Internet Protocol (IP)

Core characteristics and packet structure

IP treats each packet independently: packets may follow different routes, arrive out of order, be duplicated, or be lost. Basic capabilities commonly associated with IP include addressing, routing, fragmentation, and the use of a time‑to‑live (TTL) mechanism to avoid indefinite circulation. IP headers carry essential control information used by routers and hosts for delivery decisions. In IPv4, the header includes fields such as version, header length, total length, source and destination addresses, protocol identifier, TTL and a checksum; IPv6 simplified and extended the header format and expanded address size, removing the header checksum.

Addressing: IPv4 and IPv6

IP addresses identify network interfaces rather than physical devices. Two major versions are in widespread use: IPv4, which uses 32‑bit addresses, and IPv6, which uses 128‑bit addresses to accommodate the vastly larger address space required by modern networking. Address allocation, subnetting, and address notation differ between the two versions, and mechanisms exist to translate or tunnel traffic between IPv4 and IPv6 when necessary.

History and evolution

The design of IP emerged from early packet‑switching research and ARPANET work in the 1970s, and it became formalized as part of the Internet's architecture. The IPv4 specifications were widely standardized in early documents and became the backbone of the growing Internet. As demand for addresses and new features increased, a successor protocol, IPv6, was developed and standardized to provide a much larger address space and other improvements. IP's development has been guided by engineering needs and incremental refinements rather than a single central authority.

How IP is used in practice

In operation, an application hands data to a transport protocol (for example, TCP or UDP), which places it into a transport segment and passes it to IP. IP encapsulates the segment inside a packet and relies on lower‑layer technologies (Ethernet, Wi‑Fi, cellular links) to carry that packet between neighboring nodes. Routers examine the destination address and forward packets toward their destination using routing tables populated by static configuration or dynamic routing protocols. Diagnostic and control messages are handled by companion protocols such as ICMP; address resolution to map IP addresses to hardware addresses is typically handled at the link layer.

Notable distinctions, limitations and related protocols

IP is intentionally minimal and does not guarantee reliability, ordering, or congestion control — those functions are performed by transport layer protocols like TCP. Because it is a best‑effort, packet‑based service, network architects must combine IP with other mechanisms to achieve reliability, quality of service, or security. Various tunneling, translation and extension techniques exist to adapt IP to different network environments, and numerous protocols interact with IP for addressing, error reporting, and management. For broader context and protocol relationships see the Internet architecture and packet switching concepts.

• Connection model: connectionless, per‑packet delivery.
• Addressing: IPv4 (32‑bit) and IPv6 (128‑bit).
• Guarantees: best‑effort delivery; reliability provided by higher layers.
• Common uses: routing across LANs and WANs, backbone interconnection, and service delivery for applications.

For detailed formal definitions, implementations and protocol behavior, consult technical standards and protocol descriptions in the relevant engineering documents and reference materials. Further reading and resources are available through standards organizations and networking texts that cover the Internet protocol suite and modern operational practices.