A pupil is connected to the school’s wireless network, but the website they need will not load. Another device can print to a nearby printer while the internet connection is down. These observations make sense once we distinguish a local area network from the wider internet, the jobs of a switch, access point, router and modem, and the addresses that guide packets. Troubleshooting begins with the communication route rather than the vague instruction to ‘restart Wi-Fi’.
Ministry of Education V3.0’s Computing reference on Networks and Home Networking explains Singapore’s 2027 G3 Computing syllabus K349 Module 4.1 and 4.2: wired and wireless transmission, LAN and WAN scope, client-server and peer-to-peer architectures, star and mesh topologies, packets and error detection, modem and network devices, MAC/IPv4/IPv6 addressing, and systematic network reasoning. It includes a fictitious family learning network, 26 original worked scenarios, a twelve-stage route-audit protocol and the eduKate Depth–Load–Transfer method.
This is an independent eduKateSG teaching reference, not an official MOE/SEAB paper or network administration authorisation. Any equipment, addresses and network cases are fictional or documentation-only examples. Avoid changing live school or workplace network configurations without permission, and do not expose private credentials or students’ personal data.
1. A Network Connects Devices, Not Merely Websites
A computer network is a system of at least two computers connected through a transmission medium to exchange data. A network can function without a connection to the public internet: a local printer and computer may exchange data within one LAN even when an ISP link has failed. Conversely, an internet connection depends on successful communications across many networks and intermediary devices. The words ‘network’ and ‘internet’ therefore should not be used as synonyms.
2. Exact Official 2027 G3 Computing K349 Scope
The 2027 G3 Computing K349 syllabus, Module 4 Networking, specifies LANs/WANs, wired/wireless media, client-server and peer-to-peer architectures, star and mesh arrangements, protocols, packet communication, parity/checksum/echo/automatic repeat request, and components of a home network. It distinguishes MAC addressing for directing data on the same LAN from IPv4 and IPv6 addressing for routing between networks, and also includes Module 4.3 security and privacy. This chapter focuses on Modules 4.1–4.2; its security examples are contextual, not a replacement for the separate cybersecurity outcomes.
3. Local and Wide Are About Scope, Not Quality
A local area network typically covers a home, classroom or office location. A wide area network connects networks over geographically broader distances. The internet interconnects many LANs worldwide and is a prominent WAN example in the K349 explanation. A LAN is not automatically faster or more reliable in every real deployment; geographic scope is the distinguishing classification used in the school task.
4. Wired and Wireless Media Bring Different Trade-Offs
Wired Ethernet provides a guided physical transmission route and can be attractive where predictable throughput, low interference and fixed equipment matter. Wireless transmission uses radio and supports mobility and flexible placement, but performance can depend on walls, competing radio use and signal quality. Neither technology guarantees privacy or speed by itself; those outcomes depend on additional equipment, settings and conditions.
5. Devices Must Be Given Their Correct Jobs
A network interface controller enables a computer to communicate over a supported medium. A switch helps forward local frames between devices using MAC information. A wireless access point connects compatible wireless clients into a LAN. A router directs traffic between different networks using IP addressing. A modem connects different access technologies or protocols associated with an ISP, as described in the syllabus. One commercial box may combine several of these functions; the functions remain distinct.
6. MAC and IP Addresses Operate at Different Scopes
A conventional MAC address is commonly represented as 48 bits, for example 02:00:00:00:00:2A. IPv4 addresses are 32 bits, often written as four decimal octets; IPv6 addresses are 128 bits, often written in hexadecimal groups separated by colons. The syllabus asks students to compare purpose, bit length, typical representations and relative permanence. MAC identifiers can be locally assigned or changed, and IP addresses may be reassigned; none should be presented as a magical unchangeable fingerprint.
7. Packets Need Rules and Error Checks
A network protocol sets rules for communication. Data may be transmitted in packets rather than one indivisible stream. Simple parity checks, checksums and echo checks may help detect certain errors, while automatic repeat request mechanisms can arrange retransmission when missing or damaged transmissions are recognised. An error check is not proof of perfect data integrity against every conceivable corruption; it has defined limitations.
