A Practical Guide to Computer Networks: How Data Moves & How Networks Are Built

Why Do We Need Computer Networks?

Imagine a world where every computer operates in isolation. If you wanted to send an email, you would have to physically carry a storage device (like a USB) to another computer. If you wanted to access a website, you’d have to request a copy of it on a disk and wait for it to be delivered to you. This is how computers would function without networks—slow, inefficient, and disconnected.

A computer network solves this by allowing multiple devices to communicate, share resources, and access information in real-time, no matter where they are located. Networks are the foundation of modern digital communication, enabling everything from emails to video calls, cloud computing and online gaming.

To understand how networks work, we first need to explore how data travels from one device to another, and what processes are involved in ensuring that communication is smooth, reliable, and secure.

---

Data Transmission: How Information Travels in a Network

Analog vs. Digital Signals: Why Do Computers Work with Digital Data?

Data in the real world exists in analog form. For example:

• Your voice when talking on the phone is analog—it is a continuous wave of sound.
• A song recorded on a vinyl record is also analog—the grooves on the record store sound waves in a continuous pattern.

But computers don’t understand analog data. They work with binary digits (0s and 1s), meaning they can only process digital information. This is why all real-world data (sound, images, video, etc) needs to be converted from analog to digital before a computer can process it.

This conversion happens through a process called digitisation, using devices like:

• Microphones (convert sound waves into digital signals)
• Scanners (convert physical documents into digital copies)
• Cameras (convert real-world images into digital format)

Similarly, when a computer sends digital data to an analog system (e.g., broadcasting radio signals), it must be converted back into analog form using digital-to-analog conversion (DAC).

Analog Signals

Analog signals are continuous waveforms that vary over time. These signals can have an infinite number of values within a given range. Examples: Human voice, audio signals, radio waves, etc.

Analog signals can be represented as sine waves, with amplitude, frequency, and phase as their key characteristics.

In communication, analog signals are transmitted over mediums like telephone lines or airwaves for radio broadcasting.

Digital Signals

Digital signals are discrete, meaning they consist of a series of fixed values (typically 0 and 1, corresponding to binary bits). These signals are represented as square waves with only two levels: high (1) and low (0).

Digital signals are commonly used in modern data transmission systems, such as computer networks, digital telephony, and the internet.

Key Differences

Continuity

Analog signals are continuous, while digital signals are discrete.

Noise

Analog signals are more susceptible to noise and signal degradation over long distances, while digital signals are less prone to noise because they only need to distinguish between two states (0 and 1).

Bandwidth Efficiency

Digital signals can be more bandwidth-efficient and easier to process in modern systems.

Converting Analog Signals to Digital

As we established earlier, every sound we hear, every image we see—it all starts as an analog signal in the real world. But computers don’t process smooth, continuous waves. They work with discrete values—ones and zeros. That’s why we need Analog-to-Digital Conversion (ADC) to transform the real world into something digital systems can understand.

How it works:

The conversion process happens in three key steps: Sampling, Quantisation, and Encoding. Each step ensures the analog signal is accurately represented in digital form.

1. Sampling — Capturing the signal at intervals:
   • Imagine you’re trying to recreate a song by writing down a note every second. The more frequently you note down the sounds, the more accurate your reproduction.
   • In ADC, the continuous analog signal is sampled at regular intervals to capture its key points.
   • The sampling rate (measured in Hertz) defines how often this happens.
   • Nyquist Theorem: To avoid losing information, the sampling rate must be at least twice the highest frequency of the signal.
     • Example: Human hearing ranges up to 20 kHz, so CDs use a 44.1 kHz sampling rate to ensure high-quality sound reproduction.
2. Quantisation — Assigning values to the samples
   • Think of a painter recreating a sunset but only having a limited set of colours. Some shades will be lost, but they’ll try to match as closely as possible.
   • In digital conversion, each sampled point is approximated to the nearest level within a set range of values, introducing a minor error called quantisation noise.
   • The number of possible levels depends on the bit depth:
     • 8-bit depth → 256 levels (lower accuracy)
     • 16-bit depth → 65,536 levels (used in CDs for better audio detail)
3. Encoding — Storing the digital representation
   • Once quantised, the values are converted into binary code (ones and zeros).
   • A 16-bit ADC means each sample is stored as a 16-bit binary number, ensuring high precision.
   • This binary data can now be processed, stored, or transmitted through digital systems.

