famousrajbhatt@gmail.

com

1
Computer Networks
 INDEX

famousrajbhatt@[Link]
[Link] TOPIC PAGE NO.
Computer Networks 1 – 57
1 Introduction 3–9
2 Devices 10 – 14
3 Physical Layer 15 – 16
4 Data link Layer 17 – 27
5 Network Layer 28 – 37
6 Transport Layer 38 – 44
7 Session Layer 45
8 Presentation Layer 46
9 Application Layer 47 – 51
10 Delays 52 – 53
11 Interview Questions 54 – 55
12 MCQs & Answer Key 55 – 57

Important topics

2
 Introduction

famousrajbhatt@[Link]
1. What is a computer network?
A computer network is a group of interconnected nodes (devices) that share resources and
information via communication links.

2. What is a node?
• A node is a device capable of sending and receiving data.
End nodes: PC, printer, server, smartphone.
Intermediary nodes: Router, Bridge, Modem.

3. What is a communication link?

A communication link is a medium to transfer data.
Types
• Wired (coaxial cables)
• Wireless (Wi-Fi, Bluetooth, etc.)

4. What are the uses of a computer network?

• Communication
• Resource sharing
• Remote access
• Collaboration
• EEE stands for Entertainment, Education, E-commerce.

5. What are characteristics of computer network?

• Fault Tolerance (Continue to work despite failures).
• Scalability (Add devices without degrading performance).
• Quality of Service (Network works smoothly without delays/interruptions).
• Security (protection of data from unauthorized access, alteration, and disruption).

6. What are the services offered by computer networks?

Email, instant messaging, video telephony, storage service, online games, web browsing, file
sharing, VOIP (Voice over Internet Protocol), etc.

7. What is a protocol?
• Set of rules for effective communication.
• The sender and receiver follow a common set of rules to communicate.

3
8. Types of data flow
i. Simplex: One-way communication (e.g., keyboard to
CPU)

famousrajbhatt@[Link]
ii. Half Duplex: Two-way communication, but only one
direction at a time (e.g., walkie-talkie)
iii. Full Duplex: Two-way communication simultaneously
(e.g., telephone)

9. What does a protocol generally contain?

Information about
• Sender
• Receiver
• Communication media

10. What are the elements of a protocol?

• Message Encoding
• Message Formatting and Encapsulation
• Message Timing
• Message Size
• Message Delivery Options

11. Types of Networks

Networks PAN LAN CAN MAN WAN

Personal Campus Metropolitan
Local Area Wide Area
Full Form Area Area Area
Network Network
Network Network Network

Leased line,
Bluetooth, Ethernet, Metro
Ethernet, MPLS,
Technology IrDa, Wi-Fi, Fiber Ethernet,
wi-fi VPN,
Zigbee optics MPLS
Satellite

Above 50
Range 1 - 100 m upto 2km 1 - 5 km 5 - 50 km km to global
range

Transmission
Very high Very high High Average Low
speed

Within Within
Within a Within cities Within
Area office university,
room like Mumbai countries
building office, etc.

Private/ Private/
Ownership Private Private Private
Public Public

Maintenance Very easy Easy Moderate Difficult Very difficult

Error Rate and

Very low Low Moderate High Very high
Cost

4
12. Client-Server vs Peer-to-Peer (P2P) Network Models

Client Server Network Peer to Peer (P2P) Network

famousrajbhatt@[Link]
Centralized with one or more dedicated Decentralized network where each
servers system functions as both client and server

Each user manages his or her own

Server manages access and resources
system” OR “their own systems

More secure with authentication and

Less secure due to lack of central control
access control

Limited scalability, Becomes unstable as

Easily scalable for growing networks
the number of peers increases

Higher cost due to server setup and

Lower cost, no dedicated server required
maintenance

Used in web apps, email systems, Used in File Sharing, P2P Games,
databases, enterprises Blockchain, and Home Networks

Architecture
• Client- Server • Peer-To-Peer Network

13. What is a topology?

Topology refers to the layout and arrangement of nodes in a computer network.
Categories of topology
i. Physical topology: placement of various nodes.
ii. Logical topology: deals with the data flow in the network.

