The development of computer network technology today is increasingly rapid along with the needs of the community for services that utilize computer networks. In a computer network system, the protocol is the most important part. The commonly used network protocol is IPv4, which still has several shortcomings in handling the number of computers in an increasingly complex network. A new network protocol has been developed, namely IPv6, which is a solution to the above problems. This new protocol has not been widely implemented in networks in the world.
IP version 6 (IPv6) is a new version of the Internet protocol designed as a replacement for Internet protocol version 4 (IPv4) defined in RFC 791. IPv6, which has a giant address capacity (128 bits), supports structured address arrangement, which allows the Internet to continue to grow and provides new routing capabilities not found in IPv4. IPv6 has an anycast address type that can be used for efficient route selection. In addition, IPv6 is also equipped with a local address usage mechanism that allows for Plug & Play installations, and provides a platform for new ways of using the Internet, such as support for real-time data flow, provider selection, host mobility, end-to-end security, or automatic configuration.
IPv4, the foundation of the Internet, is nearing the end of its capabilities, and IPv6, a new protocol, has been designed to replace IPv4. The main motivation for replacing IPv4 is because of the limitations of its address length of only 32 bits and its inability to support the need for secure communication, flexible routing, or data traffic management.
IP version 6 (IPv6) is a new version of the Internet protocol designed as a replacement for Internet protocol version 4 (IPv4) defined in RFC 791. IPv6, which has a giant address capacity (128 bits), supports structured address arrangement, which allows the Internet to continue to grow and provides new routing capabilities not found in IPv4. IPv6 has an anycast address type that can be used for efficient route selection. In addition, IPv6 is also equipped with a local address usage mechanism that allows for Plug & Play installations, and provides a platform for new ways of using the Internet, such as support for real-time data flow, provider selection, host mobility, end-to-end security, or automatic configuration.
IPv6 Advantages
Automation of various settings / Stateless-less auto-configuration (plug & play) Addresses on IPv4 are basically static to the host. Usually given sequentially to the host. Indeed, currently the above can be done automatically using DHCP (Dynamic Host Configuration Protocol), but this in IPv4 is only an additional function, in contrast to IPv6 the function to set automatically is provided as standard and is the default. In this automatic setting there are 2 ways depending on the use of the address, namely stateless and statefull automatic settings.
1. Statefull Automatic Setting
Strict management method in terms of IP address range given to the host by providing a server for managing IP address conditions, where this method is almost similar to the DHCP method in IPv4. When setting automatically, the information needed between the router, server and host is ICMP (Internet Control Message Protocol) which has been expanded. In ICMP in IPv6, it also includes IGMP (Internet Group Management Protocol) which is used in multicast in IPv4.
Figure 2. Statefull Automatic Settings
2. Stateless Automatic Settings
In this way, there is no need to provide a server for managing and distributing IP addresses, just setting up a router where the host that has been connected to the network from the router on the network gets the prefix from the address of the network. Then the host adds a bit pattern obtained from unique information to the host, then creates a 128-bit IP address and makes it the IP address of the host. In this unique information for the host, among others, the MAC address of the network interface is used. In this stateless automatic setting, behind the ease of management, on Ethernet or FDDI because it is necessary to provide at least 48 bits (as large as the MAC address) to one network, it has the disadvantage of poor address usage efficiency.
Figure 3. Stateless Automatic Settings
Changes From IPV4 to IPV6
The change from IPv4 to IPv6 basically occurred due to several things which are grouped into the following categories:
1. Address Expansion Capacity
IPv6 increases the size and number of addresses that can be supported by IPv4 from 32 bits to 128 bits. This increase in address capacity is used to support the increase in hierarchy or address groups, increase the number or capacity of addresses that can be allocated and given to nodes and simplify the configuration of addresses on nodes so that they can be done automatically. Scalability improvements are also made to multicast routing by increasing the scope and number of multicast addresses. In addition to increasing the number of address capacities that can be allocated to nodes, IPv6 also introduces a new type or type of address, namely the anycast address. This type of anycast address is defined and used to send packets to one of a group of nodes.
2. Simplification of Header Format
Some fields in the IPv4 header have been removed or can be made optional. This is used to reduce the processing overhead of common IPv6 packet handling and limit bandwidth costs in the IPv6 header.
Thus, header processing on IPv6 packets can be done efficiently.
3. Options and Extension Headers
The changes that occur in IP headers are that the coding of Options headers in IP is included to make packet forwarding more efficient, so that the length restrictions on the options headers in IPv6 packets are not too strict and it is very flexible/possible to introduce new options headers in the future.
4. Packet Flow Labeling Capabilities
New capabilities or features added to IPv6 are to allow packet labeling or classifying packets that request special handling, such as certain quality of service (QoS) or real-time.