Beyond Audio — ADC in Videos

• In video processing, ADC doesn’t just capture sound but also brightness and colour data for each frame.
• Higher resolution and bit depth lead to more detailed images, just like higher sampling rates create better sound quality.

---

How is Data Transmitted in a Network?

Every time you send an email, stream a video, or load a webpage, data moves through an intricate process behind the scenes. But how does this transmission actually happen?

Step-by-Step: How Data Travels

Whenever data moves from one device to another, it follows a structured journey:

1. Creation: A user types an email, streams a video, or makes a request to a server.
2. Conversion: If the data starts as an analog signal (like voice), it’s converted into a digital format.
3. Segmentation: Large chunks of data are broker into smaller packets to optimise transmission.
4. Transmission: These packets travel through wired or wireless communication channels.
5. Reassembly: The receiving device puts the packets back together and display the original data.

This process happens so quickly that we rarely think about it—but its efficiency depends on how data is transmitted.

Data Transmission Techniques

Data doesn’t just move randomly—it follows specific transmission methods that dictate who can send data, when, and how fast.

💡 Why Full-Duplex?

Most modern networks use full-duplex communication because it allows seamless data flow in both directions, making interactions faster and more efficient.

Serial vs. Parallel Transmission: The Speed Factor

When data moves from one device to another, it can be transmitted in two key ways:

Serial Transmission — One Bit at a time (Moderns Standard)

In serial transmission, data moves one bit at a time over a single channel. While this may seem slower, it’s actually more efficient and reliable over long distances.

• USB (Universal Serial Bus) — Modern devices use USB for high-speed data transfer.
• PCI Express (PCIe) — Even computer expansion cards now use serial communication instead of older parallel connections.
• Ethernet Networking — Internet data packets are transmitted serially, allowing faster, more stable network communication.

Parallel Transmission — Multiple Bits at Once (Older Method)

In parallel transmission, multiple bits are sent simultaneously over multiple channels. While this can increase speed, it’s more prone to interference and timing errors over long distances.

• Parallel ATA (PATA) Hard Drives — Once common, now replaced by Serial ATA (SATA) for better performance.
• Older Computer Buses — Early motherboards used parallel connections before shifting to serial-based architectures.

💡 Why Serial Over Parallel?

While parallel might seem faster, serial transmission scales better with increasing data speeds and distances, which is why modern technology has moved in this direction.

Why This Matters?

Understanding how data moves helps us appreciate the backbone of modern communication. Whether you’re sending a message, watching a video, or making a call, data transmission techniques share the speed, efficiency, and reliability of our digital world.

The next time you plug in a USB, join a Zoom call, or browse the web—you’ll know exactly what’s happening behind the scenes.

---

How Do We Organise Communication in Networks? The OSI Model

Now that we understand how data moves, the next question is:

How do different devices—laptops, smartphones, routers—communicate seamlessly across networks?

If every device had its own way of sending and interpreting data, the internet would be chaos. To solve this, networking follows structured principles—and that’s where the OSI Model (Open Systems Interconnection Model) comes in.

Why Was the OSI Model Created?

Think of sending a physical letter through the postal service:

1. You write the message (data creation).
2. You put it in an envelope and add an address (formatting & addressing).
3. The postal system ensures it reaches the right location (routing).
4. Your friend receives and reads the letter (data interpretation).

💡 Computer networks work the same way!

The OSI Model was developed as a conceptual framework to standardise communication across devices and networks. The OSI Model breaks communication into seven structured layers, ensuring devices from different manufacturers can interact smoothly.

