Comparison with IPv4 IPv6

1 comparison ipv4

1.1 larger address space
1.2 multicasting
1.3 stateless address autoconfiguration (slaac)
1.4 network-layer security
1.5 simplified processing routers
1.6 mobility
1.7 options extensibility
1.8 jumbograms
1.9 privacy

comparison ipv4

on internet, data transmitted in form of network packets. ipv6 specifies new packet format, designed minimize packet header processing routers. because headers of ipv4 packets , ipv6 packets different, 2 protocols not interoperable. however, in respects, ipv6 extension of ipv4. transport , application-layer protocols need little or no change operate on ipv6; exceptions application protocols embed internet-layer addresses, such file transfer protocol (ftp) , network time protocol (ntp), new address format may cause conflicts existing protocol syntax.

larger address space

the main advantage of ipv6 on ipv4 larger address space. length of ipv6 address 128 bits, compared 32 bits in ipv4. address space therefore has 2 or approximately 7038340000000000000♠3.4×10 addresses.

in addition, ipv4 address space poorly allocated; in 2011, approximately 14% of available addresses utilized. while these numbers large, not intent of designers of ipv6 address space assure geographical saturation usable addresses. rather, longer addresses simplify allocation of addresses, enable efficient route aggregation, , allow implementation of special addressing features. in ipv4, complex classless inter-domain routing (cidr) methods developed make best use of small address space. standard size of subnet in ipv6 2 addresses, square of size of entire ipv4 address space. thus, actual address space utilization rates small in ipv6, network management , routing efficiency improved large subnet space , hierarchical route aggregation.

renumbering existing network new connectivity provider different routing prefixes major effort ipv4. ipv6, however, changing prefix announced few routers can in principle renumber entire network, since host identifiers (the least-significant 64 bits of address) can independently self-configured host.

multicasting

multicasting, transmission of packet multiple destinations in single send operation, part of base specification in ipv6. in ipv4 optional although commonly implemented feature. ipv6 multicast addressing shares common features , protocols ipv4 multicast, provides changes , improvements eliminating need protocols. ipv6 not implement traditional ip broadcast, i.e. transmission of packet hosts on attached link using special broadcast address, , therefore not define broadcast addresses. in ipv6, same result can achieved sending packet link-local nodes multicast group @ address ff02::1, analogous ipv4 multicasting address 224.0.0.1. ipv6 provides new multicast implementations, including embedding rendezvous point addresses in ipv6 multicast group address, simplifies deployment of inter-domain solutions.

in ipv4 difficult organization 1 globally routable multicast group assignment, , implementation of inter-domain solutions arcane. unicast address assignments local internet registry ipv6 have @ least 64-bit routing prefix, yielding smallest subnet size available in ipv6 (also 64 bits). such assignment possible embed unicast address prefix ipv6 multicast address format, while still providing 32-bit block, least significant bits of address, or approximately 4.2 billion multicast group identifiers. each user of ipv6 subnet automatically has available set of globally routable source-specific multicast groups multicast applications.

stateless address autoconfiguration (slaac)

ipv6 hosts can configure automatically when connected ipv6 network using neighbor discovery protocol via internet control message protocol version 6 (icmpv6) router discovery messages. when first connected network, host sends link-local router solicitation multicast request configuration parameters; routers respond such request router advertisement packet contains internet layer configuration parameters.

if ipv6 stateless address auto-configuration unsuitable application, network may use stateful configuration dynamic host configuration protocol version 6 (dhcpv6) or hosts may configured manually using static methods.

routers present special case of requirements address configuration, sources of autoconfiguration information, such router , prefix advertisements. stateless configuration of routers can achieved special router renumbering protocol.

network-layer security

internet protocol security (ipsec) developed ipv6, found widespread deployment first in ipv4, re-engineered. ipsec mandatory specification of base ipv6 protocol suite, has since been made optional.

simplified processing routers

in ipv6, packet header , process of packet forwarding have been simplified. although ipv6 packet headers @ least twice size of ipv4 packet headers, packet processing routers more efficient, because less processing required in routers due headers being aligned match common word sizes. moreover, ipv6 doesn t implement header checksum, in contrast ipv4. furthers end-to-end principle of internet design, envisioned processing in network occurs in leaf nodes.

the packet header in ipv6 simpler ipv4 header. many used fields have been moved optional header extensions.

ipv6 routers not perform ip fragmentation. ipv6 hosts required either perform path mtu discovery, perform end-to-end fragmentation, or send packets no larger default maximum transmission unit (mtu), 1280 octets.

the ipv6 header not protected checksum. integrity protection assumed assured both link layer or error detection , correction methods in higher-layer protocols, such tcp , udp. in ipv4, udp may have checksum of 0, indicating no checksum; ipv6 requires checksum in udp. therefore, ipv6 routers not need recompute checksum when header fields change, such time live (ttl) or hop count.

the ttl field of ipv4 has been renamed hop limit in ipv6, reflecting fact routers no longer expected compute time packet has spent in queue.

mobility

unlike mobile ipv4, mobile ipv6 avoids triangular routing , therefore efficient native ipv6. ipv6 routers may allow entire subnets move new router connection point without renumbering.

options extensibility

the ipv6 packet header has minimum size of 40 octets. options implemented extensions. provides opportunity extend protocol in future without affecting core packet structure. however, study in 2015 indicated there still widespread dropping of ipv6 packets containing extension headers.

jumbograms

ipv4 limits packets 65,535 (2−1) octets of payload. ipv6 node can optionally handle packets on limit, referred jumbograms, can large 4,294,967,295 (2−1) octets. use of jumbograms may improve performance on high-mtu links. use of jumbograms indicated jumbo payload option header.

privacy

like ipv4, ipv6 supports globally unique ip addresses network activity of each device can potentially tracked. design of ipv6 intended re-emphasize end-to-end principle of network design conceived during establishment of internet. in approach each device on network has unique address globally reachable directly other location on internet.

network prefix tracking less of concern if user s isp assigns dynamic network prefix via dhcp. privacy extensions little protect user tracking if isp assigns static network prefix. in scenario, network prefix unique identifier tracking , interface identifier secondary.

in ipv4 effort conserve address space network address translation (nat) obfuscates network address spaces, hosts, , topologies. in ipv6 when using address auto-configuration, interface identifier (mac address) of interface port used make public ip address unique, exposing type of hardware used , providing unique handle user s online activity.

it not requirement ipv6 hosts use address auto-configuration, however. yet, when address not based on mac address, interface s address globally unique, in contrast nat-masqueraded private networks. privacy extensions ipv6 have been defined address these privacy concerns, although silvia hagen describes these being largely due misunderstanding . when privacy extensions enabled, operating system generates random host identifiers combine assigned network prefix. these ephemeral addresses used communicate remote hosts making more difficult track single device.

privacy extensions enabled default in windows (since xp sp1), os x (since 10.7), , ios (since version 4.3). linux distributions have enabled privacy extensions well.

in addition temporary address assignments, interfaces receive stable address. interface identifiers generated such stable each subnet, change host moves 1 network another. in way difficult track host moves network network, within particular network have same address (unless state used in generating address reset , algorithm run again) network access controls , auditing can potentially configured.

the traditional method of generating interface identifiers in use unique address assignments based on mac addressing. in favor of better privacy protection, method has been deprecated in operating systems newly established methods of rfc 7217.

privacy extensions not protect user other forms of tracking @ other layers, e.g. application layer: tracking cookies or browser fingerprinting , link layer: imsi-catcher or ibeacon