Computer Networks

A Computer Network is defined as an interconnected collection of autonomous computing devices (computers, servers, smartphones, network switches, IoT devices) that communicate with one another using standardized protocols over physical or wireless transmission media.

The foundational design goals and business drivers of computer networks include:

• Resource Sharing: Making expensive hardware (high-capacity printers, optical storage arrays, GPUs) and software programs available to any network user regardless of physical location.
• High Reliability & Redundancy: Replicating files and databases across multiple redundant machines. If one server experiences hardware failure, alternate machines seamlessly assume processing duties.
• Cost Reduction: Small interconnected workstations communicating with centralized server clusters provide vastly superior price-to-performance ratios compared to monolithic mainframes.
• High-Speed Communication: Providing near-instantaneous global messaging (email, video conferencing, real-time collaboration).
• Distributed Processing: Distributing compute-heavy tasks across multiple cooperative worker nodes in parallel.

Networks are architecturally structured into two fundamental relationship models:

Computer networks are universally classified by the physical geographic diameter they encompass:

1. PAN (Personal Area Network): Spans a radius of $1$ to $10$ meters centered around an individual. Uses short-range technologies like Bluetooth (IEEE 802.15.1), Zigbee, and USB. Connects wireless earphones, smartwatches, and laptops.
2. LAN (Local Area Network): Confined to a geographically restricted area such as a single room, office floor, school laboratory, or university campus (up to a few kilometers). Characteristics: privately owned, ultra-high data rates ($100\text{ Mbps}$ to $10\text{ Gbps}$ Ethernet / Wi-Fi), and extremely low bit-error rates.
3. MAN (Metropolitan Area Network): Spans an entire city or municipal district (typically $10$ to $50\text{ km}$). Classic examples include city cable television networks, municipal traffic camera networks, and inter-branch bank networks across a metropolis.
4. WAN (Wide Area Network): Spans vast geographical regions crossing country borders, continents, or the entire globe. Characteristics: utilizes public telecommunications backbones, leased undersea fiber cables, and communication satellites; operates at lower speeds and higher error rates compared to LANs. The Internet is the quintessential global WAN.

The physical path connecting the transmitter and the receiver across which electromagnetic or optical energy propagates is called the Transmission Medium. It is categorized into Guided (Wired / Bounded) and Unguided (Wireless / Unbounded) channels.

1. Twisted Pair Cable: Consists of pairs of insulated copper wires twisted spirally around each other. Twisting mathematically cancels out electromagnetic interference (EMI) and radio frequency interference (RFI) from external sources and suppresses crosstalk between adjacent pairs.

• Unshielded Twisted Pair (UTP): Inexpensive, lightweight, flexible, and standard in modern LANs. Categories include Cat 5e ($100\text{ MHz}$, up to $1\text{ Gbps}$) and Cat 6 ($250\text{ MHz}$, up to $10\text{ Gbps}$ up to 55m). Terminated with standard 8-pin RJ-45 connectors. Maximum run without repeaters: $100\text{ meters}$.
• Shielded Twisted Pair (STP): Features a braided metal foil shield encasing each pair or all pairs, providing superior noise immunity in heavy industrial environments, but is bulkier, stiffer, and requires proper electrical grounding.

2. Coaxial Cable: Features a central solid copper conductor encased in a thick dielectric insulator, surrounded by a cylindrical woven wire mesh shield, and protected by an outer plastic jacket.

• The concentric shielding provides exceptional noise immunity and high bandwidth compared to twisted pair.
• Types: Baseband (50-ohm, transmits single digital signal) and Broadband (75-ohm, frequency division multiplexed for cable TV and high-speed cable modems). Terminated using BNC connectors (Bayonet Neill–Concelman).

3. Optical Fiber Cable (OFC): Transmits data as pulses of light through ultra-pure cylindrical strands of silica glass or plastic. An optical fiber consists of three concentric layers: (1) Core (ultra-pure glass with refractive index $n_1$), (2) Cladding (surrounding glass layer with lower refractive index $n_2 < n_1$), and (3) Buffer / Jacket (protective polymer coating).

• Key Advantages: Astronomical bandwidth (terabits per second), complete immunity to Electromagnetic Interference (EMI), lightning strikes, and radio snooping (zero crosstalk), extremely low signal attenuation enabling repeater spacings of over $80\text{ km}$.
• Fiber Types: Single-Mode Fiber (SMF) (narrow $8-10\,\mu\text{m}$ core, single laser beam path, zero modal dispersion, long-haul backbones) vs Multi-Mode Fiber (MMF) ($50-62.5\,\mu\text{m}$ core, LED source, multiple light paths, short-range campus backbones).