8. Eight Rooms of Network Reasoning
Room one identifies the two endpoints. Room two classifies LAN versus WAN. Room three identifies wired/wireless media. Room four names the network devices and their functions. Room five traces whether local or routed addressing is relevant. Room six follows packet transmission and error handling. Room seven tests observed symptoms against the route. Room eight selects an authorised, minimally disruptive check and transfers the method to an unfamiliar configuration.
9. Twelve-Stage Endpoint–Route–Verify Protocol
Stage 1 — Name the endpoints
Is the request from a tablet to a local printer, a student laptop to a shared school computer, or a home device to a server on the internet? Destination determines which network stages are necessary.
Stage 2 — Identify the network’s geographic scope
Two computers in one classroom may be on a LAN. A request to a distant service crosses networks and needs routing. ‘Wireless’ does not by itself mean ‘WAN’.
Stage 3 — Locate the transmission medium
Identify a wired Ethernet segment or a wireless link. Distinguish a signal issue from a remote server failure; they are not interchangeable problems.
Stage 4 — Map the local interface
Every communicating computer needs an appropriate network interface controller. State how its medium connects to a switch or access point, not merely that it ‘has internet’.
Stage 5 — Find the access point and switch roles
An access point handles wireless connections to the LAN, while a switch interconnects local wired devices. A combined consumer device can embody both functions.
Stage 6 — Decide whether a router is needed
For delivery across distinct IP networks, a router selects the route using IP addressing. Purely local exchanges may remain inside the LAN without internet routing.
Stage 7 — Explain the modem or ISP boundary
Home internet access often includes equipment connecting the home’s networking protocols to the ISP’s access technology. Treat that connection as separate from an access point’s radio function.
Stage 8 — Identify the address type
Use MAC concepts for local frame delivery and IP concepts for routed delivery. Read the address’s form rather than substituting MAC and IP as synonyms.
Stage 9 — Break the message into logical packets
Packet-based communication allows data to be transmitted as units under agreed protocols. A missing unit may affect the reconstructed communication even when other units arrived.
Stage 10 — Choose a suitable error check
Use parity or checksum concepts to spot particular transmission errors. Explain that a detected error can motivate a retransmission request under ARQ.
Stage 11 — Isolate a failure from observations
If local printing works but websites fail, investigate the wider route or destination before declaring every device disconnected. Compare independent observations rather than relying on one symptom.
Stage 12 — Verify on a new network diagram
Redraw the same logic for a library or home office using different device locations. Explain each hop and limitation without making unauthorised configuration changes.
10. Original Fictional Network: Riverbend Learning Home
In an invented household, two laptops connect by Wi-Fi to an access point, one desktop computer connects by Ethernet to a switch, and a printer connects to that same LAN. A router connects the LAN to the internet-facing modem and ISP service. One learning website is hosted on another network. This topology is a teaching diagram in words, not a recommendation to buy particular hardware.
When the ISP connection fails, the desktop may still print to the LAN printer and one laptop may still send files to the desktop if permissions and local services permit. Requests to remote websites, however, cannot complete over the unavailable wide-area path. A useful diagnosis tests local service and external connectivity separately. The result clarifies network boundaries without guessing at the cause from the word ‘Wi-Fi’.
11. Twenty-Six Worked Network Decisions: A–Z
12. Worked Example A — Two home computers exchanging files
The two computers form part of a network if they connect through a transmission medium to exchange data. Public internet service is not required for this definition. Sharing may still depend on software permissions, an issue distinct from connectivity.
13. Worked Example B — One house with several devices
A home containing a laptop, printer and network-attached desktop on the same local network is a LAN. Do not label it WAN merely because one device connects wirelessly.
14. Worked Example C — Schools in distant countries
Two schools exchanging data through interconnected distant networks involves WAN-style communication. The distance and interconnection distinguish it from a single classroom LAN.
15. Worked Example D — Ethernet versus Wi-Fi
A desktop beside a network switch can use Ethernet for a stable wired route, while a mobile tablet uses wireless access for movement. Neither medium alone guarantees internet access or strong security.
