Computer Networks – Introductory Principles & Architecture
Course Outcomes
- Upon completion of the course, a student will be able to:
- Interpret the individual building blocks of a communication network and their overall architecture.
- Contrast circuit-switched and packet-switched networks and quantitatively analyze network performance.
- Identify, evaluate, and design error- and flow-control mechanisms that operate in the Data-Link layer.
- Perform IPv4 subnetting and analyze Network-layer performance when using different routing protocols.
- Compare congestion-control techniques and select suitable Transport-layer protocols (with security enhancements) for real-time applications.
- Module 1: Networking Principles & Layered Architecture – 6 hrs
- Module 2: Circuit & Packet Switching – 7 hrs
- Module 3: Data-Link Layer – 8 hrs
- Module 4: Network Layer – 8 hrs
- Module 5: Routing Protocols – 6 hrs
- Module 6: Transport Layer – 5 hrs
- Module 7: Application Layer – 3 hrs
- Module 8: Contemporary Issues – 2 hrs
Reference Material
- Textbook: Behrouz A. Forouzan, Data Communication and Networking (5th Ed., 2017, McGraw-Hill).
- Additional references:
- Kurose & Ross, Computer Networking – A Top-Down Approach (6th Ed.).
- Stallings, Data and Computer Communication (10th Ed.).
- Evaluation: CAT, written assignments, quizzes, Final Assessment Test (FAT).
Communication vs. Data Communication
- Communication: Transfer of information (voice, data, video) over wired or wireless media.
- Data Communication: Exchange of data between two devices using a transmission medium (e.g., copper wire, fiber). Key term telecommunication implies “communication at a distance.”
- End nodes can be PCs, printers, cameras; intermediate nodes include switches, routers, and the Internet cloud.
- Primary goal: resource sharing (files, printers, web access) through an agreed architecture (physical + logical design).
Fundamental Characteristics of a Data-Communication System
- Delivery
- Data must reach the correct destination device/user, and only that device/user.
- Accuracy
- Information must arrive unaltered; corrupted bits render data unusable.
- Timeliness
- Especially for voice/video, data must arrive when expected (real-time constraints).
- Jitter
- Variation in packet inter-arrival time; e.g., if video frames intended for 30\,\text{ms} spacing arrive sometimes at 30\,\text{ms} and other times at 40\,\text{ms}, perceptible quality degradation occurs.
Five Core Components of a Data-Communication Model
- Message – The actual information (text, images, audio, video).
- Sender – Device initiating transmission (PC, workstation, telephone, camera).
- Receiver – Endpoint that accepts the message (computer, TV, printer).
- Transmission Medium – Physical path: twisted-pair, coaxial, fiber-optic, free-space radio, etc.
- Protocol – Formal set of rules enabling interoperability. Without matching protocols, connected devices cannot communicate (analogy: French vs. Japanese speakers with no common language).
Data-Flow Directions
- Simplex – Unidirectional; e.g., keyboard → computer, monitor ← mainframe.
- Half-Duplex – Bidirectional but not simultaneous; devices alternate roles (walkie-talkies).
- Full-Duplex – Simultaneous two-way transfer; requires either two physical links or echo cancellation (telephone conversation).
Network Definition & Line Configurations
- Network: Collection of devices (nodes) connected by communication links.
- Point-to-Point
- One dedicated link between exactly two devices; entire capacity reserved (e.g., serial cable between two routers).
- Multipoint (Broadcast)
- Two or more devices share a single link.
- Spatial sharing – Multiple nodes transmit simultaneously via separate frequencies or codes.
- Temporal sharing – Nodes take turns (time-division multiplexing).
Protocol-Level Communication Paradigms
- Message Encoding → Transmission over Channel → Reception & possible retransmission modes.
- Unicast: One-to-one.
- Multicast: One-to-selected-many.
- Broadcast: One-to-all.
- Peer-to-Peer (P2P): Symmetric nodes acting both as client and server.