5. Authentication and Privacy Capabilities
Additional capabilities to support authentication, data integrity and critical data are also specified in IPv6 addresses. The biggest change in IPv6 is the expansion of the IP address from 32 bits in IPv4 to 128 bits. This 128 bit is a continuous address space by eliminating the concept of class. In addition, changes were also made to the way IP addresses are written. If in IPv4 32 bits are divided into 8 bits each separated by "." and written with decimal numbers, then in IPv6, the 128 bits are separated into 16 bits each separated by ":" and written in hexadecimal. In addition, a tiered structure is also introduced to make routing management easier. In CIDR (Classless Interdomain Routing) the routing table is reduced by combining routing information from an organization into one.
Table 1. Address Space Division in IPv6
| Allocation | Prefix (binary) | Fraction of Address Space |
|-----------------------------------------------------------|------------------|----------------------------|
| Reserved | 0000 0000 | 1/256 |
| Unassigned | 0000 0001 | 1/256 |
| Reserved for NSAP Allocation | 0000 001 | 1/128 |
| Reserved for IPX Allocation | 0000 010 | 1/128 |
| Unassigned | 0000 011 | 1/128 |
| Unassigned | 0000 1 | 1/32 |
| Unassigned | 0001 | 1/16 |
| Unassigned | 001 | 1/8 |
| Provider based Unicast Address | 010 | 1/8 |
| Unassigned | 011 | 1/8 |
| Reserved for NeutralInterconnect-Based Unicast Addresses | 100 | 1/8 |
| Unassigned | 101 | 1/8 |
| Unassigned | 110 | 1/8 |
| Unassigned | 1110 | 1/16 |
| Unassigned | 1111 0 | 1/32 |
| Unassigned | 1111 10 | 1/64 |
| Unassigned | 1111 1101 | 1/128 |
| Unassigned | 1111 1110 | 1/512 |
| Link Local Use Addresses | 1111 1110 10 | 1/1024 |
| Site Local Use Addresses | 1111 1110 11 | 1/1024 |
| Multicast Addresses | 1111 | 1/256 |
| | 1111 | |To understand the hierarchical structure of IPv6 addresses, see the example of the provider address. First, the 128-bit address is divided into several fields that can change in length. If the first 3 bits of the address are "010", then this is the space for the provider. While the next n bits are the registry ID, which is a field that indicates the place/institution that provides the IP address. For example, the IP address provided by InterNIC, the field becomes "11000". Then the next m bits are the provider ID, while the next o bits are the Subscriber ID to distinguish organizations registered with the provider.
Then the next p bit is the Subnet ID, which marks a collection of hosts that are topologically connected in the network of the organization. And the last q=125(n+m+o+p) bit is the Interface ID, which is the IP address that marks the hosts contained in the groups that have been marked by the Subnet ID.
The Subnet ID and Interface ID are freely given by the organization. The organization is free to use the remaining p+q bits of the IP address in providing IP addresses within its organization after getting the first 128-(p+q) bits of the IP address. At that time, the administrator of the organization can divide it into sub-networks and hosts in the appropriate bit length, if necessary it can also be made more structured. Because the bit length of the provider ID and subscriber ID can change, the address given to the provider and the number of IP addresses that can be given by the provider to users can be given freely according to needs. In IPv6, the routing control section of the address field is called a prefix, which can be considered equivalent to the network address in IPv4.
Getting to Know IPV6 Classes
There are several important IPv6 classes, namely:
- Aggregatable Global Unicast Addresses: includes IPv6 addresses with the initial bit 001.
- Link-Local Unicast Addresses: includes IPv6 addresses with initial bits 1111 1110 10.
- Site-Local Unicast Addresses: includes IPv6 addresses with initial bits 1111 1110 11.
Multicast Addresses: includes IPv6 addresses with initial bits 1111 1111.
In the IPv4 protocol, special addresses such as 127.0.0.1 are known which refer to localhost, this address is represented as 0:0:0:0:0:0:0:1 or ::1 in the IPv6 protocol. In addition, in IPv6, another special address is known 0:0:0:0:0:0:0:0 as an unspecified address which should not be given as an identifier on an interface. In general, the unicast address format is as follows:
Figure 11. Unicast Address Format
Interface ID is used as a unique identifier for each host in a subnet. In its use, the interface ID is generally 64 bits with IEEE EUI-64 format. If an Ethernet media with a 48-bit MAC address is used, the formation of the interface ID in IEEE EUI-64 format is as follows:
Suppose the MAC address is 00:40:F4:C0:97:57
- Add 2 bytes, namely 0xFFFE in the middle of the address so that it becomes 00:40:F4:FF:FE:C0:97:57
- Complement (change bit 1 to 0 and vice versa) the second bit from the back in the initial byte of the address formed, so that what is complemented is '00' (in hexadecimal) or '00000000' (in binary) becomes '00000010' or '02' in hexadecimal.