16. Worked Example E — Wired printer but failed internet
A desktop prints to a local printer despite an ISP outage. The observation is possible because local network communication does not necessarily depend on the external internet path.
17. Worked Example F — Router and switch are not interchangeable
A switch normally forwards local traffic using frame/MAC information, while the router forwards packets between different IP networks. Many consumer routers include built-in switches; combined packaging does not erase the functional distinction.
18. Worked Example G — Wireless access point’s role
A wireless access point provides a radio link into the LAN. If a laptop can reach the access point but remote websites fail, the problem may lie farther along the path; a Wi-Fi symbol is not proof the ISP works.
19. Worked Example H — What a modem does
The modem at an ISP boundary handles access-technology/protocol conversion for the subscribed internet connection. A modem is not simply ‘the thing that makes all wireless signals’.
20. Worked Example I — Network interface controller
A desktop connects to a network using a suitable network interface controller. A wired controller uses its physical interface and a wireless controller uses radio capability; both support communication via corresponding media.
21. Worked Example J — Client-server example
A library catalogue server provides a shared resource to student clients. Centralised management can be useful, but dependence on one service can create availability concerns if that server is down.
22. Worked Example K — Peer-to-peer example
Two authorised computers share selected files directly within a small group. This is a peer-to-peer interaction rather than one central dedicated server delivering all resources; permissions and access control still matter.
23. Worked Example L — Star layout
Several desktop Ethernet links converge on a central switch. This is a star topology at that level; it simplifies connections but the central switch can be a shared failure point.
24. Worked Example M — Mesh layout
Multiple network nodes have alternative connections instead of relying on only one path. A suitably designed mesh can offer alternative routes, but additional connections increase configuration and management complexity.
25. Worked Example N — Protocol as a shared rule
Two systems need compatible rules about communication format and meaning. A protocol is not a physical cable or a unique computer’s name; it describes how participants exchange and interpret data.
26. Worked Example O — One packet is corrupted
An error-detection mechanism identifies a damaged received packet. An automatic repeat request protocol can cause retransmission under defined conditions, rather than treating all transmitted data as automatically lost.
27. Worked Example P — Even parity
A byte of data has three 1-bits. To make the overall parity even, append a parity bit of 1, bringing the count of ones to four. This can detect a single bit flip but cannot detect every even-numbered pattern of flips.
28. Worked Example Q — A two-bit parity failure
Under even parity, flipping two bits can preserve an even total and evade a simple parity check. Therefore passing parity does not prove the message arrived without any corruption.
29. Worked Example R — Checksum concept
A sender computes a checksum over data according to an agreed rule and a receiver recomputes it. A mismatch can reveal corruption; a match does not guarantee detection of every possible alteration.
30. Worked Example S — Echo check
A sender transmits information and checks the receiver’s echoed response. A disagreement suggests communication error. Echoing is an approach to detection, not automatic correction of all problems.
31. Worked Example T — Automatic repeat request
If an expected packet or acknowledgement is missing or a corrupted packet is detected under the protocol, the sender may retransmit. Timeouts and acknowledgements are parts of a reliable communication strategy.
32. Worked Example U — MAC representation
A commonly represented 48-bit MAC identifier, such as 02:00:00:00:00:2A, is displayed as six groups of two hexadecimal digits. It is used for local-interface identification and switching concepts, not a public street address.
33. Worked Example V — IPv4 representation
An IPv4 address has 32 bits, typically written as four decimal values separated by dots, such as documentation address 192.0.2.7. Its role is network-level addressing and routing across IP networks.
34. Worked Example W — IPv6 representation
An IPv6 address has 128 bits, commonly written as eight hexadecimal groups; 2001:db8::1 is a documentation example. The larger address length is not evidence that a given internet connection is faster.
35. Worked Example X — Same LAN versus other LAN
A frame switched within one LAN relies on appropriate local MAC delivery, while traffic to a distant IP network is routed using IP destinations. One real transmission can involve both kinds of addressing at different stages.
