EnglandComputer ScienceSyllabus dot point

What hardware and protocols move a packet of data across a network, and what does each layer of the TCP/IP stack do?

Networks (LAN and WAN, network topologies, client-server and peer-to-peer), network hardware (NIC, switch, router, WAP), the need for protocols and protocol layering, the TCP/IP four-layer stack, and packet switching.

An OCR H446 answer on networks: LANs and WANs, topologies, client-server and peer-to-peer models, network hardware (NIC, switch, router, wireless access point), the need for protocols and layering, the TCP/IP four-layer stack, and packet switching.

Generated by Claude Opus 4.814 min answerUpdated 2026-06-02

Reviewed by: AI editorial process; not yet individually human-reviewed

Have a quick question? Jump to the Q&A page

Quick answer

A LAN connects devices over a small area (one site); a WAN connects devices over a large area (between sites, like the internet), usually over infrastructure a third party owns. Topologies describe the layout: a star has each device connected to a central switch (resilient, the common choice); a mesh connects devices to many others (very resilient, costly). In a client-server network, dedicated servers provide resources and central control; in peer-to-peer, every device is equal and shares directly. Hardware includes the network interface card (NIC), switch (forwards frames within a LAN by MAC address), router (forwards packets between networks by IP address) and wireless access point. Protocols are agreed rules for communication, organised into layers so each can change independently. The TCP/IP stack has four layers: application, transport (TCP), internet (IP) and link. Packet switching splits data into packets that travel independently and are reassembled at the destination.

Jump to a section

What this dot point is asking
The answer
Examples in context
Try this

What this dot point is asking

OCR wants the types and topologies of networks, the client-server and peer-to-peer models, the main network hardware, why protocols and layering are needed, the four layers of the TCP/IP stack, and how packet switching works. Expect a "describe the TCP/IP layers" question and a "client-server versus peer-to-peer" discussion.

The answer

Network types, topologies and models

Key fact

A LAN (Local Area Network) connects devices across a small geographic area such as a building or campus, typically over hardware the owner controls. A WAN (Wide Area Network) connects devices over a large area, often between sites or countries, usually over third-party infrastructure; the internet is the largest WAN. A star topology connects every device to a central switch, so one cable fault isolates only one device and the network is easy to manage, but the central switch is a single point of failure. A mesh topology connects devices to many others so there are multiple paths, giving high resilience at high cabling cost (full wireless mesh is common in IoT). In a client-server network, powerful servers provide services and central control; in a peer-to-peer network, every device is an equal that both requests and provides resources, which is cheap for small networks but lacks central management.

Network hardware

Protocols, layering and the TCP/IP stack

Key fact

A protocol is an agreed set of rules governing communication so that different devices and software can interoperate (format, timing, error handling). Layering organises protocols into self-contained layers, each providing a service to the layer above and using the one below, so any layer can be modified or replaced independently, which makes networks modular and easier to develop. The TCP/IP four-layer stack is: the application layer (services for user programs, choosing protocols such as HTTP, FTP, SMTP), the transport layer (TCP splits data into numbered segments, manages reliable delivery, reassembly and ports), the internet/network layer (IP adds source and destination IP addresses and routes packets across networks), and the link/network-access layer (handles physical transmission on the local network using MAC addresses).

Trace data down and up the TCP/IP stack

A browser sends a request to a web server. Trace what each layer adds on the way out, and what happens on arrival.

step 1: Application layer (sending)

The browser uses HTTP to form the request message (for example "GET this page"). It passes the message down to the transport layer.

step 2: Transport layer (sending)

TCP splits the message into segments, numbers them so they can be reassembled in order, and adds port numbers (port 80 or 443 for web) to identify the application. Segments go to the internet layer.

step 3: Internet and link layers (sending)

The internet layer (IP) adds the source and destination IP addresses, forming packets, and decides the route. The link layer adds the MAC address of the next hop and transmits the bits physically.

step 4: Reassembly at the server

At the server the layers are stripped in reverse: the link layer receives the frames, the internet layer checks the IP address, the transport layer reorders the numbered segments and reassembles the message, requesting any missing ones, and the application layer hands the complete HTTP request to the web server software.

Examples in context

A home network uses a star topology around a router that also acts as a switch and wireless access point, connecting the LAN to the internet (a WAN). Streaming a video relies on packet switching, so a congested link can be routed around. The layered TCP/IP model is why you can switch from wired to wireless (changing the link layer) without changing the browser (the application layer). OCR links this to DNS and web technologies, and to network security, which adds protection at several layers.

Try this

Q1. State one advantage of a star topology over a bus or ring. [1 mark]

Cue. A single cable fault affects only one device, and it is easy to add devices or isolate faults (resilient and manageable).

Q2. State which layer of the TCP/IP stack adds IP addresses and routes packets. [1 mark]

Cue. The internet (network) layer.

Q3. Explain one benefit of organising network protocols into layers. [2 marks]

Cue. Each layer is self-contained, so it can be changed or replaced without affecting the others, making the system modular and easier to develop and maintain.

Exam-style practice questions

Practice questions written in the style of OCR exam questions on this dot point, with worked answer explainers. The year tag is the paper they imitate, not the source.

OCR 20196 marksExplain what is meant by protocol layering and describe the purpose of each layer of the TCP/IP stack.

Show worked answer →

Layering (up to 2): protocols are organised into self-contained layers, each providing a service to the layer above and using the layer below, so a layer can be changed or replaced without affecting the others; this makes the system modular and easier to develop and maintain.

The four layers (up to 4): the application layer provides services to user programs and chooses the protocol (HTTP, FTP, SMTP). The transport layer (TCP) splits data into segments, numbers them, and manages reliable delivery and reassembly using ports. The internet (network) layer (IP) adds source and destination IP addresses and routes packets across networks. The link (network access) layer handles the physical transmission on the local network using MAC addresses. Markers reward the modular benefit of layering plus the role of each named layer.

OCR 20216 marksA company is choosing between a client-server and a peer-to-peer network for an office of 60 computers. Discuss which is more appropriate.

Show worked answer →

Levels-of-response question; reward a developed comparison reaching a judgement.

Client-server: dedicated servers provide resources and central control, so security, backups, user accounts and software updates are managed centrally and consistently; performance scales for many users, but servers cost money and need expert administration, and a server failure affects everyone relying on it.

Peer-to-peer: every computer is equal, sharing resources directly with no central server, which is cheap and simple to set up for a few machines; but with 60 computers, security and backups are inconsistent, finding shared resources is harder, and performance suffers as peers serve each other.

For 60 computers in an office, central management of security, accounts and backups matters, so client-server is more appropriate despite its cost. Peer-to-peer suits only very small networks. Top marks need the central-management argument linked to the office size.

Related dot points

Sources & how we know this

OCR AS and A Level Computer Science (H046, H446) specification — OCR (2015)