居表
DNS    Domain Name System                  域名系统
CIN    Centralized Identifier Network       集中式身份网络
一种集中式身份网络路由协议CIN，包括应用层、数据链路层和物理层，还包括身份/位置层，所述身份/位置层位于应用层和数据链路层之间；所述身份/位置层解耦IP地址的身份属性和位置属性；主机的身份/位置映射项需放入全网身份位置映射数据库ILMD和身份位置映射缓存ILMC。
每台主机和路由器都有独立唯一的标识，主机的身份HID即为该标识，主机的位置LOC则为其注册路由器的标识RLID；当主机发送一个路由协议包，该主机用自己的HID作为Source ID，用目的主机HID作为Destination ID，如果目的主机在同一个局域网内，则直接传输到目的主机，如果不在同一个局域网内，在传输路由协议包的过 程中必须把Source ID域和Destination ID域重写为源路由器和目的路由器标识RLID，而源主机和目的主机被添加到扩展字段中，在路由器之间通过RLID进行路由。
所述身份/位置存储映射系统负责对身份到位置映射存储和解析，具体为：在控制端放置了全网身份位置映射数据库ILMD，在路由器端放置了身份位置映射缓存ILMC；当一台主机从一台路由器移动到另一台接入路由器时IR，接入路由器会将<HID,HLID>传到ILMD，同时，ILMC作为ILMD的子集，用来给接入路由器输入包快速解析身份位置映射；如果没有命中，路由器会查询控制端该映射并缓存于ILMC。
CIN协议采用控制/转发分离策略，具体为：控制器根据应用和用户需求采取相应路由策略，响应控制器查询请求并下发路由表项。
所述的集中式身份网络路由协议CIN包括路由计算，控制器从各个路由器获取LSAs建立全网LSDB，LSAs包含了一个路由器到另一个路由器的链路信息；建立一个矩阵W，Wij表示路由器i到路由器j的链路开销，默认运用弗洛伊德算法即可算出任何两点的最短路径及下一跳，各应用可根据需要定制路由；每次LSA改变时都会重新计算路由。
各个主机均处于稳定位置时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)H1发送数据包给它的入口路由器Ingress Router，数据包的源/目的ID域需为HID1和HID2；(B)每个路由器维持一个最近使用的身份-位置映射缓存ILMC,当入口路由器收到H1的数据包，它先在本地ILMC表中查询，如果查询命中，转到步骤(D)，不命中，转到步骤(C)；(C)入口路由器通过查询控制器ILMD获得HID2的位置信息RLID2，进一步将它缓存到它的ILMC；(D)：入口路由器重写该数据包的源ID为RLID1，目的ID为RLID2；HID1和HID2被添加到扩展字段中，数据包会被路由到另一路由器；(E)：当入口路由器到另一路由器路径经过的节点中都有到另一路由器的路由表时，包会逐跳传到另一路由器，否则缺失路由表项的路由器会发送请求到控制器，控制器计算路由并分发到相应路由器；(F)：当另一路由器接收到包，它检查目的ID域为路由器自己的ID，另一路由器分析包的目的主机ID并转发到目的主机。
当主机处于移动状态时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)控制器发送消息通知第二路由器删除RHT中H2表项，第二路由器缓存<HID2,RLIDx>到ILMD；(B)：入口路由器发送从H1到H2的数据包，由于入口路由器的ILMC的滞后更新发送该数据包到第二路由器，第二路由器本地RHT查询失败查询本地ILMC，如果查询失败或者还是第二路由器本身，则第二路由器查询控制器H2对应的RLID并缓存到自身ILMC；(C)：第二路由器发送通知消息到入口路由器通知入口路由器H2的最新身份-位置映射关系，同时，第二路由器重写该数据包的目的ID并路由到查询控制器H2对应的路由器；(D)：入口路由器更新H2的身份-位置映射关系。
每个路由器只需维护直连路由器的状态且不需计算、存储自治域路由表。
所述身份/位置层以“010”开头的保留地址用来当做其身份/位置标识。
本发明借鉴SDN中控制-转发分离思想，提出一种新的集中式身份网络路由协议CIN(Centralized Identifier Network)，它能够改进HIP协议的上述缺点：本发明采用增加Ipv6扩展首部的方式实现身份位置分离网络与当前IPv6网络的兼容，且不需要修改主机网络协议栈；本发明提出一种不需要汇聚服务器RS，简单且有效的方法应对移动性问题；本发明中每个路由器只需维护直连路由器的状态且不需计算、存储自治域路由表(交给控制器)，大大降低路由器计算存储资源。
