Understanding IPv4 and IPv6 Addressing

1. The Need for IP Addressing

1.1 The Analogy: Sending a Letter
• The Message: The data you want to send.
• Source Address: Your IP address. Without it, the recipient cannot reply.
• Destination Address: The recipient's IP address. Without it, network devices (routers)
cannot deliver the message.
• Conclusion: An IP Address is a Layer 3 logical address essential for communication
between networks.
Think of the IP header as an envelope. The data is the letter inside. The source IP is the "return
address," and the destination IP is the "delivery address." If either is missing or incorrect, the postal
service (the network) cannot function.

1.2 The Governing Bodies

• IANA (Internet Assigned Numbers Authority): The global coordinator for IP address
allocation.
• RIRs (Regional Internet Registries): IANA delegates address distribution to five RIRs
based on geography.
• African Network Information Center (AFRINIC): Supports the continent of Africa
• Asia-Pacific Network Information Centre (APNIC): Supports regions of Asia and the
Pacific
• American Registry for Internet Numbers (ARIN): Supports regions of Canada, the
USA, and parts of the Caribbean
• Latin America and Caribbean Network Information Centre (LACNIC): Supports
Latin America and parts of the Caribbean regions
• Réseaux IP Européens Network Coordination Centre (RIPE NCC): Supports Europe,
the Middle East, and Central Asia

Figure 1: IP network blocks delegation

An Autonomous System (AS) is simply any organization that is responsible for managing a large
number of internet routing networks such as an ISP within a country or region. IANA assigns AS
numbers to the various RIRs around the world. Then, the RIRs allocate these AS numbers to
network operators such as ISPs. The ISPs use their AS numbers to share their network routing
prefixes with other ISPs using the Border Gateway Protocol (BGP).
This hierarchical system prevents chaos. It ensures that two different ISPs in different parts of the
world don't assign the same public IP address to different customers, which would cause routing
conflicts.

Figure 2: Interconnected ISPs sharing network routes

1.3 The Core Problem: Address Exhaustion

• IPv4: Deployed in 1983. It is a 32-bit address, providing ~4.3 billion addresses.
• IPv6: Developed in 1999. It is a 128-bit address, providing a virtually limitless number.
The explosion of internet-connected devices (computers, phones, IoT sensors) meant that 4.3 billion
was not enough. IPv6 was created as a long-term solution.

2. Public vs. Private Address Spaces

To combat IPv4 exhaustion, IANA created two separate address spaces.

Figure 3: IPv4 public address space

2.1 Public IP Address Space
• Definition: Addresses that are routable on the internet.
• Must be unique across the entire internet.
• Assigned to devices directly connected to the internet (e.g., web servers, your home router's
external interface).

2.2 The Problem with Classful Addressing

• Inflexible: Organizations were forced to use the default subnet mask.
• Massive Wastage: A company with 2,000 devices needed a Class B block (65,534
addresses), wasting over 63,000 addresses.
• Large Routing Tables: Every classful network was a separate entry in a router's table,
slowing down internet routing.

Figure 4: Classful IPv4 addresses

2.3 Private IP Address Space (RFC 1918)

• Definition: Addresses that are NOT routable on the internet.
• Can be reused by any private network globally.
• Purpose: To conserve public IPv4 addresses.

Figure 5: Private IPv4 address space

Private IPs are like internal extension numbers in a large office building. The building has one
public street address (public IP), but inside, every office has a unique extension (private IP). You
can dial extension 101 ([Link]) from inside, but someone outside the building cannot call
101 directly; they must call the main number (public IP) first.
A device with only a private IP cannot communicate directly with the internet. It needs a translator
—Network Address Translation (NAT).
3. Network Address Translation (NAT)
Network Address Translation (NAT) is an IP service that is commonly found on almost all private
networks within organizations. NAT allows a private IPv4 source address to be translated into a
public IPv4 address, allowing devices on a private network to communicate with devices on a
public network.

Figure 6: NAT operations

3.1 How NAT Works

1. A PC with a private IP (e.g., [Link]) sends a packet to a web server on the
internet.
2. The packet arrives at the NAT-enabled router.
3. The router translates the source IP from the private address ([Link]) to its own
public IP (e.g., [Link]).

