Networks - Lesson 2
Lesson 2 of 6 Networks Series

Network Topologies

Discover how the physical arrangement of devices on a network affects performance, cost and fault tolerance. Compare star, bus, ring and mesh topologies and understand exactly which fails when a cable or device breaks.

GCSE and A-Level 6 sections Fault simulator + quiz

What is a Network Topology?

A hospital IT manager needs to build a new network for 80 computers across three floors. One doctor insists every computer must stay online even if multiple cables fail. The finance director wants to keep costs as low as possible. The head of IT says the network must be easy to manage and expand. These three requirements point to three completely different topologies.

Think about it: Why might the layout of cables between computers affect how well the network copes with failures? Could you design a layout where cutting one cable cuts off the whole network?

A network topology describes the arrangement of devices and the connections between them. There are two types: physical topology (how cables and hardware are actually arranged) and logical topology (how data flows through the network). At GCSE, questions focus on physical topology.

Topology determines three things that exam questions always ask about: fault tolerance (what breaks when something fails), cost (how much cable and hardware is needed) and performance (how efficiently data travels).

Topology
The physical or logical arrangement of devices and connections in a network.
Fault tolerance
How well a network continues to function when a device or cable fails.
Node
Any device connected to a network - computers, printers, servers, switches.
Single point of failure
A component whose failure would bring down the entire network.

Star and Bus Topologies

The two most commonly tested topologies. Star is the modern standard; bus is historical but still appears in every exam paper.

Star
All devices connect to a central switch or hub. The most common topology in modern networks.
SWITCH PC A PC B PC C PC D PC E PC F
  • If one cable fails, only that device is affected - all others remain online
  • Easy to add new devices without disrupting the network
  • Performance is consistent - devices do not share bandwidth with each other
  • Easy to identify and isolate faults
  • If the central switch fails, the entire network goes down
  • Requires more cable than bus topology
  • Central switch is a single point of failure
What fails when... One cable breaks: 1 device disconnects, 5 remain online.
Central switch fails: all 6 devices disconnect immediately.
Bus
All devices connect to a single shared cable (the backbone). Historical - rarely used today.
T T PC A PC B PC C PC D Shared backbone cable
  • Uses less cable than star - cheaper to install
  • Simple to set up for a small number of devices
  • If the backbone cable breaks, the entire network fails
  • All devices share bandwidth - performance drops as more devices are added
  • Data collisions occur when two devices transmit simultaneously
  • Difficult to diagnose faults and to add new devices
  • Terminators are required at each end of the backbone
Single point of failure Any break in the backbone cable disconnects ALL devices. There is no alternative path for data.

Ring and Mesh Topologies

Ring topology creates a closed loop; mesh topology creates multiple paths. Mesh is the basis of how the internet itself is structured.

Ring
Devices connected in a closed loop. Data travels around the ring in one direction.
PC A PC B PC C PC D PC E
  • Data travels in one direction, reducing collisions compared to bus
  • Equal access for all devices - no device dominates bandwidth
  • Can handle heavy traffic better than bus topology
  • A break in the ring can disrupt the entire network
  • Adding or removing a device disrupts the network
  • Fault diagnosis is more difficult - data must travel the whole ring
  • Slower than star for large networks as data passes through every device
What fails when... One cable breaks: data cannot complete the loop - all devices may be affected.
One device fails: the ring is broken at that point, disrupting the whole network.
Mesh
Full mesh: every device connects to every other. Partial mesh: some direct connections. Used in the internet.
PC A PC B PC C PC D Full mesh: 4 nodes, 6 connections
  • Extremely high fault tolerance - multiple paths exist between devices
  • Data can be re-routed if a connection fails
  • Very fast - data can take the most direct path
  • No single point of failure (full mesh)
  • Very expensive - a full mesh with n nodes needs n(n-1)/2 connections
  • Complex to set up and manage
  • Large amount of cabling required
Highly fault tolerant A full mesh with 4 devices has 6 connections. Multiple connections must fail before any device is isolated. The internet uses partial mesh for this reason.
Mesh connections formula

A full mesh network with n nodes requires n(n-1) / 2 connections. A network with 10 devices needs 45 connections; with 20 devices, 190 connections. This is why full mesh is rare in practice - partial mesh is used instead, providing redundancy without the full cost.