在一实施例中，图2给出CIN协议栈，图3给出CIN包格式，我们的设计使得主机协议栈不需要做任何改动且兼容IPv6，我们用身份/位置层代替传统网络IP层。本发明解耦IP地址的身份属性和位置属性，每台主机和路由器都有独立唯一的标识，主机的身份(HID)即为该标识，主机的位置(LOC)则为其注册路由器的标识(RLID)。为兼容传统IPv6，我们将IPv6以“010”开头的保留地址用来当做CIN的身份/位置标识，即"4XXX:XXXX:XXXX:XXXX"/"5XXX:XXXX:XXXX:XXXX"，X为0-f，如图4。
根据这些定义，主机处理CIN包与现在处理IP包并无不同，但 是在传输的过程中需要做如下工作：当主机发送一个CIN包，该主机用自己的HID作为Source ID，用目的主机HID作为Destination ID，如果目的主机在同一个局域网内，则直接传输到目的主机，如果不在同一个局域网内，在传输包的过程中必须把Source ID域和Destination ID域重写为源路由器和目的路由器标识(RLID)，而源主机和目的主机被添加到扩展字段中，在路由器之间通过RLID进行路由，数据包格式变化如图5所示。
身份位置分离网络的一个问题是身份到位置映射存储和解析的过程。本发明中我们在控制端放置了全网身份位置映射数据库(ILMD)，在路由器端放置了身份位置映射缓存(ILMC)。当一台主机从一台路由器移动到另一台接入路由器时(IR)，接入路由器会将<HID,HLID>传到ILMD。ILMC作为ILMD的子集，用来给接入路由器输入包快速解析身份位置映射，如果没有命中，路由器会查询控制端该映射并缓存于ILMC。
本发明定义了身份网络中一种支持移动性的集中式路由方法，身份控制器集中式路由构建和维护，路由节点更新链路状态时通知控制器，身份控制器集中式路由进行查询请求服务，路由节点简单维护与直连路由器链路状态，身份控制器实现路由计算后向路由节点下发路由表项机制，无需修改主机协议栈，增加IPv6扩展字段实现身份位置分离，同时无需汇聚服务器，无需在移动时立刻通知对等节点的移动性支持机制。
本发明采用控制/转发分离策略。如图6，路由器并无拓扑发现、路由计算功能，而需要询问控制器相应下一跳，控制器可以根据应用和用户需求采取相应路由策略，响应控制器查询请求并下发路由表项。
总体设计框图如图7。
路由计算：
控制器从各个路由器获取LSAs建立全网LSDB，LSAs包含了一个路由器到另一个路由器的链路信息。因此可以建立一个矩阵W，Wij表示路由器i到路由器j的链路开销，默认运用弗洛伊德算法即可算出任何两点的最短路径及下一跳，各应用可根据需要定制路由。每次LSA改变时都会重新计算路由。
我们规定
Figure PCTCN2015096860-appb-000001
FLOYD-WARSHALL算法是一种动态规划算法。设
Figure PCTCN2015096860-appb-000002
为从节点i到j的所有中间节点全部取自集合{1,2,…,k}的一条最短路径权重。当k＝0时，从节点i到节点j的一条不包括编号大于0的中间节点的路径将没有任何中间节点，因此，
Figure PCTCN2015096860-appb-000003
递归定义
Figure PCTCN2015096860-appb-000004
如下：
Figure PCTCN2015096860-appb-000005
矩阵
Figure PCTCN2015096860-appb-000006
就是最后的最短路径。
以下定义一种实现的主要数据包格式：
在IPV6的基础上新定义了四种扩展首部：服务承载网(SCN)首部、逻辑网归属首部、源主机ID首部、目的主机ID首部。
服务承载网的格式如下：
Figure PCTCN2015096860-appb-000007
目前CIN中定义了五种数据包类型，其功能和类型代码如下：
Figure PCTCN2015096860-appb-000008
1)IdnAdjPacket
数据格式：逻辑网数量|逻辑网1|逻辑网1流量|逻辑网2|逻辑网2流量|……|邻居数量|邻居1身份|接入邻居1端口ip|传输时间|邻居2身份|…….
各路由器与直连路由器周期发送ping包(通过IdnRegisterPacket ping命令，见本小节3))查看其连接状态，如果ping包丢失超过一定次数(可设置，默认为5次)，则在本地邻接表中置邻居状态为NONACTIVE，并在IdnAdjPacket包中删去该路由器以更新控制器LSDB信息。
