A network is two or more computers (or devices) connected so they can share data and resources. Paper 1 asks you to describe network hardware, compare models and topologies, and explain how data finds its way across the internet.
2.1 Networks including the internet
Why network?
A network lets users share files, printers, internet connections and software, communicate (email, messaging, video calls), and manage things centrally: backups, updates and security can be controlled from one place. The drawbacks are the cost of the extra hardware, the risk of hacking and malware spreading, and that if a key part fails many users are affected.
LAN and WAN
| LAN (local area network) | WAN (wide area network) | |
|---|---|---|
| Area covered | A small area: one building or site | A large area: cities, countries or the world |
| Ownership | Usually owned and managed by one organisation | Often uses lines leased from telecommunications companies, or is public |
| Speed | Generally higher | Generally lower |
| Connections | Cables, switches, wireless access points | Routers, leased lines, satellite, public networks |
| Example | A school network | The internet |
Client-server and peer-to-peer
In the client-server model, one or more servers provide services (file storage, web pages, email, printing, security) and clients request them.
In the peer-to-peer (P2P) model, every computer is equal (a peer): each can both share its own files and use those of others, and there is no central server.
| Client-server | Peer-to-peer | |
|---|---|---|
| Control | Central control; easy to manage security, backups and user accounts | No central control; each user manages their own computer |
| Cost | Server hardware and a network administrator are expensive | Cheap: no special server needed |
| Reliability | If the server fails, clients lose the service (until it is restored) | Failure of one peer does not stop the others, but files on that peer are unavailable |
| Performance | The server can be overloaded, but is built for the job | Slows down when many peers share, but no single bottleneck |
| Security | Easier to secure | Harder: no central security |
| Best for | A school, hospital or company with many users and shared data | A small home network, or file sharing between a few friends |
Thin client and thick client
- A thick (fat) client does most of its own processing and stores its own software and data. It can still work if the network fails, but is more expensive and harder to maintain and secure.
- A thin client relies on the server to do most of the processing and storage. It is cheaper, easier to maintain and more secure (little data is held on it), but it cannot work without a network and the server, and the server needs to be powerful.
Topologies
A topology is the layout of the connections in a network.
- Bus: all devices share a single backbone cable. Cheap and easy to set up, but one break in the backbone can stop the whole network, performance falls as traffic rises (collisions), and data can be seen by every device.
- Star: every device connects to a central switch (or hub). A fault in one cable affects only that device, and it is easy to add devices and to find faults, but if the central switch fails the whole network fails, and it needs more cable.
- Mesh: devices are connected to several others (a full mesh connects every device to every other). Data can take many routes, so it is very reliable and there are few collisions, but it needs a lot of cabling and is expensive and complex. A partial mesh connects only some devices to more than one other. The internet and wireless mesh networks use this idea.
- Hybrid: a combination of two or more topologies, for example several star networks joined together by a bus.
Cloud computing
Cloud computing means storing data and running applications on remote servers accessed over the internet, instead of on a local device.
- A public cloud is owned by a provider and shared by many customers (for example online storage services).
- A private cloud is used by a single organisation, either hosted by itself or by a provider, so it has more control and security.
Benefits: access from any device and place; no need to buy or maintain servers; storage and processing can be increased or decreased as needed (scalable); the provider handles backups and updates. Drawbacks: it needs an internet connection; there is a security and privacy risk because data is on someone else's servers; the customer depends on the provider (and its prices); and a data centre failure can affect many customers.
Wired and wireless
| Wired | Wireless | |
|---|---|---|
| Speed and reliability | Usually faster and more reliable | Slower, and affected by interference and walls |
| Security | More secure: a physical connection is needed | Easier to intercept: needs encryption |
| Mobility | Devices must be near a socket | Devices can move around within range |
| Cost | Cables and installation cost | No cabling, easier to extend |
| Range | Limited by cable length | Limited by signal strength |
Transmission media
- Copper cable (twisted pair, coaxial): cheap and common, but the signal weakens over distance and it is affected by interference.
- Fibre-optic cable: carries pulses of light through glass or plastic. It has very high bandwidth and speed, covers long distances and is not affected by electrical interference, but it is more expensive and harder to install.
- Radio waves (including WiFi): used for short-range wireless networks; easy to set up, but affected by interference and can be intercepted.
- Microwaves: used for line-of-sight links between towers, with a high data rate, but they need a clear path.
- Satellite: reaches remote places and covers huge areas, but has a noticeable delay (latency) and can be affected by weather.
Hardware used in a LAN
- Switch: connects devices in a LAN. It learns each device's address and sends data only to the intended device, which reduces collisions.
- Server: a computer that provides services to the other computers.