Topology Comparison

Exam questions regularly ask you to compare topologies across specific criteria. This table covers the key dimensions.

Topology Fault Tolerance Cost Performance Scalability Used where
Star Medium - switch failure = total failure; cable failure = 1 device Medium - more cable than bus; switch required High - dedicated paths, no collisions Easy - add device without disruption School networks, offices, homes
Bus Low - backbone failure = total failure Cheap - least cable, no central device Low - shared bandwidth, collisions Difficult - disrupts network to add devices Legacy networks only
Ring Low-Medium - one break can fail all Medium - similar to bus Medium - no collisions but sequential Difficult - disrupts ring to add devices Industrial networks, some WANs
Full Mesh Very High - multiple paths always available Very High - n(n-1)/2 connections Very High - direct paths, no bottleneck Very Hard - connections grow exponentially The internet (partial), critical systems

Topology Fault Simulator

The most common exam question on topology is: "What happens to the network if a cable or device fails?" Click on a cable or device below to simulate a failure and see exactly which devices lose connection.

Network Fault Simulator
Click any cable (line) or device to simulate a failure. Watch how fault tolerance differs between topologies.
Click any cable (line) or device to simulate a failure
Devices still online
5 / 5
All devices connected
Select a topology and click a cable or device to simulate a failure.
Star topology: The central switch is the only single point of failure. Individual cable breaks only affect one device - a major advantage over bus topology.
Classify the Scenarios
Which topology best fits each use case? Drag each scenario to the correct column.
30 computers in a school office, easy to manage
Internet backbone connecting major cities
1990s small office, minimal cable cost
Home broadband router connecting 6 devices
Military network requiring maximum fault tolerance
Industrial factory floor control systems
Star
Bus
Ring
Mesh

Mesh Connection Calculator

Use the formula n(n-1) ÷ 2 to calculate how many connections a full mesh network needs. See how rapidly the number grows as you add more devices.

Number of devices (n)
6
220
Calculation (worked example)
Connections required for 2 to 10 devices

Choosing the Right Topology

Exam questions frequently describe a scenario and ask you to recommend and justify a topology choice. You need to link the requirements of the scenario directly to the characteristics of the topology.

Choose Star when...
  • The network must be easy to manage and expand
  • Individual device failures must not affect others
  • Cost is reasonable but not the absolute priority
  • Examples: school, office, home network
Choose Mesh when...
  • The network absolutely cannot fail under any circumstances
  • Multiple simultaneous failures must be tolerated
  • Budget is not a constraint
  • Examples: internet backbone, military, hospitals
Mark-scheme language

When justifying a topology choice, always state the requirement then link it to the topology feature. Wrong: "Star topology is good." Correct: "Star topology is most suitable because individual cable failures only disconnect one device, meaning the other 29 computers in the office can continue working - the school cannot afford network downtime during lessons." Marks come from the link, not the label.

Lesson Quiz - 5 questions
Based on real exam question styles. Choose your answer to reveal instant feedback.
Question 1 of 5
A school has 40 computers all connected to a central switch. One cable between a computer and the switch breaks. How many computers lose their connection?
Question 2 of 5
Which topology uses the least cabling and requires no central device?
Question 3 of 5
A full mesh network has 6 computers. How many connections are required?
Question 4 of 5
Which of the following is a disadvantage of star topology compared to bus topology?
Question 5 of 5
A hospital requires a network where even if multiple connections fail, all critical systems stay online. Which topology is most appropriate and why?
0
/5
Extended thinking

1. A small business with 8 computers is choosing between star and bus topology. Give two reasons to choose star and one reason why bus might still be considered.

Star advantages (choose 2):
1. If one cable fails, only that device loses connection - the other 7 computers continue working. With a bus, the entire network fails.
2. Easy to identify and fix faults - the faulty cable connects only one device to the switch, so the problem is isolated.
3. Easy to add new devices without disrupting the existing network.

Bus consideration:
Lower cost - bus uses less cable and requires no central switch, which could matter for a small business with a very limited budget.

Note: At GCSE, star topology is almost always the recommended choice for modern networks. Answers should link reasons to the scenario.
Extended thinking

2. The internet is described as a "network of networks" using a partial mesh topology. Explain why full mesh is not used for the internet, and why partial mesh is preferred.

Why not full mesh: The internet has millions of routers. A full mesh would require millions x (millions - 1) / 2 connections - an impossible number of direct physical links. The cost and complexity would be prohibitive.