36. Worked Example Y — Address permanence
A device may use a MAC address that can be locally administered or randomised, while IP addresses can change with network attachment or assignment. Do not assume either address uniquely identifies one human being permanently.
37. Worked Example Z — Troubleshoot a partial failure
Two local devices exchange data but a remote website cannot load. First distinguish the functioning local route from the external link and destination, then check authorised network and service evidence. Do not conclude that the whole LAN is broken.
38. When a Star Network Fails
If a particular client cable breaks, that node may disconnect while other clients on the central switch continue communicating. If the central switch itself fails, multiple connected devices can be affected. This difference illustrates why fault location matters more than simply naming a topology. Real systems may contain several nested stars rather than one textbook diagram.
39. Mesh Is About Alternative Paths, Not Guaranteed Performance
A mesh topology can connect nodes using multiple paths, potentially improving resilience if one link stops working. But interference, limited bandwidth, routing logic and equipment capabilities can still affect results. A star network with high-quality central equipment may outperform a poorly configured mesh in a particular setting; context determines the appropriate conclusion.
40. MAC and IP Are Not Competing Addressing Systems
MAC addresses help local frame delivery and network interface identification. IP addresses help route traffic beyond a particular LAN. When a laptop accesses a website, local and network-level addressing cooperate at different stages. Treating the two as mutually exclusive loses the layered logic of packet communication.
41. A Longer IP Address Is Not a Better Speed Test
IPv6 has 128 bits while IPv4 uses 32. This expands address representation; it does not itself prove lower latency, fewer outages or greater bandwidth. Those depend on the route and deployment conditions. A well-structured explanation compares address capacity and purpose without inventing performance claims.
42. A Firewall Belongs to the Security Layer of the Story
The K349 Module 4 also introduces firewalls, encryption and privacy. A firewall applies rules to permitted or blocked traffic; it does not replace switching, routing or a functioning network interface. A connection can fail because a rule denies traffic even when the physical link works. Diagnose access policy separately from physical connectivity.
43. Error Detection Is Not Error Correction
Parity, checksums and echo checks can identify some inconsistencies. Automatic repeat request obtains another transmission when conditions indicate a problem. A detected error may be handled by retransmission, not because the checksum alone reconstructs the correct original data. This distinction protects students from overclaiming a protocol’s capability.
44. Troubleshooting Should Follow the Route
Begin by checking whether the endpoint is powered and connected. Determine if the problem is local to one device, shared by the LAN or specific to remote access. Compare a second local service and a known external destination if you have authority. This reduces unnecessary disruptive changes and gives a rationale for each diagnostic question.
45. Privacy and Security Are Not Synonyms for Internet Connection
Security controls protect systems and data against improper access, tampering and other threats, whereas privacy concerns appropriate handling of personal information. Being connected to a LAN does not grant permission to view another pupil’s files. Network functionality and authorised access should be discussed separately.
46. Avoid Publishing Private Address and Password Data
Screenshots used for homework can reveal network names, access tokens or identifying device details. For exercises, use documentation-only addresses or invented values instead of copying a family’s real router settings. Safe learning demonstrates architecture with hypothetical diagrams and does not require exposure of credentials.
47. The Three-Pupil Network Diagnosis Workshop
Pupil A draws the LAN devices and media; pupil B labels local versus routed addresses and packet paths; pupil C proposes non-invasive evidence for isolating a fault. The tutor rotates responsibilities, then gives each learner an unfamiliar topology and requires a complete independent explanation. The group succeeds when no pupil needs another to supply all the arrows.
48. eduKate Depth–Load–Transfer Diagnostic
Depth: can the pupil explain why local printing might work during an ISP outage? Load: can they coordinate NIC, access point, switch, router, modem, MAC/IP addressing and a lost packet in one diagram? Transfer: can they describe a new mesh or star configuration and diagnose a different symptom independently? The Clementi three-pupil tutorial standard contributes the diagnose–repair–guided-practice–independent-check teaching method.
