Data Transmission
Sending "hello" across the internet involves a lot more than it looks like: your message gets chopped up, wrapped in several layers of addressing, routed independently piece by piece, and reassembled on arrival, possibly in a completely different order than it was sent. This page makes every one of those steps visible.
Packet switching: proof it arrives out of order
The message "HELLO NET" is about to be split into 5 packets and sent from A to B across this small network. Each packet is free to take a different route, so it can arrive at a different time. Step through and watch the arrival log, then watch reassembly put it back in the right order regardless.
Controls
Arrival log (at B)
Exam tips
- Every packet carries its own header containing the destination address and a sequence number, so it can find its own way and be put back in order regardless of what order it arrives in.
- Packets from the same message can genuinely take different physical routes, because each router makes its own independent decision about the best next hop at that moment.
- This is precisely why packet switching is more resilient than a single fixed connection (circuit switching): if one route is congested or fails, packets simply get sent another way.
IP addressing
An IPv4 address is just 4 bytes, each 0-255, the same binary work from earlier, now used as a real address. Build one below and check whether it's public or private.
Try an address
Exam tips
- Private addresses (like 192.168.x.x) only work within a local network, they're not reachable from the wider internet directly.
- IPv4 has roughly 4.3 billion possible addresses, which sounds like a lot until you realise every phone, laptop, and smart device needs one, this is exactly why IPv6 exists.
DNS: turning names into numbers
You never actually type an IP address to visit a website. DNS is the system that looks one up for you, and it's genuinely hierarchical, like asking increasingly specific people for directions.
Controls
The TCP/IP model: wrapping data in layers
Before your message ever reaches a cable, it gets wrapped in three separate headers, one per layer, each adding information that a different piece of equipment along the way needs. This is called encapsulation, and it happens in reverse (decapsulation) when the message arrives.
Controls
| Layer | Adds | Example protocol |
|---|---|---|
| Application | The actual data / request | HTTP, FTP, SMTP |
| Transport | Port numbers (which app on the destination device) | TCP, UDP |
| Internet | IP addresses (which device, anywhere on the internet) | IP |
| Link | MAC addresses (which physical device on this local network) | Ethernet, WiFi |
Common protocols
A protocol is just an agreed set of rules for how two things communicate. Different jobs need different rules.
Bandwidth vs latency: a genuine race
Bandwidth is how much data can flow per second, once it's flowing. Latency is the delay before anything arrives at all. They're easy to confuse, but this race proves they're not the same thing: watch which connection actually wins change depending on the message size.
Message size
Exam tips
- High bandwidth does not mean low latency, a satellite link can move huge amounts of data per second, but every single bit still has to travel to space and back first.
- For tiny messages, latency dominates the total time. For huge transfers, bandwidth dominates. Neither number alone tells you which connection is actually faster.