Key Points to Remember:

1. Shape
2. Working
3. Advantages and Disadvantages

Types: Bus, Ring, Star, Mesh, Tree, Hybrid.

5
Bus Topology
• Definition: All devices share a single central cable.

famousrajbhatt@[Link]
• Data Flow: Bidirectional along the bus.
• Pros:
o Easy to set up.
o Minimal cabling.
• Cons:
o Cable failure breaks network.
o Limited devices and slow with traffic.
• Use: Suitable for small networks or temporary network setups.

Ring Topology
• Definition: Devices form a closed loop, each connected to two others.
• Data Flow: Typically, one-directional using a circulating token (some versions support
bidirectional flow).
• Pros:
o Simple setup.
o Fast data transfer.
• Cons:
o One failure affects the whole network.
o Harder to troubleshoot.
• Use: Token-based, orderly data networks.

Star Topology
• Definition: All devices connect to a central hub/switch.
• Data Flow: Via central hub to destination.
• Pros:
o Easy setup & fault isolation
o One device failure doesn't affect others
• Cons:
o Hub failure breaks network
o More cabling needed
• Use: Common in homes/offices.

Mesh Topology
• Definition: Every device is connected to every other device directly.
• Data Flow: Data can take multiple paths between devices.
• Advantages:
o Highly reliable and fault-tolerant.
o If one link fails, data routes through other paths.
o Provides high privacy and security.
• Disadvantages:
o Expensive and complex to install.
o Requires lots of cables and ports.
• Use: Critical networks requiring high availability and redundancy.

6
 Tree Topology
• Definition: Hierarchical topology combining multiple star networks connected to a central

famousrajbhatt@[Link]
bus backbone.
• Data Flow: Data flows from leaves to root and vice versa through the backbone.
• Pros:
o Scalable and easy to manage.
o Fault isolation is easier than bus topology.
o Supports future expansion.
• Cons:
o If the backbone fails, entire network
segments go down.
o More cabling required than bus or star.
• Use: Large networks with hierarchical structure, like university campuses.

Hybrid Topology
• Definition: Mix of two or more topologies
(e.g., bus + ring + star).
• Data Flow: Depends on combined types.
• Pros:
o Flexible & scalable.
o Customizable design.
o Easier fault isolation.
• Cons:
o Complex setup.
o Higher cost.
• Use: Large, tailored networks.

Test your understanding

i. Which network topology is the most expensive to implement and maintain?
ii. Which topology allow data to flow in only one direction?
iii. Which topology ensures communication continues even if one link fails?
iv. Which topology offers centralized control at minimal cost for connecting branch offices to a
central hub?

7
14. What is Switching?
Switching is a method of transmitting data

famousrajbhatt@[Link]
between devices by selecting the best available
path.

i. Circuit Switching
• Circuit switching is a switching technique in which a dedicated communication path
is established between two endpoints for the duration of a transmission.
• Phases:
1. Connection Establishment.
2. Data Transfer.
3. Connection Termination
• Used in: Traditional telephone networks
(a dedicated line is reserved during a call).
• Pros: Reliable, fixed path.
• Cons: Wastes resources if idle.

ii. Message Switching

• Uses store-and-forward mechanism.
• Message is transferred as a complete unit → stored → and then forwarded.
• Suitable for email and non-real-time data.
• Not suitable for streaming applications due to intermediate message storage delays.
Pros:
• No need for a dedicated path.
• Handles variable message lengths easily.
Cons:
• Slower due to message storage at each
node.
• Not ideal for real-time communication.

iii. Packet Switching

• Data is broken into small packets, each sent independently.
• Uses store-and-forward at each node.
• Efficient for internet, emails, and most data networks.
• Supports multiple paths and dynamic routing.
Pros:
• Efficient bandwidth use.
• Robust and scalable.
• Efficiently handles bursty traffic, where data is sent in irregular bursts.
Cons:
• Possible delays and packet loss.
• Requires complex routing.