49. Parent and Tutor Questions
Ask: ‘Are the two endpoints on the same LAN?’, ‘Which device supplies the radio link?’, ‘Which forwards local frames?’, ‘Which routes between networks?’, ‘How would one detect a damaged packet?’, and ‘What observation would distinguish a router fault from a remote service failure?’ Require a defensible answer rather than one remembered diagram.
50. The Network Error Ledger
Error: Wi-Fi means internet. Repair: identify the ISP and routed path. Error: switches route every global internet connection. Repair: distinguish local MAC forwarding from IP routing. Error: parity guarantees no corruption. Repair: discuss undetected even-bit changes. Error: IPv6 is faster because its address is longer. Repair: separate address size from throughput. Error: a MAC address uniquely identifies one person. Repair: explain network interfaces and changing assignments.
51. Frequently Asked Questions
Can a LAN work without the internet?
Yes. Local devices can exchange authorised data through a LAN even if the external ISP link is unavailable.
Is a router the same as a switch?
They perform distinct functions: routing between IP networks versus local switching using MAC information, though one commercial device may combine both.
What does an access point do?
It connects compatible wireless clients into a LAN via a radio link.
What is the difference between IPv4 and IPv6?
IPv4 addresses are 32 bits, and IPv6 addresses are 128 bits, with different typical presentation formats.
Is a MAC address always permanent?
No. Locally administered and randomised MAC addresses exist; do not treat one as a permanent personal identity.
Do packets always arrive correctly?
No. They can be lost or corrupted; protocols may detect certain errors and arrange retransmission.
Does even parity catch every change?
No. Some changes affecting an even number of bits may escape simple parity detection.
Which topology is best?
It depends on the purpose, site, reliability, budget and design. A named topology does not guarantee the best performance in every case.
Does the 2027 syllabus include home network devices?
Yes. K349 Module 4.2 names a modem, router, switch, wireless access point and the correct connection roles.
Should a student alter a school router to demonstrate knowledge?
No. Demonstrate concepts using authorised classroom equipment or fictional topology diagrams.
52. Independent Depth and Transfer Tasks
Practice 1 — Draw a LAN and WAN distinction
Sketch two computers, a local switch and printer, then add a router, modem and distant server. State which traffic needs the wider link.
Practice 2 — Explain a combined home device
A router box provides Wi-Fi and Ethernet ports. Identify the separate access point, switching and routing functions even though they share one enclosure.
Practice 3 — Address identification
Classify 02:00:00:00:00:2A as a common MAC format, 192.0.2.7 as an IPv4 format and 2001:db8::1 as an IPv6 documentation address.
Practice 4 — Parity check
Give an even parity bit for the pattern 1011000, which contains three ones. State why two flipped bits might evade this check.
Practice 5 — Topology reasoning
Describe one possible point of failure in a star topology and one reason a mesh might retain connectivity after a single link failure.
Practice 6 — Packet retransmission
A packet fails its receiver’s checksum check. Explain detection and how an automatic repeat request arrangement can recover.
Practice 7 — Diagnose an unfamiliar outage
A tablet can reach a local network printer but not a remote class portal. Identify at least two plausible later-path causes and what permitted observations would distinguish them.
Practice 8 — Independent transfer
Design a fictional small learning hub with four computers and wireless tablets. Justify wired versus wireless segments, addressing roles and a privacy limitation.
53. Closing Principle
Computer networking is about endpoints, media, devices, addresses, packets and agreed rules. Diagnose the complete route, distinguish local communication from the internet and explain which mechanism could detect or recover a transmission problem. Real competence means transferring those ideas to a network the learner has not already memorised.
Related eduKateSG References and Official Sources
Continue from Reference 301 — Data Representation, Reference 311 — Computing Spreadsheets, Reference 52 — Internet Infrastructure and the MOE V3.0 reference library. The immutable Clementi teaching model explains the small-group diagnostic method. Official curriculum: 2027 K349 Module 4, Networking.
MOE V3.0 Reference Sequence Navigation
Previous reference: Reference 305. Browse the MOE V3.0 library to find adjacent subjects.