4. The web server replies to the public IP ([Link]).

5. The router translates the destination address back to the private IP ([Link]) and
forwards the packet.
A. Advantages

• Conserves the Public IPv4 Address Space: The primary benefit of using NAT is to help
conserve the public IPv4 address space by allowing organizations to use private IPv4
addresses on their internal networks and be assigned a single public IPv4 address on their
modem/router.

• Hides the Internal Network: Using NAT allows an entire organization’s private network to
be hidden behind a single public IPv4 address.

• Maintains Internal Addressing Consistency: Network professionals can maintain

consistency in their private IPv4 addressing scheme within their companies.
B. Disadvantages

• Impacts Network Performance: As traffic is sent to the router or modem to be translated,

there is some delay as the router or modem has to perform the actual translation process.

• Incompatibility with IPsec VPNs: Since NAT modifies the IPv4 addresses on the packet,
Virtual Private Network (VPN) solutions that use IP security (IPsec) to establish a secure
logical tunnel over an unsecure network does not work well.

• Loss of End-to-End Connectivity: NAT modifies the IPv4 addresses within the packet, so
end-to-end connectivity is lost between a sender and receiver.

3.2 Types of NAT

A. Static NAT

Figure 7: Static NAT

• One-to-One mapping. A single private IP is permanently mapped to a single public IP.

• Use Case: Hosting an internal server (e.g., a web or email server) that needs to be accessible
from the internet.
Like having a dedicated public phone line for the CEO's office. All calls to that public
number go directly to one specific internal extension.
B. Dynamic NAT
• Many-to-Many mapping. A pool of public IPs is shared by many private IPs on a first-
come, first-served basis.
• Use Case: Less common. Used when you have a limited number of public IPs and a larger
number of internal devices that need occasional internet access.
Like a company with 10 external phone lines for 50 salespeople. The first 10 salespeople to
make a call get a line. The 11th must wait until a line is free.
 Figure 8: Dynamic NAT

C. Port Address Translation (PAT) / NAT Overload

• Many-to-One mapping. This is the most common type, used in every home router.
• It uses source port numbers to track thousands of internal connections using a single public
IP.
This is like the receptionist using a different "department code" for each employee's
outgoing mail. All mail leaves with the same return address ([Link]), but with a
unique department code (e.g., [Link]:15001, [Link]:15002). When
replies come back addressed to a specific department code, the receptionist knows exactly
which employee to deliver it to.

Figure 9: Port address translation

4: The Structure of IPv4 and IPv6
4.1 Fundamentals of IPv4: It's All About Binary
• The Core Concept: An IPv4 address is a 32-bit binary number. The dotted-decimal format
([Link]) is just for human readability.

• Conversion is Key: To truly understand subnetting, you must be able to convert between
decimal and binary.
Converting Binary to Decimal (The Power of Two Method):
Each octet is 8 bits. Each bit represents a power of 2, from 2⁷ (128) on the left to 2⁰ (1) on the right.
You add the values for every bit that is a '1'.
Example: Convert 11000000 to Decimal

Figure 10: Converting the first octet 11000000

Converting Decimal into Binary:

The following are some guidelines to ensure the results are accurate:
• Always convert one octet at a time
• Always begin by subtracting the highest power of 2, which is 27 = 128, while working
• downwards to the lowest power of 2, which is 20 = 1
• If you can subtract the decimal number from the radix value, place a 1
• If you are unable to subtract the decimal number from the radix value, place a 0
• If you get a 0, then subtract the decimal number from the next lower radix value

Figure 11: Converting 172

into binary
Example: Convert the IP address [Link]

• 192 -> 128+64 -> 11000000

• 168 -> 128+32+8 -> 10101000

• 1 -> 1 -> 00000001

• 1 -> 1 -> 00000001

• Full Binary: 11000000.10101000.00000001.00000001

4.2 Fundamentals of IPv6: A New Paradigm

• Length: 128 bits, written as eight 16-bit hexadecimal blocks.
• Example Full Address: 2001:0db8:85a3:0000:0000:8a2e:0370:7334

• Simplification Rules (Crucial for readability):

1. Remove Leading Zeros: In each block, you can remove leading zeros.
• 0db8 -> db8

• 0370 -> 370

• Result: 2001:db8:85a3:0000:0000:8a2e:370:7334

2. Compress Successive Zero Blocks: Replace the longest contiguous block of 0000
blocks with a double colon ::. This can only be done once per address.