Why partial mesh works: Each router connects to several others, not all of them. This provides redundancy (multiple paths exist if one link fails) without the exponential cost of full mesh. Data is routed intelligently around failures. A packet from London to New York might travel via different routes each time, selecting whichever path is fastest and available.

Key point: partial mesh gives most of the fault tolerance benefit of full mesh at a fraction of the cost.
Extended thinking

3. Evaluate star topology for use in a large secondary school with 500 computers across 20 classrooms. Consider performance, fault tolerance, cost and management.

Performance: Star topology gives each device a dedicated connection to the switch, avoiding collisions and providing consistent performance - suitable for 500 simultaneous users.

Fault tolerance: A single cable failure only disconnects one device. However, if a classroom switch fails, that entire classroom loses connectivity. Hierarchical star (switches connecting to a core switch) mitigates this.

Cost: Requires significant cabling (each device to the nearest switch) and multiple switches across 20 classrooms. However, equipment costs have dropped significantly, making this acceptable.

Management: Centralised management - the network administrator can monitor and control all switches from one location. Adding a new device only requires one cable run to the nearest switch.

Conclusion: Star topology is the appropriate choice for this school. The advantages (fault isolation, performance, easy management) far outweigh the cost of additional cabling, which is standard practice in modern school networks.
Printable Worksheets

Practice what you've learned

Three levels of worksheet. Download, print and complete offline.