Unguided media broadcast electromagnetic signals through atmospheric air or vacuum without physical conduits:

• Simplex: Strictly unidirectional communication. One device transmits exclusively; the other receives exclusively (e.g., keyboard to CPU, broadcast television).
• Half-Duplex: Bidirectional communication, but only one direction at a time. When one transmits, the other must listen (e.g., Walkie-Talkie).
• Full-Duplex: Simultaneous bidirectional communication in both directions simultaneously (e.g., switched Ethernet, telephone conversations).

A Network Topology defines the geometric layout or physical and logical arrangement of nodes, workstations, and connecting links within a computer network.

1. Bus Topology (Linear Bus): All nodes connect directly to a single continuous central coaxial or twisted-pair cable called the backbone. At each physical end of the backbone, a resistive Terminator (typically 50-ohm) absorbs traveling electrical signals to prevent signal reflection.

• Advantages: Inexpensive, minimal cabling, easy to set up for small temporary networks.
• Disadvantages: A single break in the central backbone cable halts the entire network; difficult fault diagnosis; heavy packet collisions as node count increases.

2. Star Topology: Every node connects independently to a central concentrator device (a multi-port Hub or Switch) via a dedicated point-to-point drop cable. This is the universal standard for modern Ethernet LANs.

• Advantages: High fault tolerance—failure of a single cable or node isolates only that node while the rest of the network operates normally; easy installation, reconfiguration, and centralized traffic monitoring.
• Disadvantages: Central dependency—if the central switch fails, the entire network collapses; requires significantly more total cable footage than bus topology.

3. Ring Topology: Each workstation is connected to exactly two neighboring nodes, forming an unbroken closed circular loop. Data circulates unidirectionally in the form of a special electronic bit-pattern called a Token.

• Advantages: Equal channel access; no packet collisions since only the token-holding node transmits; performs well under heavy traffic load.
• Disadvantages: In a standard single ring, a failure in any single workstation or cable segment disables the entire loop; adding or removing nodes disrupts communication. (Solved by Dual-Ring FDDI architecture).

4. Mesh Topology: Every device has a dedicated point-to-point physical link to every other device in the network (Full Mesh), or to multiple selected devices (Partial Mesh).

• Advantages: Maximum fault tolerance, absolute redundancy, zero traffic congestion between pairs, robust privacy and security.
• Disadvantages: Prohibitive cabling cost, excessive physical I/O ports required per device, complex installation and maintenance.

5. Tree Topology (Hierarchical): An extension of star topology where multiple star-configured clusters are interconnected via a central bus backbone cable, organized in root-parent-child hierarchies.

Computer networking problems are too complex to solve with a monolithic software program. Instead, networking relies on Layered Architecture, where complex communication tasks are decomposed into sequential, modular tiers. Each layer provides specific services to the layer above it, while concealing the implementation details of the layers below it.

Formulated in 1984 by the International Organization for Standardization (ISO), the Open Systems Interconnection (OSI) model defines seven distinct functional layers:

1. Layer 1: Physical Layer (PDU: Bits): Responsible for transmitting raw unstructured bit streams (0s and 1s) over physical communication channels. Defines mechanical, electrical, functional, and procedural specifications: pinouts, voltages, data rates, bit synchronization, and cable media.
2. Layer 2: Data Link Layer (PDU: Frames): Provides error-free point-to-point node transfer across the physical link. Packages bits into Frames, adds physical source and destination MAC Addresses, performs flow control (stop-and-wait, sliding window), and detects transmission errors using Cyclic Redundancy Check (CRC). Sub-divided into LLC (Logical Link Control) and MAC (Media Access Control).
3. Layer 3: Network Layer (PDU: Packets): Responsible for end-to-end routing and forwarding of packets across diverse internetworks. Assigns logical IP Addresses (IPv4/IPv6), calculates shortest paths using routing algorithms, handles packet fragmentation and reassembly, and manages network congestion.
4. Layer 4: Transport Layer (PDU: Segments): Responsible for true end-to-end process-to-process communication. Manages source and destination Port Numbers (addressing specific running software processes), breaks application messages into segments, controls transmission flow, and guarantees reliability through acknowledgments and retransmissions (TCP) or fast delivery (UDP).
5. Layer 5: Session Layer (PDU: Data): Establishes, maintains, coordinates, and terminates communication sessions between software applications. Controls dialog discipline (simplex, half-duplex, full-duplex) and inserts synchronization checkpoints into long file transfers to resume from interruptions without restarting.
6. Layer 6: Presentation Layer (PDU: Data): Acts as the network translator. Resolves syntax and semantics differences between disparate computers. Core duties: (1) Data Translation (ASCII to EBCDIC), (2) Data Compression (reducing bandwidth consumption), and (3) Cryptographic Encryption and Decryption (SSL/TLS security).
7. Layer 7: Application Layer (PDU: Data): The highest layer, providing direct network service interfaces to user software applications. Contains protocols like HTTP/HTTPS (web browsing), FTP (file transfer), SMTP (electronic mail), and DNS (domain resolution).