• 2001:db8:85a3:0000:0000:8a2e:370:7334 ->
2001:db8:85a3::8a2e:370:7334

Figure 12: IPv6 address structure

As shown in the preceding snippet, the first three hextets represent the Global Routing Prefix
portion, which contains the first 48 bits of the address. This portion of the IPv6 address is assigned
by the service provider such as the ISP. The fourth hextet (16 bits) represents the Subnet ID, which
is used by the ISP to create subnetworks of the network block. The last 64 bits represent the
Interface ID portion. Combining the Global Routing Prefix, Subnet ID, and Interface ID, a client is
assigned a unique 128-bit IPv6 address on its network interface card.

5. Types of IPv4 and IPv6 Addresses

5.1 Automatic Private IP Addressing (APIPA)
What is APIPA?
• A failover mechanism built into Microsoft Windows operating systems
 • Range: [Link] to [Link] with subnet mask [Link]

• Automatically activates when DHCP client cannot locate a DHCP server

How APIPA Works:
1. DHCP Discovery Failure: Device boots up and broadcasts DHCP Discover messages
2. Timeout Period: Device waits for DHCP Offer (typically 30-60 seconds)
3. Self-Assignment: If no DHCP server responds, the device automatically assigns itself an IP
from the APIPA range
4. Conflict Detection: Device performs duplicate address detection (DAD) to ensure the
selected IP isn't already in use on the local network

Figure 13: APIPA address on a client

5.2. EUI-64 (Extended Unique Identifier) Fundamentals

What is EUI-64?
Definition:
• A standard method defined by IEEE for creating 64-bit interface identifiers
• Converts 48-bit MAC addresses into 64-bit interface identifiers
• Primarily used in IPv6 Stateless Address Autoconfiguration (SLAAC)
Purpose in IPv6:
• Provides automatic generation of unique interface IDs
• Ensures global uniqueness when combined with IPv6 network prefix
• Enables devices to self-configure IPv6 addresses without DHCP
The following steps describe the EUI-64 process of converting a 48-bit MAC address of a network
adapter to create a 64-bit address that will be used as the interface ID portion to create a unique
IPv6 address:

1. First, split the 48-bit MAC address into half, separating the Organizational Unique Identifier
(OUI) and the device portions, as shown in the following diagram:

Figure 14: EUI-64 process step 1

2. Next, insert the hexadecimal value, FFFE, in the middle of the 48-bit MAC address, as
shown in the following diagram:

Figure 15: EUI-64 process step 2

3. Next, flip the seventh bit within the first byte so that a 0 will become a 1 or a 1 will become
a 0. This bit indicates if the NIC is administered locally (0) or is globally unique (1), as
shown in the following diagram:

Figure 16: EUI-64 process step 3

4. Next, convert the binary into hexadecimal to view the EUI-64 portion of the address, as
shown in the following diagram:

Figure 17: EUI-64 process step 4

5. Lastly, putting it all together, the EUI-64 bit IPv6 address that will be assigned to the
device’s network adapter is 2001:DB8:0:1111:FE99:47FF:FE75:CEE0.

5.3. Unicast
• Unicast addresses can be IPv4 and IPv6 addresses that are uniquely assigned to the network
adapter of a device.
• Using a unicast address allows one-to-one communication between a sender and receiver
device over a network.
• As shown in the diagram, each device is assigned a unique IPv4 address on the network.
• Additionally, the IP address that’s configured on each device is a unique unicast address,
which allows any device to communicate with another device.
 Figure 18: Unicast communication

5.4. Multicast
• Multicast addresses exist within both the IPv4 and IPv6 address spaces. Multicast allows
one-to-many communication between devices on a network.
• Enterprise-grade routers within a large organization are usually configured with a dynamic
routing protocol, which allows each router to automatically learn new networks and
maintain an up-to-date routing table by exchanging routing information between themselves.
• These routers send and receive messages to a multicast address group, which is only used by
devices running the same dynamic routing protocol.