Recall
Topology Fundamentals
Key term matching, topology naming from diagrams, and fill-in-the-blank advantage/disadvantage tables.
Download
Apply
Fault Analysis Tasks
Diagrams showing broken cables/devices - identify which topology, how many devices affected, and suggest improvements.
Download
Exam Style
Justify and Evaluate
Scenario-based questions requiring topology recommendations with full written justifications. Mirrors exam question style.
Download
Networks Flashcards
Review all Networks terms with spaced-repetition flashcards. Filter by lesson.
Open Flashcards
Networks Lesson 2 - Teacher Resources
Network Topologies
Teacher mode (all pages)
Shows examiner notes on exam practice pages
Suggested starter (5 min)
Give every student a piece of paper and ask them to write their name at the top. Now say: "You need to pass a message to every other person in the room - draw a line for every connection you would need." After 90 seconds, ask: "How many of you drew lines from a central person? How many drew lines between everyone? Now - if I remove the central person, what happens?" They have discovered star and mesh topologies, and the central point of failure, before you have defined a single term.
Lesson objectives
1Define network topology and distinguish between physical and logical topology.
2Describe and draw star, bus, ring and mesh topologies, identifying nodes and connections.
3State at least two advantages and two disadvantages of each topology.
4Explain the fault tolerance of each topology, identifying single points of failure.
5Apply the formula n(n-1)/2 to calculate connections in a full mesh network.
6Justify a topology choice for a given scenario using specific criteria.
Key vocabulary (board-ready)
Topology
The physical or logical arrangement of devices and connections in a network.
Star topology
All devices connect individually to a central switch. Each device has its own dedicated cable. A single cable failure only affects one device; a switch failure affects all.
Bus topology
All devices share a single backbone cable with terminators at each end. A backbone failure disconnects the entire network. Now largely obsolete.
Ring topology
Devices are connected in a closed loop. Data travels in one direction. A single cable break can disable the whole ring.
Mesh topology (full)
Every device is directly connected to every other. Uses n(n-1)/2 connections. Highly resilient but expensive and complex.
Hybrid topology
A combination of two or more topology types (e.g. a star-of-stars, or a star connected via a mesh backbone).
Single point of failure
A component whose failure causes the entire network (or a large section of it) to stop working. Bus: the backbone. Star: the central switch.
Node
Any device connected to a network - computer, printer, router, server etc.
Suggested lesson plan (60 min)
0-5 min: Starter: name/connection paper activity. Reveal the concepts of star and mesh emerge naturally from student drawings. Define topology formally.
5-20 min: Topology Explorer - work through each topology as a class using the interactive SVG cards. Students sketch each topology in their notes. Focus: what connects to what, and what breaks when.
20-30 min: Fault Simulator - student-led in pairs. Students predict what happens when each cable or device is clicked, then test their prediction. Discuss results as a class.
30-40 min: Drag activity and comparison table. Students complete the classify activity then fill in a topology comparison grid in their notes covering fault tolerance, cost, performance, and scalability.
40-52 min: Scenario-based exam practice. Students write a justified topology recommendation for one think-deeper scenario under timed conditions (12 minutes). Pair-share before class feedback.
52-60 min: Exit ticket on mini-whiteboards. Review the most common wrong answer as a class. Students write one thing they will remember and one thing they are still unsure about on a sticky note.
Discussion prompts
A building has 12 floors, each with 10 computers on a floor switch. All floor switches connect to one central core switch. Is this star, mesh, or hybrid? What are the resilience implications of the design? What would you change to eliminate the single point of failure?
Using n(n-1)/2, calculate the connections needed if just 100 major internet service providers were fully meshed. Then explain why this makes full mesh impractical at internet scale, and what topology the internet actually uses instead.
A school switches from bus to star topology. More cable is needed and a switch must be bought. Identify at least three specific operational benefits that justify these extra costs in a school environment.
Ring topology sends data in one direction around the loop, which some argue makes it "fair" since every device gets a turn. But what happens to a ring when a single device (not just a cable) fails? How does this compare to a bus topology failure?
Common misconceptions
X"Star topology has no single point of failure" - the central switch IS a single point of failure. If the switch fails, all devices lose connection. Students often confuse the resilience of individual cable failures with overall fault tolerance. The correct statement is: "a single cable failure in star only affects one device."
X"Bus topology is still widely used" - bus topology is almost entirely obsolete in practice. It appears in exam questions for historical comparison and to test knowledge of its disadvantages, not because students will encounter it. Clarify this is a legacy topology.
X"Ring topology always fails if one cable breaks" - depends on implementation. Some ring networks use bidirectional rings, so data travels the other way if one segment fails. At GCSE, treat a single break as causing failure unless stated otherwise.
X"The internet uses star topology because everything connects to servers" - the internet uses partial mesh between core routers. Individual users connect in a star pattern to their ISP, but the internet backbone is a complex partial mesh.
X"More connections always means better performance" - mesh topology's connections grow as n(n-1)/2. Cost and complexity grow rapidly, and managing thousands of connections becomes a maintenance problem. Full mesh is used only where resilience is critical and cost is not a limiting factor.
Exit ticket questions
Name two topologies that have a single point of failure if their main connecting component breaks.
[2 marks - Bus: backbone cable; Star: central switch. Ring also acceptable if discussing a device failure.]
A company has 5 servers that must all communicate directly with each other. Calculate the minimum number of connections required. Show your working.
[2 marks - n(n-1)/2 = 5x4/2 = 10 connections]
Give one reason why the internet uses partial mesh rather than full mesh.
[1 mark - full mesh would require an impractical number of connections / too expensive / too complex to manage at scale]
State one advantage and one disadvantage of a star topology compared to a bus topology.
[2 marks - Advantage: individual cable failure only affects one device (not the whole network). Disadvantage: requires more cable / requires a central switch which is a single point of failure]
Homework idea
Students design a network for this scenario: "A new medical practice is opening with 8 computers, a shared printer, a server, and wireless access for a waiting room tablet. Draw and label a network topology diagram for this practice. Justify your chosen topology with reference to fault tolerance, cost, and scalability. Identify any single points of failure and explain how they could be addressed." Students bring the annotated diagram to the next lesson for peer review using a mark scheme you provide.
Classroom tips
The Fault Simulator is the most engaging element of this lesson. Give students 3-4 minutes to explore freely, then ask "prediction questions" before they click: "Predict what happens when the switch fails - write it down first." This metacognitive step significantly improves retention.
When teaching mesh calculations, build the table on the board: n=2 gives 1 connection, n=3 gives 3, n=4 gives 6, n=5 gives 10. Ask students to spot the pattern. The exponential growth becomes visible and far more memorable than simply quoting the formula.
For the star vs bus single-point-of-failure concept, use physical drama: ask one student to be the "switch" and pass messages between others. Remove the switch. Then simulate bus by having everyone pass a single note down a chain. Both activities make the failure modes visceral and memorable.
Avoid teaching "topology = shape". Topology is about the arrangement of connections and the resulting properties (fault tolerance, performance, cost). The diagram shape is just a representation. Students who understand properties rather than shapes perform significantly better in scenario questions.