IPv4 and IPv6: A Comparison

History

The groundwork for today’s Internet was the result of the Department of Defense’s Defense Advanced Research Projects Agency Network, or DARPANET. DARPANET was born from the desire to build a network that would continue to function even if a segment was lost. In the late 1960s there was a growing interest by universities to build a network connecting their current research infrastructures. An increasing number of universities were now able to afford and use computers as part of their education and research. Computers were becoming a vital part of research. Research institutes wanted to be able to share their efforts between different parts of the universities. There was also a mounting importance for universities to work in conjunction with other universities by sharing electronic data. ARPANET was first used in 1971 and made use of IP’s predecessor, the Network Control Protocol or NCP. 

IPv4 became more finalized as the official RFC’s for TCP\IPv4 were documented. RFC’s 760 and 761 were written in 1980 as outlines for the new protocols, which were TCP and IP. RFC’s 791, 792 and 793 were published in September 1981 by DARPA. The initial TCP\IP RFC’s were instrumental in IPv4’s further implementation worldwide. A documented standard allowed other research organizations and universities to begin testing and standardizing on one set of protocols. 

Now that standards were available, testing and implementation on a wide scale was necessary. In 1980, ARPA started converting systems to TCP/IP. In 1983, ARPANET was split into two different networks: MILNET for military research and ARPANET for continuing the current path of research. In 1983 it was also mandated by ARPA that all systems connected to ARPANET use TCP/IP. 

Need for Replacement

By the late 1980s more and more organizations, primarily research oriented, had connected to the Internet. With this growth came more hosts that required IP addresses, lessening the pool of available addresses. By 1989 there were approximately 100,000 hosts connected to the Internet. That figure jumped to approximately 1,000,000 by 1992. Numbers increased dramatically in 1993 with the release of GUI browsers for HTML. Browsers like Netscape helped make the World Wide Web a commercial tool. With the World Wide Web and email becoming more focused on the consumer, Internet connectivity moved from being focused at Universities and research institutes to businesses and homes. 

Uses of new technologies have extended IPv4’s lifetime. Network Address Translation (NAT) and Classless Inter-Domain Routing (CIDR) are the two key elements that have allowed for the extension of IP addresses in IPv4. NAT allows multiple systems on a private network to share one public IP address. NAT can also help prevent unwanted access from outside the private network. 

CIDR is a type of addressing within IPv4 that makes use of 13 to 27 prefixes. The prefix designates how many of the initial address bits are used to identify the private network. The remaining bits identify the local host. 

CIDR and NAT were crucial in making the IPv4 address pool last as long as it has. The expansive need for CIDR, NAT and BGP has resulted in overly large routing tables. The extensive routing tables result in unnecessary administrative overhead in building, maintaining and replacing routing equipment. As these technologies are no longer needed in IPv6 routing, table size will decrease dramatically. 

Depleted Address Pool
IPv4 and the Internet were started and planned for research and communication between the US Government research agencies and Universities. The planners did not anticipate the worldwide interest and commercial growth of the Internet. The US focus of the Internet resulted in a low number of IP addresses assigned and available to Europe and Asia. 

Some large US companies, such as Xerox, Ford and Eli Lily, are assigned Class A IPv4 addresses, while major countries in Europe and Asia are stretching the limits of IPv4 with multiple Class C addresses. Asia currently has slightly more IPv4 addresses than US based Level 3 Communications, whose three Class A domains are equivalent to 48 million IPv4 addresses. Rob Batchelder, Research Director Internet Infrastructure at the industry think tank Gartner Group, puts this into perspective "Asia, In particular, is encountering some real problems with address space depletion. IPv4 address space is largely titled towards the US because we’re the ones who ‘invented’ the Internet."

It is easy to wonder why companies who do not return or reallocate the IPv4 Class A or B addresses they hold would not return them. The process of returning an address is extremely difficult and has only been done once, by Stanford University. Additionally, the return of IP addresses could result in a market that sees the remaining pool being squandered as a profit is made in selling those addresses and creating a false demand for them.

Disruptive Technologies
Adding to the shortfall of IPv4 addresses are technologies that some refer to as "Disruptive Technologies." The term "Disruptive Technologies" was coined by Pete Loshin in his book titled, "IPv6 Clearly Explained." The idea is that IPv6 will be implemented not because of system upgrades and the replacement of legacy equipment but rather due to the need and use of IPv6 in areas that could not be foreseen by those who laid the Internet groundwork. 

The emergence of online gaming is hastening the implementation of IPv6. Major game system manufacturers are producing or making plans to sell products that allow consumers to play games across the Internet. These gaming systems have been growing in popularity and their ability to network makes them that much more popular. The increased demand for IP addresses follows a path, similar to the increasing number of network capable gaming systems in homes.

The exponential growth of home networking has created a greater demand for IPv6. Home networking is not limited to connecting a few computers in a home; it expands to all devices in your home that require IP addresses. Cellular phones, televisions, video equipment and refrigerators are among the IP capable devices that are on the market now and the growth of this market shows no signs of slowing. Market estimates state that over 1 billion cell phones will be sold by 2005 and approximately 15% of cars will have some type of IP based device by 2010.