- Network interface card (NIC): a circuit board or chip in a device that allows it to connect to a network by cable. Each has a unique MAC address. A wireless NIC (WNIC) does the same for a wireless connection.
- Wireless access point (WAP): connects wireless devices to a wired network.
- Cables: copper or fibre-optic, to carry the data.
- Bridge: connects two LANs, or two segments of a LAN, so they act as a single network. It decides whether data needs to cross to the other side.
- Repeater: regenerates (boosts) a signal so it can travel further along a cable.
The router
A router connects different networks together (for example a home LAN to the internet). It reads the destination IP address of each packet and uses a routing table to decide the best next route. Most home routers also act as a switch and wireless access point, assign IP addresses to devices (using DHCP) and include a firewall. The router has a public IP address on the internet side.
Ethernet and collision handling
Ethernet is the most common wired LAN standard. If two devices on a shared medium send at the same moment, their signals collide and the data is corrupted. Ethernet uses CSMA/CD (Carrier Sense Multiple Access with Collision Detection):
- A device listens to check that the channel is free (carrier sense).
- If the channel is free, it sends its data and keeps listening.
- If it detects a collision, both devices stop, send a jamming signal, and each waits a random length of time before trying again.
Wireless networks avoid collisions with a related method (CSMA/CA). A modern switched star network makes collisions rare, because each device has its own link to the switch.
Bit streaming
Bit streaming sends data as a continuous flow of bits, so it can be played while it is still arriving.
- On-demand: a recording that was made earlier is stored on a server, and the user chooses when to start it (a film or a song). The file is buffered and played; the user can pause and skip.
- Real-time (live): an event is captured and sent as it happens (a live sports match). The data is sent as it is created, so it cannot be fully buffered.
The bit rate (bits per second) is how fast the data arrives. If the connection's broadband speed is lower than the bit rate the stream needs, the buffer empties and the video stalls or drops in quality. A higher bit rate gives better quality but needs more bandwidth.
The internet and the World Wide Web
Hardware that supports the internet
- Modem: converts the digital signal from a computer into the form needed for the line (and back again), for example over a telephone line.
- Public Switched Telephone Network (PSTN): the worldwide telephone network that can also carry data.
- Dedicated (leased) lines: permanent, high-quality connections rented from a telecoms company, used by businesses for a guaranteed speed.
- Cell phone (mobile) network: radio links to masts, used for mobile internet.
- Routers and servers link the networks together and store and send web content.
IP addresses
Every device connected to the internet has an IP address that identifies it, so data can be sent to the right place. A packet carries the destination IP address, and routers use it to forward the packet.
- IPv4 uses 32 bits, written as four numbers from 0 to 255 separated by dots, for example
192.168.0.1. There are about 4.3 billion possible addresses, which is now not enough. - IPv6 uses 128 bits, written as eight groups of four hexadecimal digits separated by colons, and gives an enormous number of addresses.
- A public IP address is visible on the internet; a private address is used inside a LAN and is not routed on the internet. Addresses can be static (fixed) or dynamic (given out temporarily by DHCP).
Subnetting
An IP address has two parts: a network ID (which network the device is on) and a host ID (which device on that network). Subnetting splits one large network into smaller subnetworks (subnets) by using some of the host bits as extra network bits. A subnet mask (for example 255.255.255.0) shows which bits of the address are the network part.
Benefits: traffic stays inside each subnet, which reduces congestion; each subnet is easier to manage and to secure (for example, the finance department can have a subnet of its own); and a router only has to pass data between subnets when it is for a device on another one.
Public and private addresses, and security. A private address is not visible on the internet, so a device on a LAN cannot be reached directly from outside: the router hides it behind its one public address (network address translation), which makes it safer. A public address can be reached from anywhere on the internet, so a device or server using one is more exposed and needs a firewall and other protection. A static address never changes, so it suits a web server or printer that must always be found at the same place, but an attacker can target it easily. A dynamic address is given out by DHCP for a limited time, which is easier to manage and harder to track, but it is not suitable for something that must be found at a fixed address.
URLs and the Domain Name Service
A Uniform Resource Locator (URL) is the address of a resource on the web, for example:
https://www.example.com/shop/index.html
httpsis the protocol (HTTPS is HTTP with encryption).www.example.comis the domain name (the host)./shop/index.htmlis the path to the file on the server.
People find names easier to remember than IP addresses. The Domain Name Service (DNS) translates a domain name into an IP address:
- The browser sends the domain name to a DNS server (usually the ISP's).
- If that server holds the record, it returns the IP address. If not, it asks other DNS servers up a hierarchy until the address is found.
- The browser uses the IP address to contact the web server and request the page.
Practise this chapter
Every past-paper question on these topics, with the mark scheme.