Notes · Computer ScienceUK · A-Levels

Fundamentals of communication and networking

This chapter explains how computers talk to one another. It begins with the principles of data transmission - serial and parallel, synchronous and asynchronous, bit rate and bandwidth - then covers network topologies and types, the TCP/IP four-layer model that runs the internet, packet switching and addressing, network security and wireless, and the web technologies and client models built on top. Calculations of bit rate and transfer time sit alongside the descriptive content.

5 sections~17 min reading time3 competenciesLevel Foundation 1 · Standard 3 · Advanced 1

T·0888 / 14

Exam profile

AO1 · Know transmission methods, topologies, the TCP/IP model, packet switching, security measures and web technologiesAO2 · Calculate bit rate and transfer time, and describe how a packet is encapsulated and routed through the stackAO3 · Evaluate topology, security and client-computing choices for a scenario

Operators:explaincalculatedescribecompareevaluate

basic level

AS-Level expects serial vs parallel transmission, bit rate, simple topologies and the client-server model.

higher level

The full A-Level adds the TCP/IP stack, packet switching and routing, IP addressing and DNS, network security, wireless (CSMA/CA), and web technologies.

Depth

Text

Contents · 5 sections

Fundamentals of communication and networking

Communication basics: serial, parallel and bit rate#

FoundationAQA 7517 4.9.1DfE GCE Computer Science - communication

Data can be transmitted serially or in parallel. Serial transmission sends bits one after another down a single wire; it is cheaper, needs fewer wires, and does not suffer from timing differences between wires, so it works reliably over long distances and at high speed (USB, Ethernet and the internet are all serial). Parallel transmission sends several bits at once down multiple wires; it can be faster over very short distances but suffers from skew (bits arriving at slightly different times) and crosstalk as distance and speed rise, which is why serial has largely replaced it except inside the machine.

Transmission also differs in how sender and receiver stay in step. Synchronous transmission uses a shared clock signal so that a continuous stream of bits is sent in exact time with the receiver - efficient for large, steady flows. Asynchronous transmission sends data in small units framed by start and stop bits, with no shared clock, so each unit is self-timing - simpler and used where data is bursty (a keyboard). The start/stop bits add overhead but let the two ends resynchronise on every character.

Three related quantities describe capacity and must not be confused. Bit rate is the number of bits per second actually transmitted (bps). Baud rate is the number of signal changes (symbols) per second; if each symbol encodes more than one bit, the bit rate exceeds the baud rate -

\text{bit rate} = \text{baud rate} \times \text{bits per symbol}

. Bandwidth is the range of frequencies a channel can carry, and a greater bandwidth allows a higher bit rate. Latency is the delay before a transfer begins - separate from throughput.

The practical calculation you must be able to do is transfer time: since

\text{bit rate} = \text{bits transferred} \div \text{time}

, the time to send a file is its size in bits divided by the bit rate. The commonest mistake is forgetting to convert bytes to bits (multiply by 8) before dividing by a rate quoted in bits per second. Being fluent in this conversion, and in the bit-rate/baud-rate distinction, earns reliable AO2 marks.

\text{transfer time} = \dfrac{\text{file size in bits}}{\text{bit rate}}

Transfer time

Convert the file size to bits (bytes x 8) and divide by the bit rate in bits per second.

\text{bit rate} = \text{baud rate} \times \text{bits per symbol}

Bit rate from baud

Encoding more than one bit per signal change lets the bit rate exceed the baud rate.

Worked example

Calculating a transfer time

How long does it take to transfer a 4.5 MB file over an 18 Mbps connection? Take 1 MB = 10^6 bytes and 1 Mbps = 10^6 bits per second.

File size in bits
$4.5\text{ MB} = 4.5 \times 10^6 \text{ bytes} = 4.5 \times 8 \times 10^6 = 36 \times 10^6$ bits.
Divide by the bit rate
$\text{time} = \dfrac{36 \times 10^6 \text{ bits}}{18 \times 10^6 \text{ bps}} = 2$ seconds.

Result: The transfer takes 2 seconds. The key step is converting the 4.5 MB to 36 Mbit before dividing by the 18 Mbps rate.

Typical mistakes

Confusing bit rate (bits per second) with baud rate (symbols per second) - they are equal only when each symbol carries exactly one bit.
Forgetting to multiply a file size in bytes by 8 to get bits before dividing by a bit rate in bps.

