Analysis and Comparison of IPv6 and NAT

Analysis and Comparison of IPv6 and NAT 1 Introduction In recent years, the Internet based on IPv4 protocol has been booming. Both the network coverage and network capacity, the number of Internet users and the business type are growing rapidly. Computer Networks, telecom networks, and cable television networks are moving towards an integrated IP-based next-generation network (NGN ). While the IP technology is favored and pursued by network builders in history, present, and future, we must be aware of the many imperfections in the IPv4 technology. Quality of Service (QoS), security, and address shortage are some of the key problems facing IPv4 networks, these potentially "Fatal" shortcomings are hindering the further Popularization and Application of IP technology and services. This article discusses solutions to address shortage. There are two types at present: www.2cto.com
(1) "Open Source", that is, IPv6, the next generation IP protocol with a larger address length, is a medium-and long-term solution; (2) "throttling ", that is, try to improve the utilization of existing IPv4 address resources. For example, the use of private multiplexing and various network address (port) Translation technologies are some short-and mid-term solutions. 2. Increase address utilization. Here we will first introduce the "throttling" technology. Currently, the following methods have been proposed or used to improve IPv4 address utilization: * address recycling: IP Standardization Organization Internet Engineering Task Group (IETF) has issued a call, it is hoped that the organizations and organizations allocated "excessive" addresses in their early years will be able to return the address resources they are unlikely to use, but it seems that the results have little effect at present.
* More stringent Address Allocation Policy: The Address Allocation Policy of an IANA organization previously assigned an IP address based on the "expected maximum scale, to reduce the IP Address allocation rate, the policy has been changed to "current scale required ". * Classless Inter-Domain Routing (CIDR): As class B addresses are nearly exhausted, CIDR technology is used to break the division of address types (such as class A, Class B, and class C, multiple consecutive Class C addresses are used for aggregation and allocation. The adoption of CIDR technology has slowed the growth of Internet route tables in recent years. * Dial-up access: when a user performs dial-up access, the user uses the Point-to-Point Protocol (PPP) and Dynamic Host Configuration Protocol (DHCP) to dynamically allocate IP addresses as needed to achieve address reuse.
* Variable-length Subnet Mask (VLSM): an enterprise's Enterprise Network (Intranet) is divided into subnets of different sizes to improve address space utilization. * Private address and Network Address Translation: "private" IP addresses are used in the enterprise network. Addresses of different enterprise networks can be reused. When a user in a private address network communicates with the Internet, network address translation must be performed between the two, because the address used by the enterprise network and the Internet address are not in a unified address space. The private address range specified by IETF is 10/8, 172.16/12,192.168/16 (RFC 1918 ). This is the most widely used and the most efficient solution to address shortage. Of course, enterprises only choose to use the private address defined in RFC 1918, sometimes not only because of address reuse, but also other considerations such as security and control.
3 NAT problems www.2cto.com when a CEN uses a private IPv4 address, it is not in the same address space as the Internet address (Public IPv4. Therefore, when devices in these networks communicate with the Internet, address translation must be performed at the network boundary. Typical translation technologies include Network Address Translation (NAT) and network address port translation (NAPT), which are collectively referred to as NAT. [RFC 2993]: * When a CEN communicates with other networks, NAT affects network performance and reduces network throughput. * The original Internet end-to-end design concept (end-to-end transparency) is broken. It is applicable only to Client/Server models and not to Peer-to-Peer (P2P) application models. For example, in this model, the promotion of the PUSH Service is very limited. * A single point of failure may occur on devices that maintain the connection status and dynamic ing information. * The Multi-home technology that increases the connection reliability between enterprise network sites and the Internet is more complex. * Security services are blocked at the network layer.
* Address conflict (overlap) may occur when the enterprise network needs to be integrated with other private networks. * NAPT and Realm Specific IP (rsip) increase operation complexity. * The authentication mechanism of Simple Network Management Protocol (SNMPv3) in version 3 is more complex and even ineffective. * Many high-level application protocols are incompatible with [RFC 3027]. For example, some applications (such as IPSec, Kerberos/5) cannot traverse NAT, while some (such as FTP, H.323, SIP, SMTP, RSVP, DNS, SNMP, etc) the Application Layer Gateway (ALG) is required. With the help of ALG technique, each new application must consider adding support for this application in NAT, which violates the idea that IP technology is independent from high-level applications.