- Client/Server: Dedicated server provides services to multiple clients.
Physical & Logical Topologies
- Mesh
- Every node directly connected to every other node.
- Number of required links: \frac{n(n-1)}{2} for n devices.
- Pros: high redundancy, robust, simple routing.
- Cons: expensive cabling, complex I/O ports.
- Star
- All nodes attach to a central hub/switch.
- Pros: easy to add/remove nodes; single cable fault isolates one node.
- Cons: hub is single point of failure.
- Bus
- Single backbone cable with drop lines and taps.
- Pros: inexpensive, easy to install.
- Cons: difficult fault isolation; limited cable length and traffic.
- Ring
- Each node has exactly two neighbors; data travel in one logical direction.
- Repeaters can regenerate signal; failures require bypass mechanisms.
- Hybrid
- Combination (e.g., star backbone interconnecting several bus segments) offers scalability and fault tolerance.
Essential Network Components
- Nodes
- End nodes: PCs, printers, IP cameras.
- Intermediate nodes: hubs (legacy), switches, routers.
- Media
- Wired: copper, fiber, coax.
- Wireless: IR, RF, microwave, satellite.
- Services
- Email, WWW browsing, cloud storage (Google Drive), file sharing, online gaming, instant messaging.
- Ethernet Straight-Through Cable: Connect dissimilar devices (PC ↔ switch).
- Ethernet Crossover Cable: Connect similar devices (switch ↔ switch, PC ↔ PC without hub).
- Fiber-Optic Cable (FOC)
- Immune to EMI, extremely high bandwidth, long distance.
- Coaxial Cable – Legacy LANs, cable TV.
- USB Cable – Short-range peripheral connectivity.
- Infrared (IR) – TV remotes, short-range line-of-sight links.
- Radio Waves – Bluetooth, Wi-Fi (2.4 GHz / 5 GHz ISM bands).
- Microwaves – Cellular backhaul, point-to-point links.
- Satellite – GPS, global Internet backhaul, broadcast TV.
Classification of Computer Networks
- LAN (Local Area Network)
- Coverage: room → campus; wired Ethernet (hub/switch) or Wi-Fi.
- MAN (Metropolitan Area Network)
- Connects multiple LANs within a city (e.g., municipal fiber loop).
- WAN (Wide Area Network)
- Spans large geographic areas; relies on telecom carriers.
- Internet
- Public, global-scale WAN interconnecting heterogeneous networks.
Addressing in Computer Networks
IPv4 Address
- Logical, location-dependent identifier.
- 32-bit value represented as four octets in dotted-decimal form (e.g., 192.168.0.1).
- Range 0.0.0.0 to 255.255.255.255.
- Assigned manually (static) or dynamically (DHCP).
IPv6 Address (mentioned indirectly)
- 128-bit logical address, represented in hexadecimal colon-separated notation; expansion beyond transcript scope.
MAC Address
- Physical, hardware-burned identifier on NIC.
- 48 bits, expressed in hexadecimal (e.g., 00!:!1A!:!2B!:!3C!:!4D!:!5E).
- Globally unique; not tied to device location.
Port Number
- 16-bit value 0!\text{–}!65535 identifying a process/service endpoint within a host.
- Assigned and managed by OS; visible via tools such as resmon (Resource Monitor).
Conceptual & Practical Connections
- Course maps directly to OSI/TCP-IP layered principles: physical media → protocols → addressing.
- Real-time multimedia stresses timeliness & jitter, motivating QoS and congestion-control mechanisms (Modules 5 & 6).
- Switching paradigms (Module 2) build on the physical/logical topologies introduced here.
- Error detection (Module 3) ensures the accuracy characteristic, implementing CRC, Hamming, etc.
- IPv4 subnetting (Module 4) refines the addressing concepts introduced (Classful vs. Classless).
- Ethical & practical implication: ensuring global uniqueness of MAC addresses prevents identity collision; proper protocol adherence ensures interoperability between diverse vendors.