Unit I — Data Communication Basics & Network Topologies
Data communication is the exchange of data between two devices via a transmission medium. Effective data communication must satisfy four characteristics: delivery (to the correct destination), accuracy (error-free), timeliness (within acceptable delay), and jitter (consistent delay variation for multimedia streams).
---
1. Five Components of Data Communication
| Component | Role | Example |
|---|---|---|
| Message | The data or information being communicated | Text, audio, video, image, binary file |
| Sender | The device that originates the message | Computer, telephone, camera, IoT sensor |
| Receiver | The device that receives the message | Computer, printer, speaker, smart display |
| Transmission Medium | Physical path from sender to receiver | Copper wire, fibre optic cable, air (Wi-Fi) |
| Protocol | Rules governing how communication takes place | TCP/IP, HTTP, IEEE 802.11, Bluetooth |
Exam Tip: These five components appear in 2–3 mark questions every year. Mnemonic: "Many Students Read Material Patiently" — Message, Sender, Receiver, Medium, Protocol.
---
2. Data Flow (Transmission) Modes
| Mode | Direction of Flow | Example Use Case |
|---|---|---|
| Simplex | One direction only; receiver cannot transmit back | Keyboard → CPU, TV broadcast, sensor to server |
| Half-Duplex | Both directions, but not at the same time | Walkie-talkie, CB radio, older WLAN |
| Full-Duplex | Both directions simultaneously | Telephone call, modern Ethernet, video call |
---
3. Line Configuration
- Point-to-Point: A dedicated link connects exactly two devices. Full bandwidth is available to both devices. Higher security (no eavesdropping from other nodes). Example: leased-line WAN link between two offices.
- Multipoint (Multi-drop): A single link is shared by three or more devices. All devices share the same bandwidth. Channel access protocols (CSMA/CD, token-passing, polling) manage transmission order. Example: coaxial bus LAN.
---
4. Network Topologies
| Topology | Shape | Advantages | Disadvantages | Example Use |
|---|---|---|---|---|
| Bus | Single linear backbone; all nodes connect to it | Low cost, simple installation, less cable | Cable fault stops entire network; collision-prone; hard to troubleshoot | Early Ethernet 10Base2, coaxial LANs |
| Star | Central hub/switch; each node has its own link | Fault isolation (one link fails, others work); easy to add nodes; simple management | Central device failure stops all communication; more cable required | Modern office LANs, home Wi-Fi |
| Ring | Nodes connected in a closed loop | Equal access; predictable performance; no collisions with token passing | Single ring break can stop network; difficult to add/remove devices | FDDI, Token Ring, SONET rings |
| Mesh | Every node directly connected to every other node | Highest fault tolerance; dedicated paths; no single point of failure | Very expensive: n(n-1)/2 connections for n nodes; complex installation and management | WAN backbone, military networks |
| Hybrid | Combination of two or more topologies | Flexible, scalable, optimised for different areas | Complex design and management | Enterprise networks (Star-Bus, Star-Ring) |
Study Deep: Topology Selection Factors
- Bus was the original Ethernet topology (10Base2) but is obsolete. A single cable cut stops the entire network, making troubleshooting time-consuming.
- Star dominates today because switches provide fault isolation — only the device with the broken link is affected. Adding or removing nodes requires no network downtime.
- Mesh is used in WANs because redundant paths prevent any single link failure from disrupting traffic. The internet backbone is a partial mesh.
- Ring topology uses a circulating token (Token Ring, FDDI) to eliminate collisions. Dual ring (FDDI) provides fault tolerance: if one ring breaks, traffic routes on the other.
- Hybrid topologies are standard in practice. A campus network is typically Star-Bus: departments use star topology (switches), connected by a high-speed backbone (ring or bus).
---
5. Distributed Processing
Distributed processing splits computing tasks across multiple computers:
- Security: Data can remain at the local site; central server breach does not expose everything
- Reliability: Failure of one node does not stop the whole system
- Efficiency: Users access local resources quickly, reducing wide-area traffic
- Scalability: New nodes can be added without redesigning the whole infrastructure
---
6. Categories of Networks
| Category | Full Name | Coverage Area | Speed | Example |
|---|---|---|---|---|
| PAN | Personal Area Network | ~10 metres | Moderate | Bluetooth headset, smartwatch, USB |
| LAN | Local Area Network | Building or campus | 1–100 Gbps | Office Ethernet, home Wi-Fi |
| MAN | Metropolitan Area Network | City / metro area | 10 Mbps–10 Gbps | Cable TV network, city Wi-Fi mesh |
| WAN | Wide Area Network | Country / global | Varies | Internet, private leased lines |
Exam Tip: Know the distinguishing characteristic of each network type: PAN (Bluetooth/personal devices), LAN (high speed, private, building), MAN (city-scale), WAN (global/internet). The internet is the world's largest WAN.
---
Study Deep: Internet, Intranet, and Extranet
- Internet: The global public network of networks, using TCP/IP protocols. No single owner or central authority.
- Intranet: A private internal network using internet protocols (HTTP, DNS, SMTP) but accessible only to organisation members. Examples: company internal websites, SharePoint portals.
- Extranet: An intranet that allows controlled access to authorised external users (partners, customers, suppliers). Usually secured using VPN tunnels or dedicated links.