- The interface ID obtained in IEEE EUI-64 format is 0240:F4FF:FEC0:9757.
Figure 1. Network - tunneling (IPv6 transition)
Table 3. Comparison of IPv4 and IPv6
| Ipv4 | Ipv6 |
|-----------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------|
| Panjang alamat 32 bit (4 bytes) | Panjang alamat 128 bit (16 bytes) |
| Dikonfigurasi secara manual atau DHCP IPv4 | Tidak harus dikonfigurasi secara manual, bisa menggunakan address autoconfiguration. |
| Dukungan terhadap IPSec opsional | Dukungan terhadap IPSec dibutuhkan |
| Fragmentasi dilakukan oleh pengirim dan pada router, menurunkan kinerja router. | Fragmentasi dilakukan hanya oleh pengirim |
| Tidak mensyaratkan ukuran paket pada link-layer dan harus bisa menyusun kembali paket berukuran 576 byte. | Paket link-layer harus mendukung ukuran paket 1280 byte dan harus bisa menyusun kembali paket berukuran 1500 byte |
| Checksum termasuk pada header. | Cheksum tidak masuk dalam header |
| Header mengandung option. | Data opsional dimasukkan seluruhnya ke dalam extensions header. |
| Menggunakan ARP Request secara broadcast untuk menterjemahkan alamat IPv4 ke alamat link-layer. | ARP Request telah digantikan oleh Neighbor Solitcitation secara multicast. |
| Untuk mengelola keanggotaan grup pada subnet lokal digunakan Internet Group Management Protocol (IGMP). | IGMP telah digantikan fungsinya oleh Multicast Listener Discovery (MLD). |Getting to Know IPV6 Addresses
1. Unicast Address (one-to-one)
Used for one-on-one communication, by designating one host. This Unicast address consists of:
- Global, addresses are used for example for provider addresses or geographic addresses.
- Link Local Address is an address used in one link only. What is meant by link here is a local network that is connected to each other at one level. This address is created automatically by a host that has not received a global address, consisting of a 10+n bit prefix starting with "FE80" and a 118-n bit field indicating the host number. Link Local Address is used in automatic IP address assignment.
- Site-local, an address that is equivalent to a private address, which is used only within the site. This address can be given freely, as long as it is unique within the site, but cannot send packets to this address outside the site.
- Compatible.
Figure 4. Unicast Address Structure
Figure 5. Sending Packets to Unicast Address
2. Multicast (One-to-Many)
Which is used for 1 to many communication by designating a host from the group. This Multicast Address in IPv4 is defined as class D, while in IPv6 the first 8 bits of space starting with "FF" are provided for the multicast Address. This space is then divided again to determine the range of validity. Then the Blockcast address in IPv4 whose host part address is defined as "1", in IPv6 is included in this multicast Address. Blockcast addresses for communication in the same segment separated by a gateway, just as multicast addresses are sorted based on the destination range.
Figure 6. Multicast Address Structure
Figure 7. Sending Packets on Multicast Address
3. Anycast Address
Which designates the host of the group, but the packet is sent only to one host. In this type of address, an address is given to several hosts, to define a group of nodes. If a packet is sent to this address, the router will send the packet to the nearest host that has the same Anycast address. In other words, the packet owner submits to the router the most "suitable" destination for sending the packet. The use of this Anycast Address is for example against several servers that provide services such as DNS (Domain Name Server). By providing the same Anycast Address to these servers, if a packet is sent by the client to this address, the router will select the nearest server and send the packet to that server. Thus, the load on the server can be distributed evenly. For this Anycast Address, no special space is provided. If several hosts are given the same address, then the address is considered an Anycast Address.
Figure 8. Sending Packets on Anycast Address
IPv6 Address Representation
Model x:x:x:x:x:x:x:x where 'x' is a hexadecimal value of 16 bits of the address portion, because there are 8 'x' then the total number is 16 * 8 = 128 bits. An example is:
FEDC:BA98:7654:3210:FEDC:BA98:7654:3210If the IPv6 addressing format contains a set of 16 bit address groups, namely 'x', which have a value of 0 then it can be represented as ::. An example is:
FEDC:0:0:0:0:0:7654:3210Can be represented as
FEDC::7654:3210 0:0:0:0:0:0:0:1Can be represented as
::1Model x:x:x:x:x:x:d.d.d.d where 'dddd' is an IPv4 address such as 167.205.25.6 which is used for automatic tunnelling. An example is:
0:0:0:0:0:0:167.205.25.6 atau ::167.205.25.6
0:0:0:0:0:ffff:167.205.25.7 atau :ffff:167.205.25.7So if you now access an address on the internet, for example 167.205.25.6, in time the format will be replaced with something like ::ba67:080:18. Like IPv4, IPv6 uses bit masks for subnetting purposes that are represented the same as the prefix-length representation in the CIDR technique used in IPv4, for example:
3ffe:10:0:0:0:fe56:0:0/60indicates that the first 60 bits are part of the network bits.