Figure 19: Multicast communication

5.5. Broadcast
• Broadcast allows one-to-all communication over an IPv4 network. Unlike IPv4 networks,
IPv6 does not use broadcast addresses.
• Simply put, a computer that’s connected to a network can send a single message to the
network’s broadcast address, which allows all devices within the same IP network to receive
the message from the sender.
• Within a network that uses a Network ID of [Link] and has a subnet mask of
[Link], the broadcast IP address will be [Link].
 Figure 20: Broadcast communication

5.6. Anycast
• Anycast addresses exist on IPv6 networks and are used for one-to-nearest communication
between devices over a network.
• Anycast allows the same unicast IPv6 address to be shared between multiple devices, such
as servers on the internet.
• When a client such as a computer sends a message to an anycast IPv6 address, the message
is delivered to the geographically nearest server that’s configured with the anycast address.

5.7. Link-local
• On an IPv6 network, each device is assigned two IPv6 addresses on its network adapter. One
of the addresses is the IPv6 unicast address, which is used by the device when
communicating with other hosts outside its local network.
• If the computer on an internal network within an organization wants to communicate with a
server on the internet, the IPv6 unicast address will be used for this type of communication.
• Additionally, an IPv6 Link-Local address is assigned to the same network adapter with the
IPv6 unicast address.
• A Link-Local address is an IP address automatically assigned to a device for
communication within a single local network segment (link) — when no external
configuration (like DHCP or manual setup) is available.

5.8. Loopback
• A loopback address is a special IP address used by a device to test its own network stack —
that is, to verify that TCP/IP is properly installed and functioning on the local host.
• It allows a computer to send and receive network traffic to itself, without using any physical
network interface or external device.
• The IPv4 loopback address ranges from [Link] to [Link] and has a default
subnet mask of [Link] (/8). However, it’s quite common for network professionals to
identify [Link] as the loopback address as it’s the first IPv4 address within the range, and
it’s mostly used when testing the loopback connectivity on a device.
 • In the IPv6 space, the loopback address is ::1/128, which allows network professionals to
perform the same loopback testing on the TCP/IP network model on the local device.

5.9. Unique local address

• A unique Local address exists within the IPv6 addressing space and ranges from FC00::/7 to
FDFF::/7.
• These unique local addresses have similarities to private IPv4 addresses on a network; they
are unique to a private network within an organization and are non-routable on the internet.
• These unique local addresses are used for local addressing only and can be assigned to
devices that do not need access to another network.

5.10. Default gateway

• A default gateway is the network device (usually a router) that acts as an access point
or exit for a host to communicate with devices outside its local network (subnet).
• If a device tries to send packets to an IP address not in its own network, the packet is
forwarded to the default gateway, which determines where it should go next.
Primary Function:
• Routes traffic from the local network to remote networks
• Acts as the forwarding point for all non-local destination traffic
• Essential for internet access and inter-network communication

Key Characteristics
Network Boundary Device:
• Resides on the same IP subnet as the local devices
• Has interfaces connected to multiple networks
• Makes routing decisions based on destination IP addresses
Mandatory for External Communication:
• Required for accessing resources outside the local subnet
• Without a default gateway, devices can only communicate locally
• Critical for internet connectivity