Active revision

A 4.5 MB file is sent over an 18 Mbps connection (take 1 MB = 10^6 bytes). Calculate the transfer time. Then find the bit rate of a link running at 2400 baud where each symbol encodes 3 bits.

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for computer science (Department for Education) · AQA A-level Computer Science 7517 specification (AQA)

Network topologies and types#

StandardAQA 7517 4.9.2DfE GCE Computer Science - topologies

Star topology

Fig. 1Each device has its own cable to a central switch. A single cable failure isolates only that device, and the switch can send data directly between ports, so performance is good - the standard modern wired LAN. The switch is, however, a single point of failure.

A network's topology is the arrangement of its connections, and the two required physical topologies are the bus and the star. In a bus topology (shown opposite) every device connects to a single shared backbone cable terminated at both ends; it is cheap and easy to set up, but the shared cable is a single point of failure and a bottleneck - only one device can transmit at a time, and a break in the backbone brings down the whole network. In a star topology every device has its own cable to a central switch (also shown); it is more robust (one cable failing affects only that device), performs better under load, and is the standard for modern wired LANs, at the cost of more cabling and dependence on the central switch.

It is important to separate the physical topology (how the cables are actually laid) from the logical topology (how data logically flows). A network can be physically wired as a star (every device to a central switch) but behave logically as a bus if the switch broadcasts to all - the physical and logical arrangements need not match, a distinction the exam probes.

Networks are also classified by scale and control. A local area network (LAN) covers a small geographical area (a building or site) and is usually owned by one organisation; a wide area network (WAN) spans large distances (cities, countries) using infrastructure often owned by third parties - the internet is the largest WAN. Independently, a network can be organised as peer-to-peer, where computers share resources as equals with no central server (simple, cheap, resilient, used for file sharing), or client-server, where dedicated servers provide resources and services to client machines (central control, security and backup, but a cost and a potential single point of failure).

Choosing a topology and model is a trade-off examined in scenario questions. A small home or informal network suits peer-to-peer; an organisation needing central security, shared data and backups suits client-server. A star is chosen for reliability and performance, a bus only where cost and simplicity dominate. A good answer justifies the choice against the scenario's priorities - cost, reliability, scale, security - rather than merely naming a topology.

Bus topology

Fig. 2Every device taps into a single shared backbone cable terminated at both ends. It is cheap and simple, but only one device can transmit at a time and a break in the backbone disables the whole network - a single point of failure.

Worked example

Choosing a network for an office

An office of eight computers needs shared central files, automatic backups and per-user access control. Recommend a model and topology and justify them.

Model
Client-server: a dedicated server holds the shared files, performs central backups and enforces user accounts and permissions - exactly the stated requirements, which peer-to-peer cannot manage centrally.
Topology
A physical star with each computer wired to a central switch: reliable (one failed cable isolates only one machine) and good performance for shared file access.
A disadvantage
Both introduce a single point of failure (the server; the switch) and cost more than a peer-to-peer bus - mitigated by backups and redundant hardware.

Result: Recommend a client-server model on a star topology: it meets the central-control, backup and security needs, accepting the extra cost and the server/switch as points of failure.

Typical mistakes

Confusing physical topology (the cabling) with logical topology (how data flows) - a physical star can be a logical bus.
Muddling peer-to-peer with client-server, or assuming client-server is always better - peer-to-peer suits small, informal networks.

Active revision

A small office of eight computers needs shared files, central backups and controlled user access. Recommend a network model and a topology, justify each against the requirements, and state one disadvantage of your choice.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Computer Science 7517 specification (AQA)

The TCP/IP stack, packet switching and addressing#

AdvancedAQA 7517 4.9.3DfE GCE Computer Science - TCP/IP

The TCP/IP four-layer stack

Fig. 3Data descends the stack, each layer adding its own header (encapsulation): the application layer's message is wrapped as a TCP segment, then an IP packet, then a link-layer frame. At the receiver it rises back up, each layer stripping its header. IP addressing lives at the network layer, port numbers at the transport layer.

The internet runs on the TCP/IP protocol stack, a set of protocols organised into four layers (shown opposite), each with a defined job and each hiding the layer below - another instance of layered abstraction. From the top: the application layer provides protocols the user's program speaks (HTTP for the web, SMTP for email, FTP for files, DNS for name lookup); the transport layer (TCP or UDP) splits the message into numbered segments, manages reliable delivery and uses port numbers to identify the application; the network (internet) layer (IP) adds source and destination IP addresses and handles routing across networks; and the link layer deals with the physical transmission on the local network (MAC addresses, frames).

