The Two Protocols Doing Different Jobs
The name TCP/IP is shorthand for two distinct protocols that work at different levels of the network stack. IP (Internet Protocol) operates at the internet layer and handles one job: getting a packet from a source address to a destination address, possibly across many intermediate networks. It does this without any guarantee of success — IP is a best-effort protocol that forwards packets and hopes they arrive.
TCP (Transmission Control Protocol) operates at the transport layer, sitting above IP. Its job is to take IP's unreliable delivery service and turn it into something applications can depend on: an ordered, error-checked, reliable stream of bytes. Together they form the foundation of nearly every internet application you use.
The term "TCP/IP" has expanded over the decades to refer not just to these two protocols but to the entire suite of internet protocols — including UDP, ICMP, DNS, HTTP, and dozens more. But IP and TCP remain the core.
How IP Addressing Works
Every device on the internet needs a unique address so packets can be delivered to the right destination. IP provides this through IP addresses. In IPv4, addresses are 32 bits long, written as four decimal numbers separated by dots — for example, 192.168.1.1. IPv6 uses 128-bit addresses to support a vastly larger address space.
When a packet leaves your device, your router reads the destination IP address in the packet's IP header and forwards it toward that destination. The next router does the same thing, and so on, hop by hop, until the packet arrives. No single router knows the full path — each one only knows the next best step. This is called hop-by-hop routing, and it is what makes the internet resilient: traffic can reroute around failures automatically.
IP also handles packet fragmentation. If a packet is too large for a particular network link, IP can split it into smaller fragments and reassemble them at the destination. The IP header contains all the information needed to identify fragments and put them back together.
How TCP Makes Delivery Reliable: The Three-Way Handshake
Before TCP sends any data, it establishes a connection using a three-step process called the three-way handshake. First, the client sends a SYN (synchronize) packet to the server, proposing a starting sequence number. The server responds with SYN-ACK, acknowledging the client's sequence number and proposing its own. Finally, the client sends an ACK confirming the server's sequence number. Both sides are now synchronized and ready to exchange data.
Once connected, TCP assigns a sequence number to every byte it sends. The receiver sends acknowledgments (ACK) telling the sender which bytes it has received. If the sender does not receive an ACK within a timeout window, it retransmits. This mechanism ensures that no data is silently lost — if a packet drops, TCP detects it and sends it again.
TCP also handles congestion. It monitors the network and adjusts how fast it sends data to avoid overwhelming routers. This is called congestion control, and it is why a TCP download automatically slows down when the network is busy rather than flooding the path with packets that would just be dropped anyway.
The Four TCP/IP Layers
The TCP/IP model organizes networking into four layers, each with a distinct responsibility:
| Layer | What It Does | Protocols at This Layer |
|---|---|---|
| Application | Defines the format and rules for application data exchange | HTTP, DNS, SMTP, SSH, FTP |
| Transport | Provides end-to-end communication between processes on two hosts | TCP, UDP, QUIC |
| Internet | Addresses and routes packets across networks | IP (IPv4/IPv6), ICMP |
| Link | Transmits frames over a single physical network segment | Ethernet, Wi-Fi, ARP |
Each layer communicates only with the layers directly above and below it. This strict separation is what allows the internet to evolve — HTTP/3 can replace TCP with QUIC at the transport layer without any changes to IP, and IP can run over new physical media without any changes to HTTP.
What Actually Happens When You Load a Web Page
A page load is TCP/IP working across all four layers at once. When you type a URL and press enter, your browser first needs the server's IP address, so it sends a DNS query (over UDP) to your configured DNS resolver. Once it has the IP address, it opens a TCP connection using the three-way handshake. If the site uses HTTPS, a TLS handshake follows immediately on top of the TCP connection.
With the connection established, your browser sends an HTTP GET request. That request is handed to TCP, which breaks it into segments, numbers them, and passes them to IP. IP adds source and destination addresses and passes each packet to your network interface. Your router forwards the packets toward the server, each packet making a series of hops across different networks.
The server's TCP stack receives the segments, reassembles them in order, and hands the HTTP request up to the web server application. The server sends back an HTTP response the same way. Your browser receives the response, reassembles it, and renders the page. The entire process — from keystroke to page render — typically takes between 50 and 500 milliseconds for a modern site.
Frequently Asked Questions
What does TCP stand for?
TCP stands for Transmission Control Protocol. It is the transport-layer protocol responsible for ensuring reliable, ordered delivery of data between two endpoints. It establishes a connection before sending data and retransmits anything that gets lost in transit.
What does IP stand for?
IP stands for Internet Protocol. It operates at the internet layer and is responsible for addressing and routing packets across networks. Every device on the internet has an IP address, and IP is what enables a packet to travel from one network to another until it reaches its destination.
Is UDP part of TCP/IP?
Yes. UDP (User Datagram Protocol) is part of the TCP/IP suite and operates at the same transport layer as TCP. TCP/IP refers to the entire suite of internet protocols, not just TCP and IP themselves. UDP provides a connectionless, lightweight alternative to TCP for applications that prioritize speed over reliability.
What is the TCP three-way handshake?
The TCP three-way handshake is how TCP establishes a connection before data is exchanged. The client sends a SYN packet, the server responds with SYN-ACK, and the client completes the handshake with an ACK. After these three messages, both sides have agreed on sequence numbers and the connection is ready for data.
Can TCP/IP be replaced?
TCP/IP is deeply embedded in the internet's infrastructure, making a full replacement impractical. However, individual pieces can be and are being replaced. QUIC, for example, replaces TCP for HTTP/3 traffic while still running over IP. The internet layer itself is transitioning from IPv4 to IPv6 — a process still underway decades after IPv6 was standardized.
What is the difference between TCP/IP and the OSI model?
TCP/IP is the actual protocol suite running on the internet, organized into four layers: link, internet, transport, and application. The OSI model is a theoretical seven-layer reference framework. TCP/IP predates OSI and does not map cleanly onto it — the OSI application, presentation, and session layers all collapse into the TCP/IP application layer.