TCP/IP Model

The TCP/IP model is a set of rules that describes how data moves from one device to another across a network — including across the entire internet. It takes the huge job of "getting data from here to there" and splits it into four smaller, well-organized steps called layers, where each layer handles one part of the journey and then hands the rest to its neighbour. Before any two devices can talk, they must agree on the same protocol — a shared set of rules for formatting and exchanging data — and the TCP/IP model is the blueprint that ties all of those rules together.

Two protocols do the heavy lifting, which is why the model carries their names. IP (Internet Protocol) is in charge of addressing and delivery: it gives every device an address and works out how to get a small chunk of data — called a packet — to the right destination. TCP (Transmission Control Protocol) sits on top of IP and makes that delivery reliable: it checks that every piece arrived, puts the pieces back in the right order, and asks for anything that went missing. Between them they answer the two big questions: "where does this go?" and "did all of it get there correctly?"

The TCP/IP model grew out of ARPANET, an experimental network funded by the U.S. Department of Defense in the 1970s. Two researchers, Vint Cerf and Bob Kahn, designed the core TCP/IP protocols so that completely different kinds of networks could connect to one another and keep working even if parts of them failed. On 1 January 1983, ARPANET officially switched over to TCP/IP — a date often treated as the birth of the modern internet. Because the protocols were built and proven in the real world first, TCP/IP is a practical, battle-tested model rather than a purely theoretical one.

TCP/IP stacks its four layers one on top of another. When you send something, your data travels down the stack from the top; when you receive something, it climbs back up. Here is the whole stack, from the layer closest to you down to the physical connection:

Each layer trusts the one beneath it to handle its part of the job. The application layer doesn't care how packets are routed; the internet layer doesn't care whether the link is copper cable, fibre, or Wi-Fi. The label on the right of each tier ("Data", "Segments", "Packets", "Frames") is the name the data goes by at that layer — we'll unpack all of this in the next lesson.

Breaking the work into layers isn't just tidy — it's what makes a network the size of the internet possible to build and maintain. Each layer can be designed, understood, and upgraded on its own, as long as it keeps the same simple agreement with its neighbours about what it takes in and hands out. Switching your connection from a cable to Wi-Fi changes only the bottom layer; every app you use keeps working, untouched. This is the same idea behind the older OSI model, which uses seven layers instead of four.

Almost everything you do online rides on TCP/IP: loading a page, sending a message, streaming a video, or joining a call. Understanding its four layers gives you a mental map for the whole of networking — where addresses live, where reliability comes from, and where to look when something breaks. It's also one of the most common networking interview topics, so it's worth knowing cold.