Although modern networks primarily use TCP/IP, the OSI model remains a valuable reference for understanding how data flows through different layers.

The 7 Layers of the OSI Model (Explained Simply)

Each layer has a specific job, like a team of specialists working together. Data moves from top to bottom when sent and from bottom to top when received.

Layer	What It Does	Real-World Example
7. Application	Directly interacts with the user.	Web browsers, email apps
6. Presentation	Translates, encrypts, and compresses data.	SSL encryption, JPEG file formats
5. Session	Manages ongoing communication sessions.	Logging into a website
4. Transport	Ensures complete and error-free data transfer.	TCP, UDP (Used for video calls, gaming)
3. Network	Finds the best path for data to travel across networks.	IP addresses, routers
2. Data Link	Groups bits into frames & manages direct device-to-device communication.	MAC addresses, Ethernet
1. Physical	Transfer raw 0s and 1s over cables or wireless signals.	Wi-Fi signals, fibre optic cables, ethernet cables

How the OSI Model Helps Us Understand Networking

Even though the internet doesn’t strictly use the OSI model, it’s still a powerful tool for learning how networks function. For example:

• Troubleshooting Network Issues — IT professionals use OSI layers to diagnose problems (e.g., is it a physical issue or a routing issue?).
• Understanding TCP/IP — While the internet follows the simplified TCP/IP model, its principles align with the Transport, Network, and Data Link layers of the OSI model.
• Security & Encryption — The OSI model helps us understand where encryption happens (Presentation Layer) and how secure connections are established (Session Layer).

1. Physical Layer (Layer 1)
   • Function: Deals with the transmission of raw bits (0s and 1s) over a physical medium.
   • Services Provided:
     • Defines hardware specifications (cables, switches, hubs, NICs).
     • Controls data encoding and signal transmission (voltage levels, modulation).
     • Manages physical topology (bus, star, ring, mesh).
     • Determines transmission modes (simplex, half-duplex, full-duplex).
   • Example Technologies: Ethernet cables, Fibre optics, Bluetooth, Wi-Fi (802.11).
2. Data Link Layer (Layer 2)
   • Function: Provide error-free transmission of frames between adjacent nodes.
   • Services Provided:
     • Framing: Organises bits into data frames.
     • MAC Addressing: Uses Media Access Control (MAC) addresses to identify devices on a local network.
     • Error Detection and Correction: Uses techniques like Cyclic Redundancy Check (CRC).
     • Flow Control: Regulates data transmission rate between sender and receiver.
     • Media Access Control: Determines how devices access the shared medium (CSMA/CD in Ethernet, CSMA/CA in Wi-Fi).
   • Example Technologies: Ethernet, MAC addresses, VLANs, Wi-Fi (802.11), PPP.
3. Network Layer (Layer 3)
   • Function: Responsible for routing and forwarding data packets between different networks.
   • Services Provided:
     • Logical Addressing: Uses IP addresses for device identification across networks.
     • Routing: Determines the best path for data transmission.
     • Packet Fragmentation and Reassembly: Handles large packets that need to be split for transmission.
     • Congestion Control: Manages network traffic to prevent overload.
   • Example Protocols: IPv4, IPv6, ICMP (ping), ARP (Address Resolution Protocol), RIP, OSPF, BGP.
4. Transport Layer (Layer 4)
   • Function: Provides end-to-end communication and ensures reliable data delivery.
   • Services Provided:
     • Segmentation and Reassembly: Breaks data into segments and reassembles them at the destination.
     • Flow Control: Prevents overwhelming the receiver.
     • Error Detection and Recovery: Ensures correct data delivery using retransmission mechanisms.
     • Connection Management:
       • Connection-Oriented (TCP): Reliable, ensures order and correctness.
       • Connectionless (UDP): Faster, but with no guarantee of delivery order.
   • Example Protocols: TCP (Transmission Control Protocol), UDP (User Datagram Protocol), SCTP.
5. Session Layer (Layer 5)
   • Function: Manages and controls sessions between applications.
   • Services Provided:
     • Session Establishment, Maintenance, and Termination: Ensures communication between applications.
     • Synchronization: Uses checkpoints to resume interrupted data transfer.
     • Dialog Control: Enables bidirectional or unidirectional communication.
   • Example Technologies: NetBIOS, RPC (Remote Procedure Call), PPTP (Point-to-Point Tunnelling Protocol).
6. Presentation Layer (Layer 6)
   • Function: Translates data into a format readable by the application layer.
   • Services Provided:
     • Data Translation: Converts data formats (ASCII to EBCDIC, big-endian to little-endian).
     • Encryption and Decryption: Ensures secure communication (SSL/TLS).
     • Compression: Reduces file sizes for efficient transmission (JPEG, GIF, MP3).
   • Example Technologies: SSL/TLS, ASCII, Unicode, JPEG, MPEG, GIF.
7. Application Layer (Layer 7)
   • Function: Provides user-facing services and enables communication between applications.
   • Services Provided:
     • Network Access for Applications: Email, file transfer, web browsing, etc.
     • File Transfer and Management: FTP and cloud storage interactions.
     • Email Services: SMTP for sending, POP3/IMAP for receiving.
     • Remote Access Services: SSH, Telnet.
   • Example Protocols: HTTP, HTTPS, FTP, SMTP, POP3, IMAP, DNS, Telnet, SNMP.