8
 Types of packet switching

Feature Datagram Approach Virtual Circuit Approach

famousrajbhatt@[Link]
Connectionless: No path is Connection-oriented: Path established
Connection
established before data transfer before data transfer

Each packet may take a different All packets follow the same established
Routing
route route

Addressing Packets carry full destination address Packets carry a virtual circuit identifier

Setup Time No setup required Setup phase before data transfer

Packets may arrive out of order or be Packets arrive in order

Reliability
lost, making it less reliable. (more reliable)

Lower, as packets use short virtual

Higher, since each packet carries the
Overhead circuit identifiers instead of full
complete destination address.
addresses.

Example Internet Protocol (IP) ATM, Frame Relay

Datagram Approach Virtual Circuit Approach

Test your understanding

i. Students all get on the same bus together. Which switching method does this remind you
of?
ii. College friends come from different places and meet up at the canteen. Which switching
technique does this represent?
iii. How does a virtual circuit reduce overhead compared to the datagram approach?

9
 Devices

famousrajbhatt@[Link]
1. Devices
Hardware: Cable, Repeater, Hub, Modem
Hardware + Software: Bridge, Switch, Router
Software: Gateway, IDS, Firewall.

2. Cables
Cables are a medium used to send data from one device to another. Twisted Cable

Types of Cables:
i. Unshielded Twisted Pair (UTP) Cables
• Commonly used in homes and offices.
• Pairs of wires are twisted to reduce electromagnetic interference
Coaxial Cable
(EMI).
ii. Coaxial Cable
• Has a single copper wire in the center.
• Used for cable TV and internet.
Fiber Optic
iii. Fiber Optic Cable
• Sends data using light signals.
• Very fast and supports long distances.

3. Repeaters
Work at the Physical Layer
A 2-port device (one input, one output)

Why Repeaters Are Needed:

• When signals travel long distances, they lose strength (attenuation).
• A repeater boosts the signal back to its original strength, so it doesn’t get lost while
traveling.

Key Points:
• Placed at regular intervals along the cable.
• Does forwarding of the signal.
• No filtering — only boosts and sends.
• Collision domain = n (where n is the number of devices) because repeater transmits
without memory or buffering
o No buffer or memory.
o Does not use store-and-forward; simply amplifies and transmits signals in real-time.

10
4. Difference Between Repeater and Amplifier:

famousrajbhatt@[Link]
Feature Repeater Amplifier

Restores signal to its original Just boosts the signal power

Purpose
strength (e.g., 1x → 2x or 3 x)

Noise Removal Removes noise from the signal Amplifies both signal and noise

5. Hub
• Works at the Physical Layer.
• It is pure hardware (no smart processing).
• Acts as a multiport repeater (has multiple
input/output ports).
Key Features:
• Forwards data to all connected devices
(broadcasts).
• No filtering – sends everything to
everyone
o This causes high traffic on the network.
• Hubs are prone to collisions since all data is broadcast.

6. Bridges
• Work at the Data Link Layer
(Layer 2)
• Used to connect two or
more LAN segments.
• More intelligent than hubs or
repeaters.

Key Features:
• Does forwarding of data
• Checks MAC addresses to decide where to send data
• Can filter traffic and reduce collisions ⇒ because of store-and-forward mechanism
• Maintains a table to learn which MAC addresses are on which side

11
 Example:
• Has two interfaces/ports:

famousrajbhatt@[Link]
o P1 (connected to M1, M2, M3, M4)
o P2 (connected to M5, M6, M7, M8)
• If M1 sends a message to M3, • If M1 sends a message to M7,
o The message goes to the Bridge first o The message goes to the Bridge first
o Bridge checks its MAC table o Bridge checks its MAC table
o It learns M3 is on the same port segment P1 o It learns M7 is on the other side of P1
o So, it does not forward the message to P2. o So, it forwards the message to P2
o This reduces unnecessary traffic on the
other segment.

This is something repeaters and hubs cannot do.

