IP Addressing

Introductory material.

An entire module devoted to IP addresses.

IP Addresses

• Structure of an IP address
• Classful IP addresses
• Limitations and problems with classful IP addresses
• Subnetting
• CIDR
• IP Version 6 addresses
IP Addresses

32 bits
version header Type of Service/TOS Total Length (in bytes)
(4 bits) length (8 bits) (16 bits)
flags
Identification (16 bits) Fragment Offset (13 bits)
(3 bits)
TTL Time-to-Live Protocol
Header Checksum (16 bits)
(8 bits) (8 bits)

Source IP address (32 bits)

Destination IP address (32 bits)

Ethernet Header IP Header TCP Header Application data Ethernet Trailer

Ethernet frame
IP Addresses

32 bits
0x4 0x5 0x00 4410

9d08 0102 00000000000002

12810 0x06 8bff

128.143.137.144

128.143.71.21

Ethernet Header IP Header TCP Header Application data Ethernet Trailer

Ethernet frame
What is an IP Address?

• An IP address is a unique global address for a network

interface

• An IP address:
- is a 32 bit long identifier
- encodes a network number (network prefix)
and a host number
Dotted Decimal Notation

• IP addresses are written in a so-called dotted decimal

notation
• Each byte is identified by a decimal number in the range
[0..255]:

• Example:
10000000 10001111 10001001 10010000
1st Byte 2nd Byte 3rd Byte 4th Byte
= 128 = 143 = 137 = 144

128.143.137.144
Network prefix and Host number

• The network prefix identifies a network and the host number

identifies a specific host (actually, interface on the network).

network prefix host number

• How do we know how long the network prefix is?

– The network prefix is implicitly defined (see class-based
addressing)
– The network prefix is indicated by a netmask.
Example

• Example: ellington.cs.virginia.edu

128.143 137.144

• Network id is: 128.143.0.0

• Host number is: 137.144
• Network mask is: 255.255.0.0 or ffff0000

• Prefix notation: 128.143.137.144/16

» Network prefix is 16 bits long
The old way: Classful IP Adresses

• When Internet addresses were standardized (early 1980s),

the Internet address space was divided up into classes:
– Class A: Network prefix is 8 bits long
– Class B: Network prefix is 16 bits long
– Class C: Network prefix is 24 bits long

• Each IP address contained a key which identifies the class:

– Class A: IP address starts with “0”
– Class B: IP address starts with “10”
– Class C: IP address starts with “110”
The old way: Internet Address Classes

bit # 0 1 7 8 31

Class A 0
Network Prefix Host Number
8 bits 24 bits

bit # 0 1 2 15 16 31

Class B 10 network id host

Network Prefix Host Number

16 bits 16 bits

bit # 0 1 2 3 23 24 31

Class C 110 network id host

Network Prefix Host Number

24 bits 8 bits
The old way: Internet Address Classes

bit # 0 1 2 3 4 31

Class D 1110 multicast group id

bit # 0 1 2 3 4 5 31

Class E 11110 (reserved for future use)

• We will learn about multicast addresses later in this course.

Problems with Classful IP Addresses

• The original classful address scheme had a number

of problems

Problem 1. Too few network addresses for large

networks
– Class A and Class B addresses are gone

Problem 2. Two-layer hierarchy is not appropriate

for large networks with Class A and Class B
addresses.
– Fix #1: Subnetting
Problems with Classful IP Addresses

Problem 3. Inflexible. Assume a company requires 2,000

addresses
– Class A and B addresses are overkill
– Class C address is insufficient (requires 8 Class C
addresses)

– Fix #2: Classless Interdomain Routing (CIDR)

Problems with Classful IP Addresses

Problem 4: Exploding Routing Tables: Routing on the

backbone Internet needs to have an entry for each network
address. In 1993, the size of the routing tables started to
outgrow the capacity of routers.