逻辑网流量为本地路由器于该逻辑网的数据传输包的总流量，单位为字节。控制器的逻辑网流量为属于该逻辑网的所有路由器在该逻辑网下的流量之和。
2)IdnRoutePacket
数据格式：目的身份ID|下一跳端口ip|逻辑网LogicCode
路由器查询不到FIB表项(根据目的主机ID和LogicCode)，通过IdnRegisterPacket的IDSQUERY命令(见3))向控制器查询路由表。控制器根据{LogicCode,源路由ID，目的主机ID}三元组，首先通过ILMD表查询该目的主机ID直连路由器，再根据{LogicCode,源路由ID，目的路由ID}查找最短路径，将结果下发到各个相关路由器。
3)IdnRegisterPacket
其中命令有：
Figure PCTCN2015096860-appb-000009
Figure PCTCN2015096860-appb-000010
命令1～3实现主机向路由器的身份注册工作，即主机向路由器添加HRT表项；
命令4实现路由器握手协议；
命令5的工作流程查看2.2.22)节，格式为”IDSQUERY:”+”目的路由器ID”；
在主机向路由器注册时，主机更新HRT表的时候会相应发送命令6或7来更新ILMD中身份主机ID与直连路由器ID的映射。
4)IdnIngressPacket/IdnDataPacket
必须在扩展首部加上逻辑网归属扩展首部，否则无法路由；
逻辑网归属的格式如下：
Figure PCTCN2015096860-appb-000011
各路由器可以被定制服务于多个逻辑网，某个逻辑网产生的数据包只能在本逻辑网内被路由。始于主机的数据传输包必须加上逻辑网归属号才能传输，否则则被丢弃。
IdnDataPacket需要额外的两个扩展字段以标识包的源主机ID和目的主机ID以正确路由到目的主机：
Figure PCTCN2015096860-appb-000012
如图5：过程1,3传输的包为为IdnIngressPacket，过程2传输的包为IdnDataPacket。
以下定义CIN数据包转发过程：
1)首先讨论一般情况：即主机均处于稳定位置的数据包转发过程:
初始状态：首先为H1向R1注册，R1将HID1放到它的RHT表，并上传<HID1,RLID1>对到ILMD；同样的过程发生在H2和R2；
步骤1：H1发送数据包给它的入口路由器(Ingress Router)，即R1，数据包的源/目的ID域需为HID1和HID2；
步骤2：每个路由器维持一个最近使用的身份-位置映射缓存(ILMC)。当R1收到H1的数据包，它先在本地ILMC表中查询，如果查询命中，转到步骤4，不命中，转到步骤3；
步骤3：R1通过查询控制器ILMD获得HID2的位置信息(RLID2)。它进一步将它缓存到它的ILMC；
步骤4：R1重写该数据包的源ID为RLID1，目的ID为RLID2；HID1和HID2被添加到扩展字段中。数据包会被路由到R2。
步骤5：当R1到R2路径经过的节点中都有到R2的路由表时，包会逐跳传到R2。否则缺失路由表项的路由器会发送请求到控制器，控制器计算路由并分发到相应路由器。
步骤6：当R2接收到包，它检查目的ID域为路由器自己的ID，R2分析包的目的主机ID并转发到目的主机。
2)其次讨论当主机处于移动状态时数据包转发流程：
初始状态：H2由R2移动到Rx，Rx将HID2加到RHT表并上传到ILMD；
步骤1：控制器发送消息通知R2删除RHT中H2表项，R2缓存<HID2,RLIDx>到ILMD；
步骤2：R1发送从H1到H2的数据包，由于R1的ILMC的滞后更新(H2的RLID仍缓存为R2)发送该数据包到R2。R2本地RHT查询失败查询本地ILMC，如果查询失败或者还是R2本身，则R2查询控制器H2对应的RLID(即RLIDx)并缓存到自身ILMC；
步骤3：R2发送通知消息到R1通知R1H2的最新身份-位置映射关系。同时，R2重写该数据包的目的ID并路由到Rx；
步骤4：R1更新H2的身份-位置映射关系。接下来所有通过R1到H2的包都被路由到Rx。
以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。对于本发明所属技术领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干简单推演或替换，都应当视为属于本发明的保护范围。
Figure PCTCN2015096860-appb-000014