As data is sent it travels down the stack, and at each layer a header is added - a process called encapsulation - so the transport layer wraps the data in a TCP segment, the network layer wraps that in an IP packet, and the link layer wraps that in a frame. At the receiving end the data travels up the stack and each layer strips off and acts on its own header (de-encapsulation). Each layer talks logically to its counterpart on the other machine, using the services of the layer beneath.

The internet uses packet switching: a message is broken into packets each carrying source and destination addresses and a sequence number (the packet structure is shown opposite), and each packet is routed independently by routers that choose the best available path hop by hop. Packets may take different routes and arrive out of order, so they are reassembled in sequence at the destination and any missing ones re-requested. Packet switching uses the network efficiently and is resilient - if a link fails, packets simply route around it - which is why the internet was designed this way.

Every device on the internet needs a unique IP address to be reachable, and because numeric addresses are hard for humans, the Domain Name System (DNS) translates human-readable names (www.example.com) into IP addresses via a distributed hierarchy of name servers - a lookup that happens before almost every connection. Understanding the layers, encapsulation, packet switching and DNS, and being able to trace a message through them, is core Paper 2 content.

The structure of a packet

Fig. 4Each packet carries a header (with the source and destination IP addresses, a sequence number for reassembly, and the protocol), the payload (a chunk of the message), and a trailer with an error check. The addresses let routers forward it; the sequence number lets the destination reassemble the message in order.

Worked example

Encapsulation down the stack

Explain what each TCP/IP layer does to a browser's HTTP request as it is prepared for transmission.

Application layer
The browser forms an HTTP request (e.g. GET /page). This is the data to be sent, expressed in the application protocol.
Transport layer
TCP splits the request into numbered segments and adds a header with source and destination port numbers (port 80/443 for web) so the right application receives it, plus sequence numbers for reliable, ordered delivery.
Network layer
IP wraps each segment in a packet, adding the source and destination IP addresses so routers can forward it across networks.
Link layer
The packet is placed in a frame with MAC addresses for the next hop and sent over the physical medium.

Result: Each layer adds its own header (encapsulation) as the request descends; the receiving computer reverses the process, each layer removing its header until the server's application gets the original HTTP request.

Typical mistakes

Getting the layer order or responsibilities wrong - IP addressing is the network layer, ports are the transport layer.
Thinking all packets of a message follow the same route and arrive in order - they are routed independently and reassembled by sequence number.

Active revision

Describe what happens at each of the four TCP/IP layers as a web browser sends a request, and explain how packet switching delivers the reply even if one network link fails midway.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Computer Science 7517 specification (AQA)

Network security and wireless#

StandardAQA 7517 4.9.4DfE GCE Computer Science - network security

Networks are exposed to threats, and several security measures defend them. A firewall sits between a network and the outside world and controls traffic by packet filtering - inspecting each packet's source, destination and port against a set of rules and blocking anything not permitted; a proxy server sits between clients and the internet, forwarding requests on their behalf to hide the internal addresses, cache content and filter access. Encryption protects data in transit so that even intercepted packets cannot be read without the key - the reason HTTPS is used for anything sensitive.

The malware you must know comes in three kinds. A virus attaches itself to a host file or program and spreads when that file is run or shared, relying on user action. A worm is self-replicating and spreads automatically across a network without needing a host file or user action, which makes it spread far faster. A trojan disguises itself as legitimate software that the user installs willingly, then carries out a hidden malicious payload (a backdoor, data theft). Defences combine antivirus software, firewalls, prompt patching, and - crucially - user education, since many attacks rely on tricking a person (phishing, social engineering).

Wireless networking (Wi-Fi) brings its own issues because the shared radio medium is open to eavesdropping and collisions. A network is identified by its SSID and secured with encryption (WPA2/WPA3); an unsecured or weakly-secured access point is an easy target. Because two wireless devices transmitting at once would collide and cannot easily detect it, Wi-Fi uses CSMA/CA (carrier sense multiple access with collision avoidance): a device listens first (carrier sense) and, to avoid collisions, may use a request-to-send / clear-to-send (RTS/CTS) handshake to reserve the channel before transmitting.

Security is layered - no single measure is sufficient. A firewall stops unauthorised connections, encryption protects the data itself, antivirus catches known malware, and careful configuration (strong Wi-Fi encryption, changed default passwords) closes easy routes in, while user awareness defends against the social attacks that technology cannot. Recommending an appropriate combination of measures for a scenario, and distinguishing the malware types, are standard exam requirements.

