1. Address resources are about to be exhausted: the number of IP address digits provided by IPv4 is 32 bits, that is, about 100 million addresses. As the number of hosts connected to the Internet increases rapidly, predictions indicate that all IPv4 addresses will be allocated between 2005 and 2010.
2. The routing table is getting bigger and bigger: Since IPv4 uses a form that is independent of the network topology structure to allocate addresses, as the number of connected networks increases, the number of routers increases rapidly. Correspondingly, the routing table that determines the data transmission route continues to increase.
3. Lack of service quality assurance: IPv4 follows the Best Effort principle, which is an advantage because it makes IPv4 simple and efficient; on the other hand, it lacks effective support for new business types emerging on the Internet, such as real-time and multimedia applications, which require certain service quality assurance, such as bandwidth, latency and jitter.
4. Inconvenient address allocation: IPv4 adopts manual configuration to assign addresses to users, which not only increases the complexity of management and planning, but also is not conducive to providing better services to users who need IP mobility.
IPv6 can solve many problems in IPv4, such as address shortage, service quality assurance, etc. At the same time, IPv6 has also made a lot of improvements to IPv4, including routing and automatic network configuration. IPv6 and IPv4 will coexist for several years during the transition period, and IPv6 will gradually replace IPv4.
1. IPv6 core technology
In terms of address length, compared with IPv4, an obvious improvement in IPv6 is that the length of 128-bit address of IPv6 can provide sufficient address space. At the same time, it also provides different types of address configurations for the host interface, including: global address, global unicast address, regional address, link local address, regional address, broadcast address, multicast group address, any broadcast address, mobile address, hometown address, transfer address, etc. Another basic feature of IPv6 is that it supports two automatic configuration methods of stateless and stateful addresses. The automatic configuration method of stateless address is: nodes that need to configure the address use a neighbor discovery mechanism to obtain a local connection address. Once this address is obtained, it uses another plug-and-play mechanism to obtain a globally unique routing address without any manual intervention. In addition, IPv6 also shows high characteristics in the following aspects.
1. Service quality
The format of IPv6 packets includes an 8-bit service flow category (Class) and a new 20-bit stream label (Flow Label). Its purpose is to allow the source node that sends the service flow and the router that forwards the service flow to mark the data packet. After receiving a data packet, the intermediate node can determine which stream it belongs to by verifying its flow tag, and then knows the QoS requirements of the data packet and forward it quickly.
2. Safety aspects
In terms of security, IPv6 is more closely integrated with IP security (IPSec) mechanism and services. Although both IP standards currently support IPsec (IP security protocol), IPv6 regards security as an organic part of its own standards. Security deployment is at a more coordinated and unified level, rather than using superimposed solutions like IPv4. IPsec in IPv6 can provide encryption/authorization of communications on the IP layer (that is, all applications running on the IP layer), seamless access to remote enterprise intranets (such as enterprise VPN networks), and permanent connections can be achieved. In addition to this mandatory security mechanism, IPSec also provides two services. Authentication header (AH) is used to ensure data consistency, while encapsulated security load header (ESP) is used to ensure data confidentiality and data consistency. In IPv6 packages, both AH and ESP are extended headers, which can be used at the same time or one of them can be used separately. As an important application of IPSec, IPv6 integrates the functions of virtual private network (VPN).
3. Mobile IPv6 aspects
Mobility is undoubtedly one of the most exciting services on the Internet. The mobile IPv6 protocol provides users with mobile IP data services, allowing users to use the same IPv6 address all over the world, which is very suitable for future wireless Internet access.
Mobility support in IPv6 is embedded in the IP protocol as a necessary protocol while formulating the IPv6 protocol. Different from IPv4's mobility support is proposed as a function attached to the IP protocol. Not all IPv4 implementations can provide support for mobility, and its efficiency is not as high as mobile IPv6. More importantly, IPv4's limited address space resources cannot provide the IP addresses required by all potential mobile terminal devices, making it difficult to implement large-scale application of mobile IP. Compared with IPv4, IPv6's mobility support cancels off-site proxying, fully supports routing optimization, completely eliminates triangular routing problems, and provides mobile terminals with sufficient address resources, making the practical application of mobile IP possible.
(a) When the mobile node is on the local network, the address configuration is performed by receiving router notifications from the router on the local network; and the network parameters are obtained.
(b) When the mobile node accesses the off-site network, it no longer receives router notifications from the local network; instead, it receives router notifications from the router on the off-site network, and the mobile node uses the received off-site network router notifications to perform mobile detection.
(c) After the mobile node has configured the transfer address on the off-site network, it will send a binding update message to the local agent, notify its own transfer address, and register.
(d) The mobile node also sends a binding update message to the communication node and notifies its own transfer address.
(e) In this way, the local agent can forward messages from the communication node to the mobile node through a tunnel.