Claims (5)

  1. 一个路由器的链路信息；建立一个矩阵W，Wij表示路由器i到路由器j的链路开销，默认运用弗洛伊德算法即可算出任何两点的最短路径及下一跳，各应用可根据需要定制路由；每次LSA改变时都会重新计算路由。
  2. 根据权利要求1所述的集中式身份网络路由协议CIN，其特征在于：各个主机均处于稳定位置时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)H1发送数据包给它的入口路由器Ingress Router，数据包的源/目的ID域需为HID1和HID2；(B)每个路由器维持一个最近使用的身份-位置映射缓存ILMC,当入口路由器收到H1的数据包，它先在本地ILMC表中查询，如果查询命中，转到步骤(D)，不命中，转到步骤(C)；(C)入口路由器通过查询控制器ILMD获得HID2的位置信息RLID2，进一步将它缓存到它的ILMC；(D)：入口路由器重写该数据包的源ID为RLID1，目的ID为RLID2；HID1和HID2被添加到扩展字段中，数据包会被路由到另一路由器；(E)：当入口路由器到另一路由器路径经过的节点中都有到另一路由器的路由表时，包会逐跳传到另一路由器，否则缺失路由表项的路由器会发送请求到控制器，控制器计算路由并分发到相应路由器；(F)：当另一路由器接收到包，它检查目的ID域为路由器自己的ID，另一路由器分析包的目的主机ID并转发到目的主机。
  3. 根据权利要求1所述的集中式身份网络路由协议CIN，其特征在于：当主机处于移动状态时，通过该集中式身份网络路由协议CIN进行数据包转发过程为：(A)控制器发送消息通知第二路由器删除RHT中H2表项，第二路由器缓存<HID2,RLIDx>到ILMD；(B)：入口路由器发送从H1到H2的数据包，由于入口路由器的ILMC的滞后更新发送该数据包到第二路由器，第二路由器本地RHT查询失败查询本地ILMC，如果查询失败或者还是第二路由器本身，则第二路由器查询控制器H2对应的RLID并缓存到自身ILMC；(C)：第二路由器发送通知消息到入口路由器通知入口路由器H2的最新身份-位置映射关系，同时，第二路由器重写该数据包的目的ID并路由到查询控制器H2对应的路由器；(D)：入口路由器更新H2 的身份-位置映射关系。
  4. 根据权利要求6或7任一所述的集中式身份网络路由协议CIN，其特征在于：每个路由器只需维护直连路由器的状态且不需计算、存储自治域路由表。
  5. 根据权利要求1所述的集中式身份网络路由协议CIN，其特征在于：所述身份/位置层以“010”开头的保留地址用来当做其身份/位置标识。
PCT/CN2015/096860 2015-12-09 2015-12-09 一种集中式身份网络路由协议cin WO2017096565A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2015/096860 WO2017096565A1 (zh) 2015-12-09 2015-12-09 一种集中式身份网络路由协议cin

