IP address

An IP address is an address in computer networks, which - like the Internet - the Internet Protocol (IP) based. She is assigned to devices that are connected to the network, and makes the devices so addressed and so accessible. The IP address can be a single recipient or group of recipients designate ( multicast, broadcast). Conversely, there can be multiple IP addresses to a computer.

The IP address is used to transport data from sender to their intended recipients can. Similar to the mailing address on an envelope data packets with an IP address can be provided that clearly identifies the recipient. Because of this address, the " post office ", the routers, to decide in which direction the packet should be transported further. Unlike postal addresses IP addresses are not linked to a specific place.

The most well-known notation of the today common IPv4 addresses consist of four numbers, can take the values ​​from 0 to 255 and are separated by a dot, eg Technically, the address is a 32-digit (IPv4 ) or 128- digit (IPv6 ) binary number.

  • 3.1 Manual Configuration
  • 3.2 Automatic Configuration
  • 3.3 Dynamic Addressing
  • 3.4 Static Addressing
  • 3.5 Multiple addresses on one network card
  • 3.6 Different networks on one physical network


To establish a communication between two radios, each of the devices must be able to send data to the other device. To ensure delivery of this data to the correct remote location, this must be clearly identified (addressed). This is done in IP networks with an IP address. For example, a web server from a web browser is addressed directly by its IP address. The browser queries to a name server at the IP address from which (for example, " www.example.com " ) associated with a domain. He then uses this IP address to send data to the web server.

IP address in the IP data packets

Each IP data packet begins with an information field for the transport of the IP layer, the IP header. This header contains two fields, in which the IP addresses are entered both the sender and the receiver before the data packet is sent. The switching is done at layer 3 of the OSI model, the network layer.



The IPv4 addresses mainly used since the introduction of version 4 of the Internet Protocol consist of 32 bits, ie, 4 octets (bytes). This means that 232, so 4,294,967,296 addresses displayed. In dotted decimal notation, the 4 octets are written as four dot- separated decimal integers in the range 0 to 255,


Due to the rapidly increasing demand for IP addresses is foreseeable that the usable address space of IPv4 will be exhausted sooner or later. Mainly for this reason, IPv6 was developed. It uses 128 bits for storing addresses, so are 2128 = 25616 ( = 340.282.366.920.938.463.463.374.607.431.768.211.456 ≈ 3.4 × 1038 ) addresses displayed. This number is sufficient to provide the IP addresses for each square millimeter of the surface of at least 665.570.793.348.866.944 ( = 6.65 · 1017 ).

Since the decimal ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd.ddd confusing and would be difficult to handle, provides one IPv6 addresses in hexadecimal dar. this to simplify further presentation that summarizes two octets of the address and displayed in groups by a colon. XXXX: XXXX: XXXX: XXXX: XXXX: XXXX: XXXX: XXXX.

Zeros can be omitted at the beginning of a block to further shortening. One or more consecutive blocks, which consist only of zeros can be replaced by :: - but no more than once in the address can be so clearly made ​​up to eight blocks.

Network part and device part

Each IPv4 address is separated by a network mask, any IPv6 addresses by specifying the prefix length, in a network and a device component ( " host part "). The network mask, so the prefix length, which specifies which bit of the address must be shared. The on the net mask masked or referred to by the prefix length bits ( network part ) are in all hosts ( computers ) a subnetwork identical. The information whether a device is on the same subnet (ie the same network part of IP address) that is needed from a host in order to make routing decisions (see following section).

Decimal Binary calculation IP address 203 000 113 195 11001011 00000000 01110001 11000011 ip address Netmask netmask 11111111 11111111 11111111 11100000 AND Netzwerkadr. 203 000 113 192 11001011 00000000 01110001 11000000 = network part IP address 203 000 113 195 11001011 00000000 01110001 11000011 ip address Netmask 11111111 11111111 11111111 11100000                                 00000000 00000000 00000000 00011111 AND ( NOT netmask ) Unit part 3 00000000 00000000 00000000 00000011 = device- part In a network mask with 27 bits set results in a network address of There remain 5 bits and thus 25 = 32 addresses for the device part. Of these, an address for the network itself and for the broadcast are still ever needed so that there are 30 addresses for devices.


Will a device send an IP packet, the network part of the source IP address and destination IP address are compared. If they match, the destination host in the same network, and the packet is sent directly to the recipient. In the case of Ethernet networks, the ARP (Address Resolution Protocol) is used to find the hardware address. The ARP operates on the second layer of the OSI model and establishes the connection to the first layer.

If the network parts, however, do not match, then it will be a routing table the IP address of a router (next hop ) is searched and sent the packet to this router. This has one or more interfaces contact with other networks and routes the packet using the same method more - he consulted his part his own routing table and sends the packet if necessary to the next router or the destination. Up to the terminal, the package can go through many networks and routers. Running through a router is also called hop ( jump), the routing method next hop routing.