How Data Theoretically Move on the Internet (With the OSI Model)

When an email is sent from one computer to another, the data flows through the seven layers of the OSI model, where each layer adds its own functionality before reaching the recipient (This is referred to as Data Encapsulation).

💡 OSI Data Encapsulation refers to the lower layer receiving packets from the higher layer, adding its own header and trailer (optional), and passing the packet onto the next lower layer.

The receiver will run the process at each layer performing the actions specified in the header/trailer of the sender, remove the header/trailer, and passes on the data packets to the layer immediately above. This process is known as data decapsulation.

Example Scenario: Sending an Email

A user, Alice, sends an email to Bob via SMTP (Simple Mail Transfer Protocol) from her email client (e.g., Outlook, Gmail) on Computer A. Bob receives the email on Computer B via POP3 or IMAP.

Step 1: Application Layer (Layer 7)

• Process Initiation: Alice composes an email in Gmail or Outlook and clicks “Send”.
• Protocol User: SMTP (Simple Mail Transfer Protocol) which handles outgoing emails.
• Data Format: The email is in a human-readable format (text, attachments, images, etc).
• Service Provided: SMTP on Alice’s computer establishes a connection with the mail server (e.g., smtp.gmail.com).

Data is then passed to the Presentation Layer.

Step 2: Presentation Layer (Layer 6)

• Data Translation: Converts the email text and attachments into a standard format (e.g., ASCII or Unicode).
• Compression: If needed, the data is compressed for efficient transmission.
• Encryption: If Alice uses TLS/SSL, the email is encrypted for secure transmission.

Data is then passed to the Session Layer.

Step 3: Session Layer (Layer 5)

• Session Establishment: The email client establishes a session with Alice’s SMTP mail server.
• Dialog Control: Ensures a stable session is maintained for the email transfer.
• Synchronisation: If the email is large, session checkpoints allow resumption in case of interruptions.

Data is then passed to the Transport Layer.

Step 4: Transport Layer (Layer 4)

• Segmentation: The email is broken down into smaller segments to ensure reliable delivery.
• Protocol Used:
  • TCP (Transmission Control Protocol) is used to ensure reliable, error-free delivery.
  • It assigns port numbers (SMTP typically uses port 25, or 587 for secure email transmission).
• Flow Control: Manages data rate so that the receiver isn’t overwhelmed.
• Error Control: Ensures correct data delivery with retransmission if needed.

Data is then passed to the Network Layer.