– Fix #2: Classless Interdomain Routing (CIDR)

Problems with Classful IP Addresses

Problem 5. The Internet is going to outgrow the 32-

bit addresses

– Fix #3: IP Version 6

Subnetting

• Problem: Organizations
have multiple networks
University
UniversityNetwork
Network
which are independently
managed Engineering Medical
– Solution 1: Allocate one or School School
more Class C address for
each network
• Difficult to manage Library
• From the outside of the
organization, each network
must be addressable.
– Solution 2: Add another
level of hierarchy to the
Subnetting
IP addressing structure
Basic Idea of Subnetting

• Split the host number portion of an IP address into a

subnet number and a (smaller) host number.
• Result is a 3-layer hierarchy

network prefix host number

network prefix subnet number host number

• Then: extended network prefix

• Subnets can be freely assigned within the organization
• Internally, subnets are treated as separate networks
• Subnet structure is not visible outside the organization
Subnet Masks

• Routers and hosts use an extended network prefix (subnet

mask) to identify the start of the host numbers

Class B 10 network host

16 bits
Network Prefix (16 bits)

with 10 network subnet host

subnetting
Extended Network Prefix (24 bits)
Subnet
mask 1111111111111111111111100000000
(255.255.255.0)

* There are different ways of subnetting. Commonly used netmasks for university
networks with /16 prefix (Class B) are 255.255.255.0 and 255.255.0.0
Typical Addressing Plan for an Organization that
uses subnetting
• Each layer-2 network (Ethernet segment, FDDI segment) is
allocated a subnet address.

128.143.71.0 / 24

128.143.0.0/16 128.143.16.0 / 24

128.143.7.0 / 24 128.143.8.0 / 24

128.143.17.0 / 24

128.143.22.0 / 24

128.143.136.0 / 24
Advantages of Subnetting

• With subnetting, IP addresses use a 3-layer hierarchy:

» Network
» Subnet
» Host
• Improves efficiency of IP addresses by not consuming an
entire Class B or Class C address for each physical network/
• Reduces router complexity. Since external routers do not
know about subnetting, the complexity of routing tables at
external routers is reduced.

• Note: Length of the subnet mask need not be identical at all

subnetworks.
CIDR - Classless Interdomain Routing

• IP backbone routers have one routing table entry for each

network address:
– With subnetting, a backbone router only needs to know one entry for
each Class A, B, or C networks
– This is acceptable for Class A and Class B networks
• 27 = 128 Class A networks
• 214 = 16,384 Class B networks
– But this is not acceptable for Class C networks
• 221 = 2,097,152 Class C networks
• In 1993, the size of the routing tables started to outgrow the
capacity of routers
• Consequence: The Class-based assignment of IP addresses
had to be abandoned
CIDR - Classless Interdomain Routing

• Goals:
– Restructure IP address assignments to increase efficiency
– Hierarchical routing aggregation to minimize route table
entries

• CIDR (Classless Interdomain routing) abandons the notion of

classes:
Key Concept: The length of the network id (prefix) in the IP
addresses is kept arbitrary

• Consequence: Routers advertise the IP address and the

length of the prefix
CIDR Example

• CIDR notation of a network address:

192.0.2.0/18
• "18" says that the first 18 bits are the network part of the
address (and 14 bits are available for specific host
addresses)
• The network part is called the prefix

• Assume that a site requires a network address with 1000 addresses

• With CIDR, the network is assigned a continuous block of 1024 addresses
with a 22-bit long prefix
CIDR: Prefix Size vs. Network Size

CIDR Block Prefix # of Host Addresses

/27 32 hosts
/26 64 hosts
/25 128 hosts
/24 256 hosts
/23 512 hosts
/22 1,024 hosts
/21 2,048 hosts
/20 4,096 hosts
/19 8,192 hosts
/18 16,384 hosts
/17 32,768 hosts
/16 65,536 hosts
/15 131,072 hosts
/14 262,144 hosts
/13 524,288 hosts
CIDR and Address assignments

• Backbone ISPs obtain large block of IP addresses space and

then reallocate portions of their address blocks to their
customers.

Example:
• Assume that an ISP owns the address block 206.0.64.0/18, which
represents 16,384 (214) IP addresses
• Suppose a client requires 800 host addresses
• With classful addresses: need to assign a class B address (and
waste ~64,700 addresses) or four individual Class Cs (and introducing 4
new routes into the global Internet routing tables)
• With CIDR: Assign a /22 block, e.g., 206.0.68.0/22, and allocated a
block of 1,024 (210) IP addresses.
CIDR and Routing Information

Company X :
206.0.68.0/22
ISP X owns:
206.0.64.0/18
204.188.0.0/15
209.88.232.0/21
Internet
Backbone ISP y :
209.88.237.0/24

Organization z1 : Organization z2 :
209.88.237.192/26 209.88.237.0/26
CIDR and Routing Information
Backbone routers do not know
anything about Company X, ISP
Y, or Organizations z1, z2.
Company X :