7. Types of Bridges: Static vs Dynamic

Feature Static Bridge Dynamic (Transparent) Bridge

Configuration Manual (by administrator) Automatic (self-learning)

Learning Capability No Yes

MAC Address Table Manually maintained Built and updated automatically

Handling New Devices Requires manual update Learns automatically through traffic

Manual reconfiguration
Port Changes Automatically adapts
needed

Minimal or no broadcasting
Initial Operation (depending on manual table Broadcasts initially to learn devices
setup)

Flexibility Low High

8. Switch
• Operates at Layer 2 (Data Link Layer)
• Functionally similar to a multiport
bridge, but more advanced in terms
of speed and switching logic
• Supports full-duplex links
(communication both ways
simultaneously)
• Creates separate communication
paths for different device pairs
(e.g., A-B, C-D)
• Results in minimal network traffic
• Collision domain per port = 1; no
collision across ports
• No collisions occur because each link is a dedicated circuit
Example:
• Commonly used to connect devices to the network, e.g., Laptop → Switch → Router

12
9. Roputer
• Works at the Network Layer (Layer 3) of the

famousrajbhatt@[Link]
OSI model.
• Connected to multiple networks.
• Forwards packets using a routing table.
• Routers drop or ICMP unreachable; flooding
is not default behavior.
• Routes packets based on IP addresses.
• Routers separate collision domains, but
collisions may still occur on half-duplex Ethernet interfaces.

Protocols Used
• RIP (Routing Information Protocol)
• OSPF (Open Shortest Path First)
• BGP (Border Gateway Protocol)

10. Gateway
• Connects two different networks (e.g., LAN to WAN).
• Acts as a translator between different network
protocols.
• Placed at the edge of LAN, connecting to the
internet.
• Gateways work across multiple OSI layers
depending on protocol translation but is
commonly associated with layer 7 (Application
Layer)
• Handles protocol conversion and sometimes
NAT/firewall.
Example:
A VoIP gateway connecting IP-based networks
with traditional telephone networks.

11. Firewall
• A security device that monitors and
controls network traffic.
• Acts as a barrier between trusted
(LAN) and untrusted (Internet)
networks.
• Placed at the network boundary,
usually before/after the router.
• Filters traffic based on IP, port, or
protocol using predefined rules.
• Can be hardware, software, or
both.
Example: pfSense, Cisco ASA, Windows Defender Firewall.

13
12. IDS (Intrusion Detection System)
• Monitors network or system activity to detect suspicious behavior or attacks.

famousrajbhatt@[Link]
• Does not block traffic, only detects and alerts.
• Can detect threats like unauthorized access, malware, or policy violations.
• Works by Analyzing traffic patterns, logs, or behavior.
• Two main types:
o NIDS (Network IDS) - monitors entire network.
o HIDS (Host IDS) - monitors individual devices.
• Example:
Snort (for NIDS), OSSEC (for HIDS)

14
 Physical Layer

famousrajbhatt@[Link]
1. Where is it located in the OSI Model?
It is the 1st layer from the bottom (7th from the top) in the OSI
Model.

2. What is the form of data? → Bits

3. Which devices operate at the Physical Layer?

• Cables
• Hub
• Repeater

4. What are the services offered by this

layer?
i. Bit-by-bit Transmission
• Data is sent as a continuous stream
of individual bits.

ii. Encoding and Decoding

• Encoding converts data into signals
suitable for transmission (e.g.,
electrical, optical).
• Decoding converts received signals
into a form that can be interpreted as
data by higher layers.

iii. Signal Transmission

• It refers to transmitting encoded data
over a physical medium such as
wires, fiber optics, or air.
• The transmission can use analog (continuous) or digital (discrete) signals, depending
on the communication method.

iv. Modulation and Demodulation

• Modulation alters a
carrier signal to encode
and transmit data.
• Demodulation retrieves
the original data from
the modulated carrier signal at the receiver.
• A modem is the device used for modulation and demodulation.

15
5. Types of connection
i. Point-to-point

famousrajbhatt@[Link]
• Connects exactly two devices directly.
• Used for direct communication
between two devices.

ii. Multipoint
• Multiple devices share a single link.
• Used for shared communication over a single
physical link.

