IPv6 - Why

Cost

  • IPv4 addresses are rare, therefore expensive.

Energy

  • No NAT required
  • End-to-end communication allows peer to peer topologies reducing the need of intermediate systems
  • IP header checksum not calculated
  • No Broadcast (Multicast used instead)
    • Neighbor discovery protocol is using Link-local unicast and is more efficent than ARP with boradcast

Simplicity

Extensibility

  • Extentability with IP extention headers

Security

  • The sheere amount of IPs makes it hard to scan networks
    • No Brute-Forcing: Standard IPv4 ping sweeps are impossible on IPv6 because a standard subnet contains 2⁶⁴ (18 quintillion) addresses.
    • Discovery Methods: Scanners rely on multicast pings (ff02::1), neighbor discovery protocol (NDP) tables, DNS records (AAAA), and targeted address lists rather than sequential sweeps.

Deployment state

Rationale

IP header checksum

The header checksum was removed in IPv6 to increase packet processing speed in routers and improve overall network efficiency.

In IPv4, every router along a packet's path had to recalculate the header checksum because the "Time to Live" (TTL) field changes at each hop. This constant recalibration placed a significant load on router processors. Since modern transmission technologies at the lower layers of the OSI model (such as Ethernet or Wi-Fi) and the upper layers (such as TCP and UDP) already possess their own robust error detection mechanisms, the duplicate checking at the IP layer was deemed redundant.

Consequently, IPv6 designers adopted a principle of "trusted transport": erroneous packets are no longer discarded at the IP layer. Instead, errors affecting payload data are detected by end devices, while link-layer protocols handle transmission errors. Eliminating this checksum allows routers to forward packets significantly faster, enhancing performance particularly at high data rates and within the internet backbone.

Links