ISPISP 206.0.68.0/22
ISP X does not know about X owns:
y sends everything which matches
Organizations z1, z2. the prefix:
206.0.64.0/18 to Organizations z1
209.88.237.192/26
204.188.0.0/15to Organizations z2
209.88.237.0/26
209.88.232.0/21
Internet
ISP X sends everything which
Backbone
matches the prefix: ISP y :
206.0.68.0/22 to Company X,
209.88.237.0/24 to ISP y 209.88.237.0/24

Backbone sends everything

which matches the prefixes
Organization z1 : Organization z2 :
206.0.64.0/18, 204.188.0.0/15,
209.88.232.0/21 to ISP X. 209.88.237.192/26 209.88.237.0/26
 You can find about ownership of IP addresses in
Example North America via http://www.arin.net/whois/

• The IP Address: 207.2.88.170

207 2 88 170

11001111 00000010 01011000 10101010

Belongs to:
City of Charlottesville, VA: 207.2.88.0 - 207.2.92.255

11001111 00000010 01011000 00000000

Belongs to:
Cable & Wireless USA 207.0.0.0 - 207.3.255.255

11001111 00000000 00000000 00000000

CIDR and Routing

• Aggregation of routing table entries:

– 128.143.0.0/16 and 128.144.0.0/16 are represented as
128.142.0.0/15
• Longest prefix match: Routing table lookup finds the
routing entry that matches the the longest prefix

Prefix Interface
What is the outgoing interface for
128.0.0.0/4 interface #5
128.143.137.0/24 ?
128.128.0.0/9 interface #2
128.143.128.0/17 interface #1

Routing table
IPv6 - IP Version 6

• IP Version 6
– Is the successor to the currently used IPv4
– Specification completed in 1994
– Makes improvements to IPv4 (no revolutionary changes)

• One (not the only !) feature of IPv6 is a significant increase in

of the IP address to 128 bits (16 bytes)
• IPv6 will solve – for the foreseeable future – the
problems with IP addressing
IPv6 Header
32 bits
version Traffic Class Flow Label
(4 bits) (8 bits) (24 bits)
Next Header
Payload Length (16 bits) Hop Limits (8 bits)
(8 bits)

Source IP address (128 bits)

Destination IP address (128 bits)

Ethernet Header IPv6 Header TCP Header Application data Ethernet Trailer

Ethernet frame
IPv6 vs. IPv4: Address Comparison

• IPv4 has a maximum of

232  4 billion addresses
• IPv6 has a maximum of
2128 = (232)4  4 billion x 4 billion x 4 billion x 4 billion
addresses
Notation of IPv6 addresses

• Convention: The 128-bit IPv6 address is written as eight 16-

bit integers (using hexadecimal digits for each integer)
CEDF:BP76:3245:4464:FACE:2E50:3025:DF12

• Short notation:
• Abbreviations of leading zeroes:
CEDF:BP76:0000:0000:009E:0000:3025:DF12

CEDF:BP76:0:0:9E :0:3025:DF12
• “:0000:0000:0000” can be written as “::”
CEDF:BP76:0:0:FACE:0:3025:DF12  CEDF:BP76::FACE:0:3025:DF12
• IPv6 addresses derived from IPv4 addresses have 96 leading zero bits.
Convention allows to use IPv4 notation for the last 32 bits.
::80:8F:89:90  ::128.143.137.144
IPv6 Provider-Based Addresses

• The first IPv6 addresses will be allocated to a provider-based

plan

Registry Provider Subscriber Subnetwork Interface

010
ID ID ID ID ID

• Type: Set to “010” for provider-based addresses

• Registry: identifies the agency that registered the address
The following fields have a variable length (recommeded length in “()”)
• Provider: Id of Internet access provider (16 bits)
• Subscriber: Id of the organization at provider (24 bits)
• Subnetwork: Id of subnet within organization (32 bits)
• Interface: identifies an interface at a node (48 bits)
More on IPv6 Addresses

• The provider-based addresses have a similar flavor as CIDR

addresses

• IPv6 provides address formats for:

– Unicast – identifies a single interface
– Multicast – identifies a group. Datagrams sent to a
multicast address are sent to all members of the group
– Anycast – identifies a group. Datagrams sent to an anycast
address are sent to one of the members in the group.