Step 5: Network Layer (Layer 3)

• Logical Addressing: Assigns the IP addresses of Alice’s computer and the mail server.
  • Example: Alice’s PC (192.168.1.2) → Mail Server (142.250.191.109).
• Routing: Determines the best path for the email packet to travel across networks.
• Packetisation: The segmented data from the transport layer is wrapped into packets.

Data is then passed to the Data Link Layer.

Step 6: Data Link Layer (Layer 2)

• MAC Addressing: Adds the sender’s and receiver’s MAC addresses (Physical device addresses).
  • Example:
    • Alice’s MAC:00:A0:C9:14:C8:2
    • Router’s MAC:00:B0:D0:86:BB:F7
• Framing: Encapsulates the email data into frames with headers for error detection.
• Error Detection: Uses Cyclic Redundancy Check (CRC) to detect transmission errors.
• Media Access Control: Determines when Alice’s computer can send data over the network (Ethernet, Wi-Fi).

Data is then passed to the Physical Layer.

Step 7: Physical Layer (Layer 1)

• Bit Transmission: Converts the email data into electrical, optical, or radio signals.
• Transmission Medium: Data travels over Ethernet cables, fibre optics, or Wi-Fi.
• Signal Transmission: The frames are transmitted bit by bit through the network.

The Email Reaches the Mail Server

1. The mail server receives the email, processes it, and forwards it to Bob’s mail server (e.g., mail.yahoo.com).
2. The data repeats the OSI process in reverse as the email is transmitted from Bob’s mail server to his computer.

Receiving the Email: Reverse Process on Bob’s Computer

1. Physical Layer: Bob’s computer receives the signal and converts it into binary data.
2. Data Link Layer: The MAC address is verified, and error-checking is performed.
3. Network Layer: The IP address is checked, and the data is routed to Bob’s device.
4. Transport Layer: TCP reassembles the segments into the complete email.
5. Session Layer: The session with Bob’s mail client is maintained.
6. Presentation Layer: If encrypted, the email is decrypted and decompressed.
7. Application Layer: Bob’s email client (e.g., Outlook, Thunderbird) displays the email.

Summary of the OSI Model in Email Transmission

OSI Layer	Sending Process (Alice’s Computer)	Receiving Process (Bob’s Computer)
7. Application	Alice composes and sends the email (SMTP)	Bob’s email client (POP3/IMAP) retrieves it
6. Presentation	Formats, encrypts, and compresses data	Decrypts, decompresses, and formats data
5. Session	Establishes a session with the mail server	Maintains and terminates session after transfer
4. Transport	Segments data, ensures reliability (TCP)	Reassembles segments into the full email
3. Network	Adds IP addresses, routes data	Verifies IP, routes to Bob’s machine
2. Data Link	Adds MAC address, creates frames	Verifies MAC, extracts packets
1. Physical	Converts data into signals for transmission	Converts signals back into binary data

How Data Actually Moves on the Internet (With TCP/IP)

Let’s say you send an email. Here’s what happens in real-world networking:

1. Application Layer (TCP/IP Model) → You type and send the email via Gmail.
2. Transport Layer (TCP/IP Model) → The email is broken into smaller packets and assigned sequence numbers.
3. Internet Layer (TCP/IP Model) → The packets are wrapped with IP addresses to find their destination.
4. Network Access Layer (TCP/IP Model) → The packets are transmitted over Wi-Fi, Ethernet, or mobile networks as electrical or radio signals.

When the recipient gets the email, the process happens in reverse!

Why This Distinction Matters

• OSI = Conceptual Model – Helps us understand networking in theory.
• TCP/IP = Practical Model – Actually used in the real world.

Even though we don’t follow OSI strictly, it’s still the foundation that shaped modern networking—so understanding it makes you a better problem-solver in the digital world.

---

Network Topologies: How Devices Are Connected

Now that we understand how data moves through a network, let’s talk about the physical and logical structure of networks—network topologies.

A network topology defines how devices (nodes) are arranged and connected, The choice of topology affects performance, fault tolerance, and scalability—making it a crucial design decision in networking.