6. Types of communication modes?

i. Unicasting: If the message is sent from the source to a single node, it
is known as unicasting. This is the most common communication
method in networks, where data is sent from one sender to one specific
receiver.

ii. Anycasting: If the message is sent from the source to any one of the
nodes, it is known as anycasting. It is commonly used in Content
Delivery Networks (CDNs), where the request is routed to the nearest or
best-performing server.

iii. Multicasting: If the message is sent to a subset of nodes from the

source then it is known as multicasting. Used to efficiently deliver the
same data to a selected group of receivers simultaneously.

iv. Broadcasting: If the message is sent from a source to all the nodes in
a network, it is known as broadcasting. Protocols such as DHCP and
ARP use broadcasting within local area networks (LANs).

16
 Data link Layer

famousrajbhatt@[Link]
1. Where is it located in the OSI Model?
It is the 2nd layer from the bottom (6th from the top) in the OSI Model.

2. What is the form of data? → Frames

3. Sublayers
i. Logical Link Control (LLC)
• Handles flow control and multiplexing between upper
layers and the MAC sublayer.
• Provides:
o Error detection notifications to higher layers
o Acknowledgments
ii. Media Access Control (MAC)
• Manages device interaction with the physical medium.
• Responsibilities:
o Frame addressing
o Provides access control to the physical transmission medium
o Encapsulates Network Layer packets into frames and forwards them to the Physical
Layer.

4. What are the services offered by this layer?

i. Framing
Encapsulates packets into frames for transmission.
ii. Physical / MAC Addressing
Adds source and destination MAC addresses to the frame header.
iii. Flow Control
• Prevents the sender from overwhelming the receiver.
• Protocols:
o Stop-and-Wait
o Go-Back-N (GBN)
o Selective Repeat (SR)
iv. Error Control
Provides error detection and optional correction.
Error Detection Techniques:
• Single-bit parity check
• Two-dimensional parity check
• Checksum
• Cyclic Redundancy Check (CRC)
Error Correction Technique:
• Hamming Code

17
 v. Access Control
Manages access to the shared medium to avoid collisions.

famousrajbhatt@[Link]
Random Access Protocols:
• ALOHA
• CSMA (Carrier Sense Multiple Access)
o 0-persistent
o 1-persistent
o p-persistent
• CSMA/CD (Collision Detection)
Controlled Access Protocols:
• Polling
• Token Passing
Channelization Protocols:
• FDMA (Frequency Division Multiple Access)
• TDMA (Time Division Multiple Access)

5. How does data link layer work?

• Receives raw bits from the Physical Layer and organizes them into meaningful frames.
• Detects and corrects errors in the received frames before passing data to the Network
Layer.
• Uses MAC addressing to ensure the frame reaches the correct device on the local
network.
• Controls access to the shared medium, ensuring devices transmit data without
collisions.
• Delivers clean, framed data to the Network Layer, which adds IP logic and handles routing.

6. Which devices operate at the Data Link Layer?

• Switch
• Bridge
• Network Interface Card (NIC)
• Wireless Access Point (WAP)

7. Framing
• Groups raw bits into frames for clear separation.
• Enables reliable node-to-node (hop-to-hop) delivery across the physical link.
• Structure of a frame:

o Flag: Like an envelope in a postal system. It defines the start and end of data (delimiter).
o Header: Contains source and destination MAC address.
o Payload Field: It is the actual data.
o Trailer: Error detection and correction bits.

18
 Types of Framing
i. Fixed-Size Framing

famousrajbhatt@[Link]
• Frame size is constant.
• No delimiters needed; length defines
frame boundaries.

ii. Variable-Size Framing

• Frame size varies; requires delimiters.
• Two methods:
a. Length Field - Specifies frame size in
the header.
b. End Delimiter - Special pattern marks end of frame.
o Byte Stuffing - Adds extra byte to avoid confusion (character oriented).
o Bit Stuffing - Adds extra bits to avoid confusion (bit-oriented).

