Keywords: IPv6 routing protocol
China Library Classification. : TP393 document identification code: a article number:1672-3791(2012) 05 (b)-0013-02 At present, various Internet-based applications are developing rapidly, which leads to a surge in Internet terminal visits. However, IPv4 currently used by the Internet is struggling due to its own problems. In order to completely solve the problems existing in IPv4, the International Internet Engineering Task Force began to study the next generation IP protocol, namely IPv6, from 1995. The address length of IPv6 is 128 bits, so the routing protocol based on IPv6 should be changed accordingly.
OSPFv3, RIPng and IS-ISv6 routing protocols under IPv6 are all commonly used intra-domain routing protocols at present. This paper will analyze and compare these protocols.
OSPFv3 routing protocol under 1 IPv6.
OSPFv3 is the third version of Open Shortest Path First defined in RFC2740. During the transition from IPv4 to IPv6, OSPFv3 has become the core routing technology in IPv6 network and the mainstream protocol of dynamic routing in the next generation network [1]. OSPFv3 basically continues the framework of OSPFv2, but it has been modified according to the characteristics of IPv6, mainly in the following aspects.
(1) supports running multiple instances on a link. OSPFv3 supports running multiple OSPF instances on a link through the InstanceID of the datagram header, while in OSPFv2, the same purpose is achieved through the authentication field of the datagram header.
(2) Use of link local address. In order to realize "neighbor discovery" and "automatic configuration", OSPFv3 uses local link addresses on separate links. Instead of forwarding packets with local link source addresses, IPv6 routers use the router's link-local IPv6 address (starting with FF80::/ 10) as the source address and next-hop address.
(3)OSPFv3 also redefines the role of LSA. In OSPFv3, the task of announcing network topology and IPv6 address information is assigned to newly introduced and existing LSAs. This can make OSPFv3 support new network protocols more conveniently, and also make OSPFv3 have good universality, which can improve the performance of IPv6 with little network upgrade and no major protocol migration.
(4)OSPFv3 also adds a variety of optional functions, and adds many optional function control areas in the message to realize the universality of routing protocols.
In a word, the difference between OSPFv3 and previous protocols is that it simplifies the message structure by providing inherent security. OSPFv3 messages can be authenticated and encrypted by using security extension headers in IPv6 messages.
RIPng protocol under IPv6.
RIPng(RIP Next Generation) is a modification based on RIP-2 protocol and a new version defined for IPv6. RIP is suitable for small similar networks and is a typical distance vector protocol. Under RIP routing protocol, routers send their own routing tables to other routers every 30s. If they don't receive the neighbor routing update message within 3min, they are marked as unreachable. If the network neighbor's routing update message is not received within 120s, it will be deleted from the routing table. This process ensures the maintenance and reliability of neighbors [2].
RIPng routes information through UDP message switching, and the end slogan used is 52 1. RIPng uses hop count to measure the distance to the destination address, also known as metric. In RIPng, the number of hops from the router to the directly connected network is 0, the number of hops from the network accessible through the router is 1, and so on. In order to limit the convergence time, RIPng stipulates that the metric value should be an integer between 0 and 15, and the number of hops greater than or equal to 16 is defined as infinity, that is, the destination network or host is unreachable.
RIPng retains the main features of RIPv2, but makes corresponding changes in the message format: First, because IPv6 adopts the address prefix, RIPng deletes the subnet mask field in RIPv2. Secondly, RIPng updates the next hop field in RIPv2 to the next hop RTE (routing table entry). In addition, RIPv2 stipulates that each message can carry up to 25 RTEs, but RIPng no longer limits this item, but is determined by the maximum transmission unit of the transmission medium.
The routing protocol based on distance vector algorithm will have the problems of slow convergence and infinite count, which will lead to the inconsistency of routing. RIPng uses horizontal segmentation technology, toxicity inversion technology and trigger update technology to solve these problems.
In a word, RIPng is still used AS an internal gateway protocol in small and medium-sized AS.
IS-Isv6 protocol under IPv6.
Is-is (Intermediate System to Intermediate System Routing Protocol) was originally a dynamic routing protocol designed by ISO for its connectionless network protocol CLNP [3]. IS-IS is a link-state protocol, which uses the shortest path first (SPF) algorithm for routing calculation.
IS-IS protocol learns the network topology of the whole autonomous system by exchanging routing information with other routers in the autonomous system (AS). The routing information of the whole Internet can be obtained through the routing information of other autonomous systems injected by the border router of the autonomous system.
With the construction of IPv6 network, dynamic routing protocol is also needed to provide accurate and effective routing information for IPv6 packet forwarding. IS-IS routing protocol, combined with its good scalability, supports IPv6 network layer protocol and can discover and generate IPv6 routes. IS-IS routing protocol supporting IPv6 protocol is also called IS-ISv6 dynamic routing protocol. In order to support the operation in IPv6 environment and guide the forwarding of IPv6 messages, IS-ISv6 simply extends IS-IS to enable it to handle the routing information of IPv6.
(1) CLV joins IS-ISV6: IETF stipulates that IS-IS supports IPv6, and mainly adds two CLVs (code-length-value). The value in the NLPID field of CLV supporting protocol is 0×8 1, which indicates that IS-ISV6 functions are supported.
(2)IS-ISv6 adjacency: IS-IS uses Hello messages to discover neighbor routers on the same link and establish adjacency. After the adjacency is established, it will continue to send Hello messages periodically to maintain adjacency. In order to support IPv6 routing and establish IPv6 adjacency, IS-ISv6 extends the Hello message: (1) adds an 8-8-bit NLPID to CLV, indicating that the current router supports IS-ISv6 function. (2) Add the CLV of the IPv6 address of the interface with IS-ISv6 function in the Hello message, and fill in the local address of the IPv6 link of the interface with IS-ISv6 function in the interface address field.
4 Comparison of three protocols
OSPFv3 adds some functions that OSPFv2 does not have, making it more universal and convenient for future protocol upgrade. OSPFv3 is mature, popular, easy to use and maintain, universal and extensible. The algorithm of RIPng is relatively simple, which makes the configuration and maintenance of RIPng network relatively simple. It is suitable for small networks with relatively simple network topology and extremely low data link failure rate. For ordinary users, such as Internet cafes, schools, institutions and other small and medium-sized enterprises, using this RIPng router can obtain higher economic benefits. In large networks, RIPng is generally not used, but OSPF with relatively complex protocol is used. IS-IS only needs to run one process on one router, so it can support both Ipv4 and Ipv6 topology calculation, and takes up less resources. The disadvantage is that the collapse of any protocol will lead to the collapse of another protocol, which is not flexible enough.
refer to
[1] Tang Yongzheng, Zhou Dawei. Research on Routing Protocol Based on Ipv6 [J]. Journal of Yancheng Institute of Technology (Natural Science Edition), 201(1).
, Huang Hua, Sun. Analysis and Comparison of Two Dynamic Routing Protocols in Large-scale Networks [J]. Science and Technology and Engineering, 2006 (9): 6 1 ~ 63.
Li Zhongnian. Analysis of EIGRP Routing Protocol [J]. China Data Communication, 2005 (6):112 ~114.