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2015/096860 WO2017096565A1 (zh) 2015-12-09 2015-12-09 一种集中式身份网络路由协议cin

Publications (1)

Publication Number Publication Date
WO2017096565A1 true WO2017096565A1 (zh) 2017-06-15

Family

ID=59012527

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/096860 WO2017096565A1 (zh) 2015-12-09 2015-12-09 一种集中式身份网络路由协议cin

Country Status (1)

Country Link
WO (1) WO2017096565A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114157724A (zh) * 2021-11-23 2022-03-08 之江实验室 一种基于源路由的身份/位置标识混合转发方法
CN115150314A (zh) * 2021-03-31 2022-10-04 腾讯科技（深圳）有限公司 跨网络域的数据包传输方法和装置、存储介质及电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102123182A (zh) * 2011-04-09 2011-07-13 山东师范大学 一种基于ipv6地址的主机身份标识与位置标识分离的方法
CN102938885A (zh) * 2011-08-16 2013-02-20 中兴通讯股份有限公司 身份位置分离与传统网络互联互通方法、ilr和asr
CN102957621A (zh) * 2011-08-31 2013-03-06 上海贝尔股份有限公司 一种基于位置和身份标识分离的通信网络系统及其设备
US20150023262A1 (en) * 2011-11-15 2015-01-22 Zte Corporation Device and method for realizing identity and locator separation network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102123182A (zh) * 2011-04-09 2011-07-13 山东师范大学 一种基于ipv6地址的主机身份标识与位置标识分离的方法
CN102938885A (zh) * 2011-08-16 2013-02-20 中兴通讯股份有限公司 身份位置分离与传统网络互联互通方法、ilr和asr
CN102957621A (zh) * 2011-08-31 2013-03-06 上海贝尔股份有限公司 一种基于位置和身份标识分离的通信网络系统及其设备
US20150023262A1 (en) * 2011-11-15 2015-01-22 Zte Corporation Device and method for realizing identity and locator separation network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
XU DONGXIAO ET AL.: "A novelscheme for seperating location identity and identification", COMPUTER APPLICATIONS AND SOFTWARE, vol. 27, no. 2, pages 20100228 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115150314A (zh) * 2021-03-31 2022-10-04 腾讯科技（深圳）有限公司 跨网络域的数据包传输方法和装置、存储介质及电子设备
CN115150314B (zh) * 2021-03-31 2023-08-25 腾讯科技（深圳）有限公司 跨网络域的数据包传输方法和装置、存储介质及电子设备
CN114157724A (zh) * 2021-11-23 2022-03-08 之江实验室 一种基于源路由的身份/位置标识混合转发方法

Similar Documents

Publication Publication Date Title
US9621508B2 (en) System and method for sharing VXLAN table information with a network controller
EP2099175B1 (en) Method for advertising route message, method, system and device for routing packets
EP2974133B1 (en) Method and system for controlling an underlying physical network by a software defined network
US9178818B2 (en) Communication apparatus
CN104734963B (zh) 一种基于SDN的IPv4和IPv6网络互连方法
US10742697B2 (en) Packet forwarding apparatus for handling multicast packet
WO2009012668A1 (en) Network architecture of mutiple address spaces, and method for host information register and data transmission
EP2456130A1 (en) System for network deployment and method for mapping and data forwarding thereof
WO2014132286A1 (en) Method for link failure detection and session transfer to a lively link in the multihoming environment of id/locator split-based networks
KR101381701B1 (ko) 데이터 메시지 처리 방법, 시스템 및 접속 서비스 노드
WO2011124132A1 (zh) 数据通信系统及方法
WO2010139115A1 (zh) 多个汇聚节点共同处理移动组播源组播业务的方法及装置
US20150055656A1 (en) Network appliance redundancy system, control apparatus, network appliance redundancy method and program
WO2016058261A1 (zh) 一种基于网络的扁平路由的实现方法
WO2017096565A1 (zh) 一种集中式身份网络路由协议cin
Cabellos et al. An Architectural Introduction to the Locator/ID Separation Protocol (LISP)
Dumba et al. A virtual ID routing protocol for future dynamics networks and its implementation using the SDN paradigm
US11196666B2 (en) Receiver directed anonymization of identifier flows in identity enabled networks
WO2011124121A1 (zh) 网间数据通讯系统及方法
JP5971348B2 (ja) 通信システム、制御装置及び制御方法
WO2015039563A1 (zh) 实现三层虚拟专用网的方法和设备
WO2017107033A1 (zh) 一种基于身份的集中控制式网络体系架构iccn
WO2018230608A1 (ja) 通信システム、通信制御装置、スイッチ装置、通信制御方法、及び、記録媒体
Lee et al. A Simple Cost Analysis of Host ID-LOC Separating protocol using SDN Features
Vistro et al. A Review and Comparative Analysis of Routing Protocols in Network

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15910040

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 04/10/2018)

122 Ep: pct application non-entry in european phase

Ref document number: 15910040

Country of ref document: EP

Kind code of ref document: A1