Explain packet switching, the structure of a packet, how routers forward packets, and contrast it with circuit switching

What this dot point is asking

SEAB wants you to explain packet switching, the structure of a packet, how routers forward packets, and how this differs from circuit switching. The central idea is that the internet splits messages into independently routed packets that share network links, rather than reserving a dedicated path - making the network efficient and resilient.

The answer

Packet switching

In packet switching, a message or file is split into many small packets. Each packet is routed independently across the network and may take a different path. At the destination, packets are reassembled in the correct order. Links are shared between many communications - capacity is used only when packets are actually sent (statistical multiplexing).

The structure of a packet

A packet has a header of control information and a payload of the actual data. The header typically carries:

• Source IP address and destination IP address (for routing).
• A sequence number (so the receiver can reassemble packets in order).
• A checksum or error-check (to detect corruption).
• A time-to-live (TTL) / hop limit (discarding packets that loop).

[ header: src IP | dest IP | seq no | checksum | TTL ] [ payload: data ]

How routers forward packets

A router connects networks. At each hop, it:

1. reads the packet's destination IP address,
2. looks it up in its routing table to choose the best next hop toward the destination,
3. forwards the packet on the chosen outgoing link.

It does this independently for every packet, so the network can route around congestion or a failed link by choosing different paths.

Reassembly and out-of-order arrival

Because packets travel independently, they can arrive out of order, be delayed differently, or be lost. The transport layer (TCP) fixes this at the destination: sequence numbers let it reorder packets, and missing packets are detected and retransmitted.

Contrast with circuit switching

Circuit switching reserves a dedicated end-to-end path for the whole communication (a traditional phone call), held even when idle. Packet switching reserves nothing, sharing links among many flows.

• Packet switching is efficient (shared capacity, no idle reservation) and resilient (reroutes around failures).
• Circuit switching gives a guaranteed constant path but wastes capacity when idle and fails entirely if the path breaks.

Examples in context

Example 1. Resilience of the internet. If a major link fails, routers detect it and forward subsequent packets along alternative paths, so a download continues with only a brief disruption. This rerouting is possible precisely because packet switching reserves no fixed path, unlike a circuit-switched call that would simply drop.

Example 2. Sharing a home connection. Several family members stream, browse and message at once over one broadband line because their packets are interleaved on the shared link, each using capacity only when sending. Circuit switching would force each to reserve a slice of the line whether or not they were using it, wasting capacity.

Try this

Q1. Give three pieces of information carried in a packet header. [3 marks]

• Cue. Source IP address, destination IP address, sequence number (also acceptable: checksum, time-to-live, protocol).

Q2. Why might packets of one message arrive out of order? [1 mark]

• Cue. Each packet is routed independently and may take a different path with a different delay.

Q3. State one advantage of packet switching over circuit switching for the internet. [1 mark]

• Cue. Efficient shared use of capacity (no idle reserved path) and resilience through rerouting around failures.