(f) If the communication node obtains the forwarding address of the mobile node by receiving the binding update from the mobile node, it can communicate directly with the mobile node without the need to use a local proxy to achieve routing optimization.
4. Multicast technology
Multicast is a network technology that allows one or more senders (multicast sources) to send a single data packet to multiple recipients (one time, simultaneous). It is suitable for one-to-multipoint or multi-point to multi-point data transmission services. The basic principle of multicast implementation is: relying on the IP protocol to complete multicast, and IP multicast forces the network to copy information packets at the fork of the data distribution tree. The implementation of IP multicast includes three parts: addressing, multicast member management and multicast routing protocol.
1. Multicast addressing: IPv6 reserves a certain address space for multicast, and its address is "111111111" higher than 8 bits, followed by a 120-bit multicast group identifier. This address is only used as the destination address of the multicast packet, and the multicast source address can only be a unicast address. The sender only needs to send data to the multicast address to transmit user data to multiple different locations without knowing any information from the receiver.
2. Multicast member management: Multicast uses the IGMP (Internet Group Manager Protocol) protocol to implement the dynamic registration process of users. The relationship between the host and the multicast router is established and maintained through the IGMP protocol. The multicast forwarding router uses the IGMP protocol to understand whether there is a receiver of a multicast group, i.e., a group member, on the network segment connected to each interface. If a member appears, the multicast router forwards the multicast packet to this network segment; if not, it stops forwarding or does not forward to save bandwidth.
3. Multicast routing protocol: The function of the multicast routing protocol is to establish and maintain multicast routing tables to make full use of bandwidth. Multicast routing protocols are divided into two types: dense mode and sparse mode. Dense mode multicast routing protocol refers to the dense distribution of multicast members across the entire network, that is, many subnets contain at least one member and have abundant bandwidth, but they are not suitable for large-scale networks that are suitable for sparse distribution of multicast members in the network and may not necessarily have sufficient bandwidth available.
2. Transition technology from IPv4 to IPv6
How to complete the conversion from IPv4 to IPv6 is the first problem that needs to be solved in the development of IPv6. At present, the IETF has set up a special working group to study the conversion problem of IPv4 to IPv6, and has proposed many solutions, mainly including the following types:
1. Network transition technology
(l) Tunneling technology: With the development of IPv6 networks, many local IPv6 networks have emerged. Tunneling technology can be used to connect local IPv6 networks through the existing Internet backbone network (i.e. tunneling) that runs the IPv4 protocol. Therefore, it is the easiest technology to be adopted in the early stages of the transition from IPv4 to IPv6. The tunneling technology method is: the router encapsulates IPv6 data packets into IPv4, and the source and destination addresses of IPv4 packets are the IPv4 addresses at the entrance and exit of the tunnel respectively. At the exit of the tunnel, the IPv6 packet is taken out and forwarded to the destination site.
(2) Network address translation/protocol translation technology: Network address translation/protocol translation technology NAT-PT (Network Address Translation - Protocol Translation) realizes mutual communication between hosts with only IPv6 installed and most applications with only IPv4 installed by combining them with SIIT protocol translation and dynamic address translation (NAT) under traditional IPv4 and appropriate application layer gateway (ALG).
2. Host transition technology
IPv6 and IPv4 are network layer protocols with similar functions. Both are based on the same physical platform, and there is no difference between TCP and UDP on the transport layer protocols loaded on it. It can be seen that if a host supports both IPv6 and IPv4 protocols, then the host can communicate with hosts that support IPv4 protocol and with hosts that support IPv6 protocol. This is the working mechanism of dual protocol stack technology.
3. Application Service System (DNS) Transition Technology
During the transition from IPv4 to IPv6, DNS services as Internet infrastructure must also support the upgrade and conversion of this network protocol. The DNS record formats of IPv4 and IPv6 are different. In order to realize DNS query and response between IPv4 and IPv6 networks, the application-layer gateway DNS-ALG combined with NAT-PT can be used to play a translation role between IPv4 and IPv6 networks. For example, IPv4's address domain name mapping uses "A" records, while IPv6 uses "AAAA" or "A6" records. Then, the DNS query request sent by the IPv4 node to the IPv6 network is a "A" record, and DNS-ALG rewritten "A" into "AAAA" and send it to the DNS server in the IPv6 network. When the server's answer reaches DNS-ALG, DNS-ALG modifys the answer, changes "AAAA" to "A", changes the IPv6 address to the IPv4 translation address in the DNS-ALG address pool, notifies NAT-PT of the mapping relationship between the IPv4 translation address and the IPv6 address, and returns the IPv4 translation address as the resolution result to the IPv4 host. The IPv4 host uses this IPv4 translation address as the destination address to communicate with the actual IPv6 host through NAT-PT.
The above technologies rely heavily on the conversion from the Internet that supports IPv4 to Internet that supports IPv6. We expect IPv4 and IPv6 to be compatible with each other during this conversion process. At present, the 6to4 mechanism is one of the more popular means of implementation.