Typical mistakes

Confusing a virus (needs a host file and user action) with a worm (self-replicating, spreads automatically) or a trojan (disguised as legitimate software).
Thinking a firewall encrypts data - it filters/blocks traffic; encryption is a separate measure that protects the data's contents.

Active revision

Explain the difference between a virus, a worm and a trojan, and recommend three measures a company should take to protect its network, justifying each. Explain why Wi-Fi uses collision avoidance rather than collision detection.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Computer Science 7517 specification (AQA)

The web: DNS and client technologies#

StandardAQA 7517 4.9.5DfE GCE Computer Science - web technologies

The client-server request-response cycle

Fig. 5The browser (client) sends an HTTP(S) request to the web server, which processes it and returns a response (the HTML page and its resources). DNS first resolves the domain name to the server's IP address; HTTPS encrypts the exchange.

The World Wide Web is a service that runs on the internet, built on the client-server model (shown opposite): a client (the browser) sends a request to a web server, which returns a response (a web page and its resources). The request and response use HTTP (Hypertext Transfer Protocol), or HTTPS when the connection is encrypted with TLS so that data such as passwords and card details cannot be read if intercepted. Before the request can be sent, DNS resolves the site's domain name to the server's IP address.

A web page is built from three complementary technologies. HTML (Hypertext Markup Language) defines the structure and content - headings, paragraphs, links, images. CSS (Cascading Style Sheets) defines the presentation - layout, colours, fonts - kept separate from the content so a site's look can change without touching its structure. JavaScript adds behaviour - code that runs in the browser to respond to the user, validate input and update the page dynamically. This separation of structure, style and behaviour is itself a design principle worth stating.

Where the work happens defines the client model. In thin-client computing most processing and storage is done on powerful central servers, and the client mainly displays results (a web app, a terminal); it is easy to manage and secure centrally and needs little client hardware, but depends heavily on the network and the server. In thick- (fat-) client computing the client machine does much of the processing and storage locally (a desktop application, a game); it works offline and uses local power, but is harder to maintain and update across many machines. The choice trades central control and low client cost against local capability and network independence.

These pieces fit together on every page load: DNS resolves the name, the browser opens an HTTP(S) connection to the server, requests the HTML, and then fetches the CSS, JavaScript and images it references, running the JavaScript locally. Understanding this flow, the roles of HTML/CSS/JavaScript, the security value of HTTPS, and the thin-versus-thick-client trade-off, rounds off the networking topic and connects it to everyday experience.

Worked example

From address bar to rendered page

Outline the sequence of events when a user types a web address and presses enter, naming the protocols and technologies involved.

Name resolution
The browser asks DNS to translate the domain name into the web server's IP address.
Request
The browser opens an HTTP (or encrypted HTTPS) connection to that IP address and sends a GET request for the page.
Response
The server returns the HTML document; the browser then requests the CSS, JavaScript and images the HTML references.
Render
The browser uses the HTML for structure, applies the CSS for presentation, and runs the JavaScript for interactivity, displaying the finished page.

Result: DNS resolves the name, HTTP(S) carries the request and response, and HTML, CSS and JavaScript together structure, style and animate the page the user sees.

Typical mistakes

Confusing the roles of HTML (structure), CSS (presentation) and JavaScript (behaviour), or thinking CSS or HTML can perform logic.
Treating the web and the internet as the same thing - the web is one service running on the internet.

Active revision

Describe, step by step, what happens from typing a web address to the page appearing, naming DNS, HTTP/HTTPS and the roles of HTML, CSS and JavaScript. Then recommend thin- or thick-client computing for a school's shared administrative system and justify it.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Computer Science 7517 specification (AQA)

References & sources

Sources

Department for Education

GCE AS and A level subject content for computer science

AQA

AQA A-level Computer Science 7517 specification

Fundamentals of communication and networking

5 sections~17 min reading time3 competenciesLevel Foundation 1 · Standard 3 · Advanced 1

Fundamentals of communication and networking

Key points

Calculating a transfer time

Key points

Choosing a network for an office

Key points

Encapsulation down the stack

Key points

Key points

From address bar to rendered page

Sources

Fundamentals of communication and networking

Key points

Calculating a transfer time

Key points

Choosing a network for an office

Key points

Encapsulation down the stack

Key points

Key points

From address bar to rendered page

Sources