Let’s explore the five major topologies and how they impact real-world networks.

1. Bus Topology – The Network Highway

How It Works:

• All devices share a single backbone cable that carries data.
• Devices connect using drop cables and taps—like exits on a highway.
• Data travels in both directions, but only one device can transmit at a time.

Advantages:

•  Easy to install – Requires minimal cabling, making it cost-effective for small networks.
• Efficient for simple applications – Works well when only a few devices need connectivity.

Disadvantages:

• Difficult to troubleshoot – A single failure can disrupt the entire network.
• Limited scalability – Adding devices weakens the signal and increases network congestion.
• Single point of failure – If the backbone cable fails, the entire network goes down.

💡 Where is it used?

• Older Ethernet networks (10BASE2, 10BASE5).
• Small office or home networks with few connected devices.

---

2. Ring Topology – The Circular Relay

How It Works:

• Devices form a closed loop, with each node connected to exactly two other devices.
• Data travels in one direction (unidirectional) or both (bidirectional).
• Repeaters amplify signals to prevent data loss over long distances.

Advantages:

• Organised data flow – Since data moves in a structured path, there’s minimal collision.
• Easy fault detection – If a device stops receiving data, it can trigger an alert.

Disadvantages:

• Breaks cause complete failure – If one device or connection fails, the entire network stops working.
• Not flexible – Adding or removing devices requires modifying two adjacent connections.
• Slower than star topology – Data must pass through multiple devices before reaching its destination.

💡 Where is it used?

• Token Ring networks (legacy) – Some older business networks.
• Fiber Distributed Data Interface (FDDI) – Used in high-speed backbone networks.

---

3. Star Topology – The Most Common Choice

How It Works:

• All devices connect to a central hub or switch, which directs network traffic.
• When a device sends data, the hub/switch forwards it to the correct destination.
• This is the most widely used topology in modern networking.

Advantages:

• Easy to expand – New devices can be added without affecting others.
• Better fault isolation – If one device fails, the rest of the network remains functional.
• High performance – Direct connections prevent delays caused by multiple hops.

Disadvantages:

• Single point of failure – If the hub/switch fails, the entire network goes offline.
• Higher cost – Requires more cables than bus or ring topologies.
• Scalability limitations – Growth depends on the number of available ports in the hub.

💡 Where is it used?

• Modern Ethernet networks – Home, office, and corporate environments.
• Wi-Fi networks – The router acts as a central hub.

4. Mesh Topology – The Ultra-Reliable Network

How It Works:

• Every device connects to every other device through dedicated links.
• Supports full-duplex (simultaneous transmission) or half-duplex (one at a time) communication.
• Used in critical applications where failure is not an option.

Advantages:

• Highly robust – Multiple paths ensure network reliability.
• No congestion – Every device has dedicated bandwidth.
• High security – Data is sent through direct, private channels.

Disadvantages:

• Expensive to implement – Requires excessive cabling and ports.
• Complex installation & maintenance – More connections mean more troubleshooting.
• Not scalable – The number of connections grows exponentially as devices increase.

💡 Where is it used?

• Military and aerospace networks (high-security, mission-critical systems).
• Financial institutions (where uptime is crucial).

5. Hybrid Topology – The Best of Both Worlds

How It Works:

• A combination of two or more topologies (e.g., Star-Bus, Star-Mesh).
• Commonly used in large networks where different sections need different structures.

Advantages:

• Highly flexible & scalable – Can be adapted based on needs.
• Optimised performance – Leverages strengths of different topologies.
• Better fault tolerance – Redundancy improves reliability.

Disadvantages:

• Complex design & maintenance – Requires careful planning.
• Higher cost – More cabling and hardware needed.
• Troubleshooting challenges – Identifying faults can be difficult in multi-topology setups.

💡 Where is it used?

• Enterprise and data center networks (balancing performance & reliability).
• Large-scale cloud infrastructures (e.g., AWS, Google Cloud).