8. Physical/ MAC addressing

• NIC (Network Interface Card): A hardware component that connects a device to a
network.
• NIC can be wired (Ethernet) or wireless.
• WAP (Wireless Access Point): A device that lets wireless devices connect to a wired
network using Wi-Fi.
• WAP connects wireless devices to a wired network and extends wireless coverage.
• MAC Address: A unique hardware identifier assigned to the NIC for identification on a local
network.

9. Flow Control
i. Stop-And-Wait
• Window size = 1 (only one frame in transit).
• Sender sends 1 frame, waits for ACK
before sending next.
• Sender is idle during wait.
• Prevents data loss if receiver is slower.
• Simple but inefficient due to idle time.
Problems in Stop-and-Wait Flow Control
• Lost Data
• Lost Acknowledgement
• Delayed ACK/Data

ii. Go-Back-N Protocol

• It is a sliding window protocol.
• It is a flow control and error control protocol.
• The sender can send N (size of sender window) frames in a row without waiting for
an acknowledgment.
• Receiver sends cumulative acknowledgment for the last correctly received in order
frame.

19
 Transmission Process
• Suppose the sender has to send 11
frames (0 to 10).

famousrajbhatt@[Link]
• Window size is 4.
• Sender sends frames: 0, 1, 2, 3.
• Receiver sends ACK for frame 0.
• Sender slides window and sends
frame 4 (now window is 1,2,3,4).
• Receiver ACKs frame 1, sender sends frame 6 (window is 3,4,5,6).
• This process continues.
Error Handling (Go-Back Mechanism)
• If a frame or ACK is lost, the sender
waits.
• If ACK for frame 2 is missing:
o Sender cannot send frame 6.
o After a timeout, sender goes back
to frame 2.
o Retransmits frames 2 onward (e.g.,
2, 3, 4...), even if some were
received correctly.
• This ensures reliable delivery but
may cause extra retransmissions.

iii. Automatic Repeat reQuest / Selective Repeat ARQ

• Multiple frames are sent; multiple frames are acknowledged.
• Only lost/damaged frames are
retransmitted.
• Receiver buffers out-of-order frames
temporarily.
• ACK sent for each correctly received frame.
• NACK sent for missing or corrupted frames.
• More efficient than Go-Back-N (no
unnecessary retransmissions).

10. Error detection and correction

If the receiver does not receive the data exactly as sent by the sender, data is said to be
corrupted and error has occurred.
Types of errors

20
i. Error detection method
a. Single Bit Parity Check

famousrajbhatt@[Link]
• Helps detect single-bit errors in transmission.
• A basic error detection technique.
• Adds 1 extra bit (parity bit) to the data.
• Ensures total number of 1s is even.
o If data has odd number of 1s ⇒ add 1
o If data has even number of 1s ⇒ add 0
• In advanced parity methods, like 2D matrix-based parity, errors involving 2 or 3 bits can
be detected but not corrected.

b. Checksum
• A computed value (often a sum or similar operation) of data bits before transmission.
• Checksum is transmitted along with the data.
• Receiver recomputes the checksum and compares it.
• If values don’t match, an error is detected.

c. Cyclic Redundancy Check

• Uses polynomial division to generate a check value.
• More robust than parity and checksum.
• Commonly used in networks and file systems for error detection.

ii. Error Correction method

a. Two types of error correction:
• Backward Error Correction:
Receiver detects error, asks sender to retransmit.
• Forward Error Correction:
Receiver uses extra bits to detect and correct errors without retransmission.

b. Hamming code for error correction

• Parity bits are added at positions that are powers of 2 (1, 2, 4, 8…).
• Each parity bit checks a fixed set of positions based on binary rules.
• Even parity is used — parity bit is 0 if total 1s (including itself) is even, else 1.
• Sender sends the full message (data + parity bits).
• Receiver recalculates parity bits to detect errors.
• The binary value formed by the positions of incorrect parity bits indicates the position
of the error.
• Receiver flips the bit at that position to correct the error.