If in IPv4 you are familiar with the division of IP classes into classes A, B, and C, then in IPv6 the class division is also carried out based on the prefix format (FP), namely the initial bit format of the address. For example:
3ffe:10:0:0:0:fe56:0:0/60So if we pay attention to the first 4 bits, namely hexa '3', we get the prefix format for the first 4 bits is 0011 (which is the value of '3' hexa in binary).
Packet Structure In IPv6
In designing this packet header, it is attempted to make the cost/value of header processing small to support more real-time data communication. For example, the start and end addresses are needed in each packet. While in the IPv4 header when the packet is broken down, there is a field to store the sequence between packets. However, this field is not used when the packet is not broken down. The header in IPv6 consists of two types, the first, namely the field required by each packet is called the basic header, while the second, namely the field that is not always needed in the packet is called the extension header, and this header is defined separately from the basic header. The basic header is always present in each packet, while additional headers are only inserted between the basic header and the data if necessary. Additional headers, currently defined in addition to use when the packet is broken down, are also defined for security functions and others. This additional header is placed after the basic header, if several headers are needed, these headers will be connected in a chain starting from the basic header and ending with the data. The router only needs to process the smallest header needed, so that processing time becomes faster. The result of this improvement, although the basic header size has increased from 20 bytes to 40 bytes, the number of fields has been reduced from 12 to only 8.
Figure 9. Basic Header Structure in IPv6
Flow Label and Real Time Process
The IPv6 packet header has a flow-label field that is used to request that the packet be given a certain treatment by the router during delivery (giving a 'flag'). For example, in multimedia applications, it is transferred as quickly as possible even though the quality is slightly reduced, while e-mail or WWW requires more accuracy than real-time properties.
Table 2. Flow Label Table in IPv6
| Label | Kategori |
|--------|--------------------------------------------------------------------|
| 0 | Uncharacterized Traffic |
| 1 | "Filler" traffic (e.g., netnews) |
| 2 | Unattended data transfer (e.g., e-mail) |
| 3 | Reserved |
| 4 | Attended bulk transfer (e.g., FTP, HTTP, NFS) |
| 5 | Reserved |
| 6 | Interactive traffic (e.g., Telnet, X) |
| 7 | Internet control traffic (e.g., routing protocols, SNMP) |
| 8-15 | Realtime communications traffic, non-congestioncontrolled traffic |Routers manage priority scales and resources such as communication capacity or processing capability, based on these flow labels. If in IPv4 all packets are treated the same, then in IPv6 with different treatments for each packet, depending on the contents of the packet, applicable communication can be realized.
IPv6 Transition (IPv4-IPv6)
To overcome the obstacles of differences between IPv4 and IPv6 and to ensure communication between IPv4 users and IPv6 users, a Hosts – dual stack and Networks – Tunneling method was created on network hardware, for example routers and servers.
Figure 10. Network - Tunneling (IPv6 Transition)
So every time a router receives a packet, the router will sort the packet to determine the protocol used, then the router will forward it to the layer above it.
Understanding Classless Inter-Domain Routing (CIDR)
If we need an IP address with a host count of 500 with IP class C, then we must have 2 subnets. Because for class C the maximum host is 254. For each subnet, it must be entered into the routing table on the router device on the network.
This causes the number of entries in the routing table to swell and will drain device resources. To overcome this, Classless Inter-Domain Routing (CIDR) can be used. CIDR is routing that does not pay attention to the class of the IP address. CIDR is discussed in RFC 1518 to 1520.
Example: To connect 500 hosts with class C IP addresses, 2 subnets are required. The IP addresses used are 192.168.0.0/255.255.255.0 with 192.168.1.0/255.255.255.0, so the routing table on the router device also has 2 subnets. By using the CIDR routing table on the device, it is sufficient to use the address 192.168.0.0/255.255.252.0, with this only 1 routing table entry is required to connect to the network.
Understanding Fiber Distribution Data Interface (FDDI)
FDDI is a standard for fiber optic networks with a speed of 100Mbps. In the OSI Model FDDI is illustrated as in Figure 3.14. The RFC that describes FDDI is RFC 1188.
FDDI works by using 2 RING-shaped paths, where if damage occurs to a station, the previous station will create a loopback so that the network is not disconnected.
Figure 3.14 How FDDI works