A router has this for each interface has its own IP address and network mask, which belongs to the respective network. Each IP packet is routed individually. The source and destination address in the IP header are set by the sender and remain unchanged during the entire journey.

Special IP addresses

Special IPv4 addresses according to RFC 6890:

After this list meet 622 199 809 of around 4.3 billion IPv4 addresses, or 14.5 % of all possible IPv4 addresses a particular purpose.

Thus, there are three types of IP addresses:

  • Unicast: Send to a certain recipients on the Internet (normal addressing).
  • Broadcast: send to all devices in the same network (subnet). This is replaced by IPv6 multicast.
  • Multicast: Send some devices in the same network ( or devices in the multicast backbone).

No longer reserved IP addresses

With the RFC 5735 about 50 million IP addresses were released. The reservation of the following address ranges has been lifted and released for distribution.

DNA - translation of host names to IP addresses

Via a global Domain Name System (DNS) can resolve names to IP addresses ( and vice versa). The name www.example.com results, for example, the IPv4 address, the name www.ipv6.uni -muenster.de the IPv6 address 2001:638:500:101:2 e0: 81ff: FE24: 37c6.

Assignment of IP addresses and network areas

IANA - Internet Assigned Numbers Authority

The allocation of IP networks in the Internet is regulated by the IANA. In the early years of the Internet, IPv4 addresses or networks were awarded directly in large blocks from the IANA to organizations, companies or universities. For example, the range / 8 and therefore 16,777,216 addresses of Xerox Corporation has been allocated. Merck & Co. received from the IANA also an area of 16,777,216 addresses ( / 8), as well as IBM ( / 8). The only German company that has been appointed a / 8 range, the debis AG ( / 8). Today, the IANA allocates blocks to regional registries.

RIR - Regional Internet Registry

Since February 2005, there are five regional registries, which are called Regional Internet Registries (RIRs ):

  • AfriNIC (African Network Information Centre) - responsible for Africa
  • APNIC (Asia Pacific Network Information Centre) - responsible for the Asia -Pacific region
  • ARIN ( American Registry for Internet Numbers ) - North America
  • LACNIC ( Latin American and Caribbean Network Information Centre) - Latin America and Caribbean
  • RIPE NCC ( Réseaux IP Européens Network Coordination Centre ) - Europe, Middle East, Central Asia.

For Germany, Liechtenstein, Austria and Switzerland so the RIPE NCC is responsible.

The Regional Internet Registries assign allocated to them by the IANA networks to local registries.

LIR - Local Internet Registry

The Local Internet Registries (LIR ) said local awarding bodies give their allocated by the RIRs addresses to their customers. The aim of the LIR usually meet provider Internet service. Customers of the LIR can be either end-users or other ( sub-) provider.

The addresses may be permanently assigned to the customer either ( static IP address, static IP address ) or when connecting to the Internet are assigned dynamically (dynamic IP address, dynamic IP address). Permanently assigned addresses are used especially for leased lines or if servers are to be operated on the IP address.

To what end users or what Local Internet Registry IP address or a subnet has been assigned, can be determined via the Whois databases of the RIRs.

Private Networks

In private, local area networks (LAN ), even IP addresses are assigned. To this should be used for IPv4 addresses from the private networks referred to in RFC 1918 (for example,,, ...). These addresses are not assigned by the IANA and not routed on the Internet. To enable nevertheless an Internet connection, the LAN internal addresses are translated into public, valid in the Internet IPv4 addresses in a router using Network Address Translation. For packets that arrive addressed to the public address, public address in turn translated back into the private addresses. Additionally, NAT enables all computers on the local network appear to the outside under the same (ie only one) valid in the Internet IPv4 address, which " addresses saves ". The assignment of a communication between a local computer with a private address and server on the Internet takes place via the port number.

Net classes

Originally, IPv4 addresses were divided into network classes A to C with different network masks. Classes D and E were designed for specific tasks. Due to the ever-growing routing tables as the classless routing CIDR (Classless Inter Domain Routing ) was introduced in 1993. So it does not matter which network class belongs to an IPv4 address.

Device Configuration

Manual Configuration

For administrators, there are configuration utilities. Under Linux, this is ip, netsh on Windows, and other Unix-like operating systems use ifconfig. To display selbige can be used with Windows ipconfig or winipcfg also available ( depending on version).

  • Windows: netsh dump
  • Windows: netsh interface ipv6 add address interface = "Local Area Connection 1" address = 2001: db8 :: 123
  • Unix (FreeBSD, Mac OS X): ifconfig eth0
  • Older ifconfig: ifconfig eth0 netmask broadcast

The specification of the parts "broadcast " or " brd " are optional. ( " Brd " stands for the automatic calculation of the broadcast address, it can also be a specific address can be specified. Ifconfig calculates the broadcast address in newer versions automatically ).