Figure 21: Default gateway address of a Windows 10

client
 Figure 22: Routing table of a Windows device

Figure 23: Routing table of a Linux-based device

6. Delving into IPv6 Concepts

• IPv4 and IPv6 addresses exist within different spaces and cannot natively communicate with
each other.
• To help ensure devices that exist on both IPv4 and IPv6 networks can communicate with
each other, various IPv6 technologies allow both versions of IP to coexist.
6.1. Tunneling
• IPv4 is still the dominant protocol across most of the internet.
• However, IPv6 public networks are increasingly being implemented worldwide.
• The transition phase between IPv4 and IPv6 creates compatibility issues, as both protocols
cannot communicate directly.
• Tunneling provides a bridge between IPv4 and IPv6 networks.
• It works by encapsulating packets of one protocol version inside packets of another
version, allowing them to traverse incompatible networks.
• When an IPv6 packet needs to travel across an IPv4 network, it is encapsulated inside an
IPv4 packet.
• The forwarding router performs this encapsulation.
 • When the encapsulated packet reaches the destination network, the IPv4 header is removed
(decapsulated), and the original IPv6 packet is delivered to its IPv6 destination.
• Tunneling enables coexistence of IPv4 and IPv6 during the transition phase.
• 6to4 tunneling is the most common technique during IPv6 deployment.
• 4to6 tunneling works in reverse, enabling IPv4 traffic to pass through IPv6 environments.
6.2. Dual stack
• As the world transitions from IPv4 to IPv6, not all networks or devices have been upgraded
to support the new protocol.
• To ensure smooth communication during this migration period, many systems operate in
dual-stack mode, allowing both IPv4 and IPv6 to coexist on the same device or network.
• Dual stack is a transition mechanism that allows a device, such as a computer, router, or
server, to run both IPv4 and IPv6 protocols simultaneously.
• Each device interface is configured with both an IPv4 address and an IPv6 address.
• The device maintains two protocol stacks — one for IPv4 and one for IPv6.

Figure 24: Dual stack NIC

6.3. Translation
• In IPv4 networks, Network Address Translation (NAT) is commonly used to convert private
IPv4 addresses into public IPv4 addresses before packets are sent out to the internet.
• As networks transition to IPv6, a similar mechanism is needed to enable communication
between IPv6 and IPv4 systems that cannot talk directly.
• NAT64 (Network Address Translation 64) is a translation mechanism that allows IPv6-
only devices to communicate with IPv4-only hosts.
• It works by translating IPv6 addresses into IPv4 addresses — and vice versa — similar to
how traditional NAT converts private IPv4 addresses into public IPv4 addresses.
6.3. Router advertisement
• In IPv6 networks, devices can automatically configure their own addresses and obtain
essential network information without manual configuration or a traditional DHCP server.
 • This automatic configuration is made possible through ICMPv6 (Internet Control
Message Protocol for IPv6) messages — specifically Router Solicitation (RS) and Router
Advertisement (RA) messages.

Router Solicitation (RS) and Router Advertisement (RA)

1. When an IPv6-enabled host (like a computer) connects to a network, it first sends a Router
Solicitation (RS) message.
• Purpose: To discover if there are any IPv6 routers on the local link.
2. The IPv6 router replies with a Router Advertisement (RA) message.
• Purpose: To provide the host with configuration information needed to obtain a
Global Unicast Address (GUA).

What Is a Router Advertisement (RA)?

• A Router Advertisement (RA) message is an ICMPv6 message sent periodically or in
response to an RS message.
• It informs IPv6 hosts about:
• How to obtain an IPv6 address, and
• Important network configuration parameters.
6.4. Stateless Address Autoconfiguration (SLAAC)
• SLAAC allows devices on a network to be configured with a GUA IPv6 address without the
need for a Dynamic Host Configuration Protocol v6 (DHCPv6) server.
Process of Obtaining an IPv6 GUA Address:
1. Router Solicitation (RS) –
The client sends an RS message to discover any IPv6 routers on the network.
2. Router Advertisement (RA) –
The router responds with an RA message, providing the network prefix and prefix length
(for example, 2001:db8:acad::/64).

3. Address Creation Using EUI-64 –

• The client takes its 48-bit MAC address from its network interface card (NIC).
• It uses the EUI-64 process to convert it into a 64-bit Interface ID.
• The Interface ID is then appended to the 64-bit network prefix, forming a complete
128-bit IPv6 address (GUA).
4. Address Creation Using SLAAC with stateless DHCPv6 –
• First, use SLAAC to create its own GUA IPv6 address.
• Second, use the router’s IPv6 Link-Local address as the default gateway for the
network.
• Lastly, use the stateless DHCPv6 server to obtain the DNS server addresses.
5. Address Creation Using Stateful DHCPv6 –
• A stateful DHCPv6 server has similar functionalities to a traditional DHCP server on
an Ipv4 network as it provides the following configurations to clients: GUA IPv6
address, Prefix length, DNS server addresses.