Internet ProtocolEdit

The Internet Protocol (IP) is the foundational addressing and routing protocol of the Internet. It operates at the network layer of the Internet protocol suite, providing a scalable mechanism to move packets from a source host to a destination host across diverse networks. IP defines how data should be addressed in packets, how those packets are routed through intermediate routers, and how to recognize and handle delivery failures. The two predominant versions in use today are IPv4, which uses 32-bit addresses, and IPv6, which uses 128-bit addresses. IP works in concert with transport protocols such as TCP and UDP, and with other upper-layer protocols that rely on IP for end-to-end communication. In modern networks, IP is embedded in everything from home broadband to global data centers and submarine cable systems.

IP is characterized by a few enduring design principles. It is connectionless and best-effort, meaning that each packet is treated independently and delivery is not guaranteed. The network is expected to be robust in the face of variable paths and delays, with higher-layer protocols responsible for reliability when needed. IP addresses serve as locators for devices on a network, and routing decisions are made by routers that forward packets toward their destination based on these addresses. As the Internet grew, IP evolved to accommodate more devices and more traffic, culminating in the broad adoption of IPv6 to address long-term scalability concerns. For an overview of the architectural context, see Internet Protocol and related discussions within the IETF and the RFC series.

History

The development of IP was part of the broader early evolution of the Internet, beginning in the research networks of the 1960s and 1970s. The core ideas were crystallized in the work of Vint Cerf and Robert Kahn, who coordinated the Transmission Control Protocol and Internet Protocol (TCP/IP) design. The transformation from experimental networks to a widely deployed standard occurred in the early 1980s when TCP/IP became the standard protocol suite for ARPANET and other connected networks. Key milestones include the publication of IPv4 as the main addressing scheme, formalized in RFC 791 (1981), and the subsequent standardization and expansion of the protocol suite to support a growing Internet.

As address scarcity became a long-term concern, the development of IPv6 began in the 1990s and matured through the work of the IETF and related bodies. IPv6 was designed to provide a vastly larger address space, better autoconfiguration, and improvements in routing efficiency and security features. The formal IPv6 specification was published in RFC 8200 (2008), with supporting documents that describe IPv6 header formats, addressing architecture, and transition mechanisms. Throughout this history, governance of IP addressing has involved organizations such as the IANA and regional Internet registries like ARIN, RIPE NCC, and APNIC.

Technical foundations

IP defines a datagram service for delivering packets across networks. Each IP datagram contains a header with routing and addressing information, followed by a payload that carries higher-level protocol data. Two essential differences distinguish the main versions:

IPv4 uses 32-bit addresses, allowing roughly 4.3 billion distinct addresses. It employs a variety of techniques to extend reach and utilization, including private address spaces defined in RFC 1918 and widespread use of NAT to conserve the public address pool.
IPv6 uses 128-bit addresses, yielding an astronomically larger address space and enabling more scalable routing and configuration options, such as SLAAC (Stateful/Stateless Address Autoconfiguration) and improved security considerations.

Key concepts and components include: - Datagram-based delivery: IP handles individual packets without guaranteed delivery; reliability is provided by higher layers (e.g., TCP) or by error-correcting features when implemented. - Addressing: IP packets carry source and destination addresses to identify endpoints in the network. - Fragmentation: IP can fragment large packets to fit the next-hop MTU, though modern practice often relies on endpoints to perform sending or path MTU discovery to minimize fragmentation. - Routing: Routers forward IP datagrams using routing tables and path information. Global routing decisions are coordinated by interdomain protocols such as the Border Gateway Protocol BGP. - End-to-end principle: IP supports a simple, generalized approach to delivering data across heterogeneous networks, while higher-layer protocols provide application-specific semantics.

For a deeper look at the packet structure and behavior, see discussions of the IPv4 header and the IPv6 header, as well as the broader TCP/IP model.

Addressing schemes and evolution

IP addressing is central to how networks locate devices. IPv4 relies on 32-bit addresses, while IPv6 uses 128-bit addresses to support growth, mobility, and improved multicast and autoconfiguration features. The address spaces reflect a shift in Internet architecture: IPv4 networks often rely on NAT to extend the usability of a limited address pool, whereas IPv6 aims to restore end-to-end reachability by eliminating the need for NAT as the primary addressing mechanism.

IPv4 addressing: Public and private address ranges; subnetting through hierarchical prefixes; address exhaustion prompted migration strategies and deployment of NAT-based techniques.
IPv6 addressing: Global unicast addresses, link-local addresses, unique local addresses, and multicast capabilities. Features such as SLAAC and DHCPv6 support flexible host configuration in large-scale networks.

Routing plays a critical role in IP operations. IPv4 and IPv6 routing rely on routers that implement the global Internet routing fabric, with protocols like BGP guiding inter-domain decisions and local routing protocols handling intra-domain paths. The ongoing transition to IPv6 involves dual-stack deployments, tunneling mechanisms, and the gradual phasing out of legacy IPv4-only configurations in many networks.

Deployment and usage

IP is deployed across virtually all networks that interconnect computers and devices. In practice, most networks today run some combination of IPv4 and IPv6, often in a dual-stack configuration to ensure compatibility with older systems while enabling modern features. The global Internet uses a layered approach where IP handles addressing and routing, while higher-layer protocols and applications manage the semantics of data exchange.

Private addressing and NAT: To mitigate IPv4 address shortages, many networks deploy private address spaces and employ NAT at the network edge. While NAT can simplify address management and security for some deployments, it also alters end-to-end connectivity and adds complexity to troubleshooting.
IPv6 adoption: Regions differ in the pace of IPv6 deployment. Where IPv6 is widely deployed, hosts can obtain globally routable addresses without the same level of NAT reliance, potentially improving direct reachability and simplifying some network architectures.
Transition mechanisms: Several mechanisms exist to facilitate the transition from IPv4 to IPv6, including dual-stack operation, tunneling methods (e.g., some historical approaches like 6to4), and translation techniques that allow IPv6-only hosts to communicate with IPv4-only peers.

IP-based routing is fundamental to how content reaches users through ISPs, data centers, and content delivery networks. The interoperation of IP with the Domain Name System (DNS)—which resolves human-readable names to IP addresses—is essential for the usability of the Internet. Networks also implement security and resilience measures at the IP layer, such as IPsec in certain configurations, and rely on tools like ICMP for diagnostic and error-reporting messages.

Security, privacy, and governance

Security considerations at the IP layer include the optional use of IPsec to provide confidentiality, integrity, and authentication for IP traffic, as well as defensive measures against spoofing and certain forms of denial-of-service (DDoS) attacks. Because IP addresses can reveal a device’s location and network topology, privacy concerns arise, motivating the use of privacy-preserving techniques such as Virtual Private Networks (VPNs) and, in some contexts, traffic anonymity tools.

Governance of IP addressing and the global routing infrastructure involves standard-setting organizations and regional registries. The IETF develops the technical standards and best practices that guide IP evolution, while regional Internet registries manage address allocation and assignment within their respective regions. The interplay between technical innovation, business interests, and regulatory considerations shapes deployment choices and the pace of change in Internet infrastructure.