The explosive growth of the internet — from a research network connecting a few thousand hosts to a global infrastructure serving billions of devices — exposed fundamental limitations in Internet Protocol version 4 (IPv4). Its 32-bit address space, once considered vast, has been fully allocated. Internet Protocol version 6 (IPv6) was designed as the long-term successor to IPv4, expanding the address space to 128 bits and introducing architectural improvements in header design, auto-configuration, and built-in security support .

This lesson explores the motivations behind IPv6, its hexadecimal notation and compression rules, the major address types (unicast, multicast, and anycast), a direct comparison with IPv4, and guidance on selecting the appropriate IPv6 address format for different communication scenarios.

Footnotes

1. IPv4 Address Exhaustion - Wikipedia - Documents the depletion of the global IPv4 address pool, IANA allocation timeline, and RIR exhaustion dates. ↩

Why IPv6? The Limitations of IPv4

IPv4 was standardized in 1981 (RFC 791) with a 32-bit address space yielding $2^{32} = 4{,}294{,}967{,}296$ total addresses. Several factors drove this pool to exhaustion 2:

1. Address Exhaustion: The Internet Assigned Numbers Authority (IANA) allocated its last blocks of IPv4 addresses to Regional Internet Registries (RIRs) in February 2011. All five RIRs have since reached exhaustion of their free pools. With billions of smartphones, IoT sensors, and cloud instances joining the internet, 4.3 billion addresses are fundamentally insufficient .

2. NAT Complexity: Network Address Translation (NAT) was introduced as a workaround for address scarcity. While NAT extended IPv4's lifespan, it breaks the end-to-end connectivity principle of the internet, complicates peer-to-peer applications, VoIP, and online gaming, and adds processing overhead on every packet traversing the NAT device.

3. Inefficient Header Processing: The IPv4 header contains a variable-length options field and a header checksum that must be recalculated at every router hop, adding latency. The fragmentation fields in the IPv4 header further complicate router processing.

4. Limited Security and Auto-configuration: IPv4 was not designed with built-in security (IPsec is optional) or stateless address auto-configuration. Features like DHCP had to be layered on separately.

IPv6 addresses all of these issues with a fundamentally redesigned protocol.

Footnotes

1. IPv4 Address Exhaustion - Wikipedia - Documents the depletion of the global IPv4 address pool, IANA allocation timeline, and RIR exhaustion dates. ↩ ↩²

2. IPv4 Exhaustion Explained: Causes, Impacts, and Solutions - ServerMania - Comprehensive overview of IPv4 exhaustion, NAT workarounds, and the push toward IPv6 adoption. ↩

IPv6 Address Structure and Notation

An IPv6 address is 128 bits long, written as eight groups of four hexadecimal digits separated by colons. Each group is called a hextet (also known as a "quartet" or "hex group") .

Full notation example:

$2001\text{:}0db8\text{:}0000\text{:}0000\text{:}0000\text{:}00ff\text{:}fe00\text{:}0042$

This address is unwieldy, so two compression rules make it more manageable:

Rule 1: Omit Leading Zeros

Within any hextet, leading zeros can be dropped. At least one digit must remain.

$2001\text{:}0db8\text{:}0000\text{:}0000\text{:}0000\text{:}00ff\text{:}fe00\text{:}0042$ $\downarrow$ $2001\text{:}db8\text{:}0\text{:}0\text{:}0\text{:}ff\text{:}fe00\text{:}42$

Rule 2: Replace Consecutive All-Zero Hextets with ::

A single contiguous sequence of one or more all-zero hextets can be replaced by a double colon (::) — but this can only be used once in any address, to avoid ambiguity.

$2001\text{:}db8\text{:}0\text{:}0\text{:}0\text{:}ff\text{:}fe00\text{:}42$ $\downarrow$ $2001\text{:}db8\text{::}ff\text{:}fe00\text{:}42$

These rules shorten only the representation; the underlying address is always 128 bits .

Footnotes

1. IPv6 Address Representation - NetworkAcademy.IO - Detailed explanation of IPv6 hexadecimal notation, the two compression rules (leading zero omission and double colon), and expansion procedures. ↩ ↩²

IPv6 Address Types

Unlike IPv4, which uses unicast, broadcast, and multicast, IPv6 eliminates broadcast entirely and introduces anycast as a first-class address type. The three primary types of IPv6 addresses are 2:

1. Unicast — One-to-One

A packet sent to a unicast address is delivered to exactly one interface. IPv6 defines several unicast scopes:

2. Multicast — One-to-Many

A packet sent to a multicast address is delivered to all interfaces that have joined that multicast group. IPv6 multicast addresses begin with ff00::/8. Notable multicast addresses include:

• ff02::1 — All nodes on the local link
• ff02::2 — All routers on the local link
• ff02::1:ff00:0/104 — Solicited-node multicast (used in Neighbor Discovery)

IPv6 uses multicast where IPv4 used broadcast (e.g., ARP is replaced by Neighbor Solicitation using multicast).

3. Anycast — One-to-Nearest

A packet sent to an anycast address is delivered to the nearest interface (by routing distance) configured with that address. Anycast is commonly used for load-balanced DNS servers and content delivery network (CDN) nodes.

Footnotes

1. IPv6 Address Types - IPCisco - Covers all IPv6 address types including global unicast, link-local, unique local, multicast, anycast, and special addresses with prefix details. ↩

2. FAQ on IPv6 Adoption and IPv4 Exhaustion - Internet Society - Internet Society's authoritative FAQ on IPv6 deployment status, transition strategies, and the relationship between IPv4 exhaustion and IPv6 adoption. ↩

IPv4 vs. IPv6: A Comprehensive Comparison

The differences between IPv4 and IPv6 extend far beyond address length. IPv6 was a ground-up redesign that addressed performance, security, and configurability concerns 2:

Footnotes

1. IPv4 Exhaustion Explained: Causes, Impacts, and Solutions - ServerMania - Comprehensive overview of IPv4 exhaustion, NAT workarounds, and the push toward IPv6 adoption. ↩

2. Navigating the IPv4 to IPv6 Transition - IR - Enterprise guide on IPv6 transition strategies including dual stack, tunneling, and translation mechanisms, with IPv4 vs IPv6 header and design comparisons. ↩

IPv6 Header Simplification

One of IPv6's most significant engineering improvements is its streamlined, fixed-length header. The IPv4 header has 12 mandatory fields plus a variable-length options field; the IPv6 base header has only 8 fields in a fixed 40-byte structure :

Key improvements:

• No header checksum: Eliminates the need for recalculation at every router hop, reducing processing overhead.
• No fragmentation fields in the base header: Fragmentation is handled end-to-end using extension headers, so intermediate routers never fragment packets.
• Extension headers: Optional functionality (like fragmentation, routing, authentication) is placed in chained extension headers referenced via the Next Header field, keeping the base header compact and fast to process.
• Flow Label: A new 20-bit field allows routers to identify and provide special handling for specific traffic flows (e.g., real-time video) without deep packet inspection.

Footnotes

1. Navigating the IPv4 to IPv6 Transition - IR - Enterprise guide on IPv6 transition strategies including dual stack, tunneling, and translation mechanisms, with IPv4 vs IPv6 header and design comparisons. ↩