4. IPv6 disputes the open source technology is introduced here. Although the design of IPv4 was based on the technical level in the middle of 1970s and the very limited operation experience at that time, the great success of IPv4-based Internet has proved that the design of IPv4 is basically very successful. When designing IPv6 in 1990s, there was a good reason to stick to the IPv4 feature to the maximum extent in IPv6, simply increasing the address length. However, based on Internet operation experience over the years, we should make some "revolutionary" changes to other parts of IPv4. Compared with the IPv4 protocol, the biggest change of the IPv6 protocol is that it significantly simplifies the header design, which is mainly reflected in: * simplified Header Format, all headers use a fixed length; * reduce sensitive parts of package processing, such as checksum and sharding. * increase the address length to bits. Although IETF has already chosen IPv6 as the next-generation Internet (NGN) protocol, there are still some controversies in the industry over IPv6 technology.
(1) Quality of Service (QoS) assurance. The IPv6 Header does have some improvement in performance. For example, the checksum field is removed, but this is not the same as the so-called QoS warranty. Currently, the technologies that address IP QoS are DiffServ, InterServ, and Multi-Protocol Label Switching (MPLS), which are applicable to both IPv4 and IPv6. In other words, IPv6 uses the same technology as IPv4 to solve the QoS problem. Without IPv6, the service quality will be guaranteed. (2) security assurance. Both IPv4 and IPv6 use the IPSec protocol to provide security assurance. The difference is that IPv4 has optional requirements for IPSec and IPv6 has mandatory requirements for IPSec. However, IPv6 only requires implementation and does not require all applications to use IPSec, because first, it is unnecessary, and second, it will have a significant impact on performance. Therefore, from this point of view, the security of the two is almost equivalent, and there is no possibility that IPv6 will be safer than IPv4. (3) better support for mobility. We should not deny that IPv6 has better support for mobility (Mobbie IP) than IPv4, but note that only when the Mobile IP business has the actual market value, is it possible to explore how to better support it? Mobile IP has not been widely used since its birth. The root cause is the lack of "Killer Application", which means there is no E-mail. WWW is an application that greatly promotes Internet development. It must or is best to use Mobile IP to support it. This may happen in the future (for example, 3G popularity), but it is not, at least not obvious or widely used.
(4) according to a report published by Internet Society, there are still some misunderstandings or sayings about IPv6: * IPv6 must be used for mobility. * IPv6 is more suitable for wireless networks. * Only IPv6 supports automatic configuration. * IPv6 solves the routing scalability problem. * IPv6 better supports fast prefix numbering. * IPv6 provides better support for Multi-home. Conclusion 5 some people think that IPv4 addresses will soon run out, so IPv6 will be commercially available in recent years, which is questionable. According to IANA, about 36% of Public IPv4 addresses (about 1.5 billion) are not allocated, therefore, the address shortage problem is not as serious and urgent as some media or manufacturers have described. Although IANA's policy for applying for IPv4 addresses tends to be strict, if the carrier does have the requirement, it can still be applied. The current IP Address Allocation speed remains the same. It is conservatively estimated that IPv4 addresses will only be used up after January 1, 2015. Second, www.2cto.com generally thinks that mobile data services require a large amount of IP address space resources and must use IPv6 addresses. However, the actual situation is that many mobile data operators do not adopt completely open Internet business models, instead, they prefer the commercial model of "Walled Garden", namely, limiting users to a closed network, because in this way, operators can easily control their businesses and users, making billing convenient, and higher security. When the "Walled Garden" business model is adopted, the mobile data network and the Internet form an isolated addressing domain. Therefore, operators almost choose to use more mature, inexpensive, and more familiar (private) IPv4 address.
Third, the-bit IPv6 address must be in the middle so that "each grain of sand on the earth is allocated to an IP address". That is to say, IPv6 address space is no longer a rare resource. Since it is not a rare resource, it is unlikely that it will become a rare resource in the future. How much practical significance does it make even more IPv6 address space? Fourth, both NAT technology and IPv6 can solve the address shortage problem. I think that the core issue of NAT is that it will undermine the end-to-end transparency of the Internet, so it is only applicable to applications in the customer/Server mode. The biggest advantage of IPv6 is that it can support Peer to Peer (P2P) applications, because it will not damage the Internet end-to-end transparency. Therefore, IPv6 must be deployed only when P2P applications (such as games) are widely used.