Mobile IP may presently be the hottest area of development that is pushing the need for IPv6, especially in Europe and Asia. The majority of key manufacturers are shipping PDA’s, cellular phones and laptops with some type of wireless Internet capability. Paralleling this is the emergence of free wireless access in cafés, restaurants, hotels, airports and various businesses. The combination of these two key factors helps push the demand for mobile devices and services and in turn the demand for IP addresses.

Comparison of IPv4 and IPv6

While IPv4 and IPv6 are similar in much of their basic framework, there are also many differences. From first glance, there are obviously differences in the addresses between IPv4 and IPv6. The graphic below shows an IP address for both versions of IP.

IPv4 Address Example: 
125.12.3.65

IPv6 Address Example: 
2145:00D5:2F3B:0000:0000:00FF:EF00:98F3

Removing zeros can also reduce the IPv6 address. Zeros can be removed when they are leading in and within any 16 bit block. The address from the previous example could be reduced using this to the following representation. Note that in the example the block of EF00 does not lose its zeros because they are at the end of the block.

IPv6 Address with Leading Zeros Removed:
2145:D5:2F3B:0:0:FF:EF00:98F3

Compressing zeros can further reduce IPv6 addresses. A contiguous block of zeros within a 16 bit block can be removed. The blocks of zeros are then represented by double colons ( :: ). 

For example, the IPv6 Multicast address of FF02:0000:0000:0000:0000:0000:0000:0002 can be reduced to FF02::2 using compression.

IPv6 Address with Compressed and Removed Zeros:
2145:D5:2F3B::FF:EF00:98F3

An IPv4 address has 32 bits, whereas an IPv6 address contains 128 bits. The 128 bits in an IPv6 address are split between the network and host addresses. There are 64 bits for the network address and 64 bits for the host address. Due to the larger address space, the number of available addresses jumps from 4,294,967,296 in IPv4 to 340,282,366,920,938,463,463,374,607,431,768,211,456 (or 3.4X10^38) in IPv6. IPv6’s address is also separated using a different format. IPv4 uses a dotted decimal and IPv6 uses a colon-hex format. The larger address space allows for clearer addressing and routing. It also allows for multiple interfaces per host and multiple addresses per interface. 

The IPv6 address space supports three types of address; Unicast, Multicast and Anycast. IPv6 Multicast addressing absorbs the role of IPv4’s broadcast addresses, which is no longer present. The biggest change is the introduction of the Anycast address. Anycast addressing allows multiple nodes to be assigned the same Anycast Address. When packets are sent to this address routing decides which node is closest to the source and routes the traffic to it. Anycast addresses could be useful in setting up mirror websites, with different physical locations being accessible through the same Anycast address. A user trying to access this site would then be routed to the closest site, resulting in a better experience.

Addressing enhancements result in reduced administrative overhead. The teaming of IPv6 Neighbor Discovery and address autoconfiguration allows hosts to operate in any location without any special support. Renumbering is made easier, resulting in less manual attention by support and network administrators. Renumbering also makes transition from ISP to ISP or network segment to segment much easier and potentially seamless. Stateless and Stateful address configuration assist in making IP configuration and planning easier. Stateless configuration works without a DHCP server, while Stateful is a configuration that has a DHCP server present. 

Address Autoconfiguration allows for a node to make use of router discovery to determine router addresses, network configuration parameters, on-link prefixes and additional addresses. What makes Address Autoconfiguration so impressive is that while it requires a multicast capable interface, it is possible without the use of DHCP. Through proper configuration and planning, this can reduce the overhead caused by DHCP management in large organizations and ISP’s.

With a new addressing scheme comes a new way of handling name resolution through DNS. The DNS changes required to support IPv6 are specified in RFC 1886. As part of the interim transition from IPv4 to IPv6, it is possible to register an IPv6 address on a DNS server as an IPv4 address. This is important if a consumer’s ISP has not moved to IPv6 for DNS and the consumer would prefer to use IPv6 DNS. The figure below shows a WHOIS lookup in which the domain has an IPv6 address and is found through IPv4 DNS.

This example shows a WHOIS registration record from the registrar Network Solutions. The initial resolution with Network Solutions is an IPv4 address, the DNS server from which the record was retrieved. 

Security is a key feature of IPv6. IPv6 is primarily focused on improved security, which makes it popular as data security becomes more and more of a hot topic in all areas of IT. There are many standardized and required security features within IPv6 without having to make changes to applications. Among the improved security features is packet signing to handle authentication. Data confidentiality through encryption helps aid security within IPv6. IPv6 includes an end to end security model that is designed to protect DHCP, DNS and IPv6 mobility. While IPv6’s improvement in security does not make IP invulnerable from attacks, it is certainly a positive and necessary addition. 

The IP routing experience differs with the implementation of IPv6. Smaller routing tables result in more efficient routing and less overhead through faster computation and aggregation. The routing structure makes use of a hierarchical structure that is also more efficient. 

IPv6 brings major changes to the IP header. IPv6’s header is far more flexible and contains fewer fields, with the number of fields dropping from 13 to 8. Fewer header fields result in a cleaner header format and Quality of Service (QoS) that was not present in IPv4. IP option fields in headers have been replaced by a set of optional extensions. The efficiency of IPv6’s header can be seen by comparing the address to header size. Even though the IPv6 address is four times as large as the IPv4 address, the header is only twice as large. Priority traffic, such as real time audio or video, can be distinguished from lower priority traffic through a priority field. The images below show the difference in the headers. Red designates fields in the IPv4 header that are no longer present in the IPv6 header.