While OSI is a comprehensive theoretical reference model, the practical global Internet operates on the TCP/IP Protocol Suite (developed by DARPA). It consolidates networking into four operational layers:

An Internet Protocol version 4 (IPv4) address is a 32-bit logical address uniquely identifying every host interface connected to an IP network. For human readability, the 32 binary bits are divided into four 8-bit groups called Octets, each converted to a decimal integer ($0$ to $255$) separated by periods—known as Dotted-Decimal Notation (e.g., 192.168.1.1).

Every IP address is hierarchically partitioned into two essential components:

• Network ID (NetID): Identifies the specific physical or logical network to which the host belongs. All hosts on the same physical broadcast segment share identical NetID bits.
• Host ID (HostID): Uniquely identifies a specific host machine or network interface card on that network.

The original Internet architecture established five address classes distinguished by their leading binary bits:

To mitigate the rapid depletion of IPv4 addresses, the IETF reserved three private address blocks that are never routed over the public Internet:

• 10.0.0.0 to 10.255.255.255 (10.0.0.0/8 - Single Class A network, 16 million private hosts).
• 172.16.0.0 to 172.31.255.255 (172.16.0.0/12 - 16 contiguous Class B networks, 1 million private hosts).
• 192.168.0.0 to 192.168.255.255 (192.168.0.0/16 - 256 contiguous Class C networks, 65,536 private hosts; standard in home Wi-Fi routers).

NAT (Network Address Translation): A border router technology that enables hundreds of private network devices sharing private IP addresses (e.g., 192.168.1.X) to access the public Internet simultaneously using a single public routable IP address.

IPv6 (Internet Protocol version 6) was engineered to overcome IPv4 address exhaustion. It uses a 128-bit address written as 8 groups of 4 hexadecimal digits separated by colons (e.g., 2001:0db8:85a3:0000:0000:8a2e:0370:7334). It provides $2^{128} \approx 3.4 \times 10^{38}$ distinct addresses, eliminating the need for NAT and integrating native IPSec security.

When data travels through an interconnected network of multiple nodes, switching techniques determine how transmission paths are allocated:

• HTTP (HyperText Transfer Protocol - Port 80): The foundational protocol of the World Wide Web. A stateless, request-response protocol running over TCP used to transfer HTML documents, images, and multimedia.
• HTTPS (HTTP Secure - Port 443): Transmits HTTP over an encrypted cryptographic tunnel using TLS/SSL (Transport Layer Security / Secure Sockets Layer). Protects against eavesdropping, session hijacking, and man-in-the-middle attacks.
• FTP (File Transfer Protocol - Ports 20 & 21): Protocol for transferring large files between client and server. Uses two separate TCP connections: Port 21 for control commands and Port 20 for data streaming.
• SMTP (Simple Mail Transfer Protocol - Port 25): A text-based protocol used by mail user agents to push outgoing emails to mail servers and between mail transfer agents.
• POP3 (Post Office Protocol v3 - Port 110) & IMAP (Port 143): Protocols used by client software to retrieve emails from a mail server. POP3 downloads emails and deletes them from the server; IMAP synchronizes emails across multiple devices while keeping mail stored on the server.
• DNS (Domain Name System - Port 53): The 'phonebook' of the Internet. Resolves human-readable domain names (e.g., targetexams.com) into 32-bit IP addresses (e.g., 77.237.235.240) using a hierarchical distributed tree of Root, TLD (.com, .org, .in), and Authoritative nameservers.
• DHCP (Dynamic Host Configuration Protocol - Ports 67 & 68): Automatically configures network host devices by dynamically leasing IP addresses, subnet masks, default gateway routers, and DNS server addresses upon connection.

Invented in 1989 by Tim Berners-Lee at CERN, the Web is an interconnected information space of hyperlinked documents accessed via the Internet:

• URL (Uniform Resource Locator): The global address of a web resource. Anatomy: https://www.targetexams.com:443/syllabus/cs?year=2026#sec1 comprises: Protocol (https), Domain Host (www.targetexams.com), Port (443), Path (/syllabus/cs), Query String (?year=2026), and Fragment Identifier (#sec1).
• Web Browser vs Web Server: A web browser (Chrome, Firefox, Safari) is a client application that renders HTML/CSS/JavaScript. A web server (Apache, Nginx, Microsoft IIS) is an application listening on port 80/443 that serves files upon receiving HTTP requests.
• Static vs Dynamic Web Pages: Static pages display identical pre-written HTML content to all visitors. Dynamic pages are rendered on-the-fly by server-side scripts (PHP, Python, Node.js) querying backend databases based on user input.