Automatic configuration

About protocols such as BOOTP or DHCP can assign IP addresses when booting the computer by a corresponding server. A range of IP addresses on the server to the administrator defines from which other computers can see an address at startup. This address is leased to the computer. Hosts that need fixed addresses can be identified in the Ethernet network via its MAC address and get a permanent address.

Advantage is the central management of the address. If automatic configuration provided by the installation of the operating system, no further settings for network access must be more done. Mobile devices such as laptops can share addresses, if not all devices can be connected simultaneously to the grid. In addition, they can be integrated without changing the configuration in different networks (for example, company, customer network, home network ), if necessary.

For IPv6, there is also the possibility of auto-configuration, which does not require server.

Dynamic Addressing

If a host for every new connection to a network is assigned a new IP address, it is called dynamic or changing address. In the LAN area dynamic addressing is distributed via DHCP, the Internet access area dynamic addressing mainly by Internet Service Providers ( ISP) is used, which offer Internet access via dial-up lines. You use dynamic addressing via PPP or PPPoE.

Advantage of the dynamic addressing is that much less than one IP address per customer is required, on average, since all customers are never online at the same time. A ratio of 1:10 to 1:20 is common. The RIPE NCC requires its LIRs evidence of the use of their assigned IP addresses. A fixed allocation of addresses is accepted only in justified cases, for example for the operation of servers or for billing purposes.

With DSL connection of the customer providers typically use also dynamically assigned IPs.

Static Addressing

Static addressing is used basically anywhere where a dynamic addressing is not technically possible or not meaningful. So get in LANs, for example, gateways, servers or network printer usually fixed IP addresses. In the Internet area static addressing is used mainly for routers to dedicated lines. Also for machine-to -machine communication is increasingly used static addressing in particular in mobile communications ( GPRS). Static addresses are usually configured manually, but can also use automatic addressing (see above ) can be assigned.

Multiple addresses on one network card

Most is (for example, network card) of a host assigned to exactly one IPv4 address for each network interface. In some cases ( see below), however, it is necessary to assign more IPv4 addresses to an interface. This is also known as an IP alias. Multiple IPv4 addresses on one network card can be used, among other things, to operate several of the same services there in parallel to make a host of different subnets reachable or a service logically be separated from the host so that it - with its IPv4 address and transparent can be moved to another hardware - for the clients.

With IPv6, the binding of several addresses to a network interface the rule to operate, for example, a link- local addition to a global address and dynamically assigned prefixes in addition to fixed or IPv6 addresses of multiple Internet providers to have on the same host. In addition, the above-mentioned reasons as for IPv4 apply.

Different networks on one physical network

On a physical network (eg Ethernet ), different networks can be (with different network address part ) attached and used simultaneously. This is used, among other things, if later, the network should be split up or if previously separate networks are combined.

Storage of IP addresses

A final legal assessment of storing IP addresses in Germany has not yet come about.

The German Federal Constitutional Court ruled on 2 March 2010 that the storage of IPs in Germany was unconstitutional in its present implementation, since the law on occasion without storing large amounts of data of all users of electronic communications services does not provide for concrete measures for data security, and also has its barriers to retrieval of this data is considered too low. The judgment committed German telecommunications provider to immediately delete the previously collected data. It found, however, that data retention under stringent safety and transparency measures and limited demand opportunities for the security agencies is generally permitted.

An information request of the prosecutor is then comply, if this is done in the context of an investigation of a serious crime within the meaning of § 100a paragraph 2 Code of Criminal Procedure. The storage of IP addresses for other purposes ( for example, when you visit a website, such as a log file ) but is not legally a complete mystery.

The district court Mitte ( Berlin), in March 2007 IP addresses to personal information within the meaning of § 3 BDSG, thus their storage would not be permitted. The District Court of Munich decided in late September 2008 that IP addresses are not to be regarded as personal data, hence their storage would generally permitted. However, the Court tied this to some specifications. Thus, the admissibility of the storage of the capabilities of that depends on, which stores the data. Has he in principle the possibility of a person to identify by their IP address ( possibly using a personalized user account ), automatic storage is not permitted. In this case, this is only allowed if the user has given his explicit consent.

It should be noted that both judgments regarding IPv4 addresses were issued, due to the larger address range of IPv6 addresses, the latter are legally classified differently under certain circumstances.

Recovery of additional information

Although an IP address itself does not directly contains additional information, additional information about the user that uses it, or won will be accepted with a certain probability with their help:

  • Geotargeting tried the real residence (at least state, region ) infer ( localization ).
  • At times, also the probability of a certain social role are determined: For example, content made available a fixed IP address of a company or a government department have a high probability to be considered official statements of this company / institution.
  • For example, if Sony assigns the owners of web -enabled Sony TVs IP addresses from a particular IP range via DHCP, can thus (eg Google ) server from the IP address conclude that a matching user as " viewers " should be classified.