21
11. MAC (Multiple Access Control)
Determines how multiple devices share the communication channel.

famousrajbhatt@[Link]
1) Random Access Protocol
• No station has priority, all are equal.
• Any station can transmit when the medium is idle.
• No fixed timing, stations transmit at random times.
• Collisions may occur if two stations send at the same time.

i. ALOHA
a. PURE ALOHA
• When a station sends data, it waits for an ACK (Acknowledgement).
• If no ACK is received, it assumes a collision has occurred.
• The station waits for a random backoff time (Tb) before retrying.
• The back-off time in Pure
ALOHA is random
• Pure ALOHA does not perform
carrier sensing.
• Used in early LAN protocols,
now mostly obsolete.
• Only transmission time is
considered; propagation delay
is ignored.

Vulnerable Time = 2 x tp

• tp = amount of time required to send a frame between the two most widely separated
stations

22
b. SLOTTED ALOHA
• Any station cannot start
transmission at any time.

famousrajbhatt@[Link]
• Time is divided into equal
slots.
• A station can only start
transmission at the beginning
of a time slot.
• If a station misses the start of
the slot, it must wait for the
next one.
• This reduces the chance of
collisions compared to Pure
ALOHA.

Vulnerable Time = tp

ii. CSMA (Carrier Sense Multiple Access)

• A station senses the medium before transmitting.
• If the medium is idle, it starts transmitting data.
• If the medium is busy, it waits until
it becomes idle.
• Fewer collisions than ALOHA due
to sensing.
• However, collisions can still occur
due to propagation delay.
• Two stations may sense the medium
as idle at the same time and
transmit, causing a collision.

CSMA ACCESS MODES

P-Persistent
1-Persistent CSMA 0-Persistent CSMA
CSMA
• Sense channel → if idle, • Sense channel → if idle, transmit • Priority-based; nodes
transmit immediately. with probability p. have predefined
• If busy, keep checking • With (1-p) probability, wait a time transmission order.
continuously. slot, check again. • If channel is idle, each
• Transmit as soon as medium • Repeat until the frame is sent. node waits for its slot.
becomes idle (with 100% • Balance between efficiency and • Used when nodes follow
chance). collision control. a fixed backoff scheme.
• High chance of collision if • Used in Wi Fi, radio
multiple nodes do the same.

23
iii. CSMA/CD (Collision Detection)
• It is a slight modification of CSMA.

famousrajbhatt@[Link]
• If a collision occurs during transmission:
o Sends a jam signal to notify others.
o Waits a random backoff time, then retries.
• CSMA/CD doesn’t use acknowledgments; it
relies on collision detection.
• Requires transmission time ≥𝟐×
propagation delay for proper collision
detection. So that collisions can be detected
before transmission ends.”
• Not used in modern networks (e.g., switched
Ethernet).

iv. CSMA/CA (Collision Avoidance)

• Carrier Sensing:
o Station senses the medium to check if it's idle or busy.
• Interframe Space (IFS):
o If medium is idle, station waits for a short time (IFS) before proceeding.
o This helps avoid overlap due to propagation delay.
• Contention Window:
o After IFS, the station selects a random backoff time (in slots).
o If medium becomes busy during backoff, the timer pauses, and resumes once it's
idle again.
• Transmission:
o Once backoff ends and the medium is idle, data is transmitted.
• Acknowledgement:
o Receiver sends an ACK.
o If ACK not received, sender assumes collision and retries.

24
CSMA/CA Flow:
• Station senses the channel:

famousrajbhatt@[Link]
Check if the medium is idle.
• If medium is idle:
Wait for Interframe Space (IFS) —a
short delay based on priority.
• After IFS:
Check medium again.
o If still idle, proceed to send data.
o If busy, wait.
• If medium was busy initially or
after IFS:
Enter Contention Window:
o Randomly select a backoff slot
before transmission attempt.
o If transmission fails, double the
contention window size and
select a new random backoff time
(exponential backoff).
o Pause timer during busy periods
and resume when medium
becomes idle.
• Send data.
• Wait for acknowledgement (ACK):
o If ACK received, transmission is
successful.
o If no ACK is received within
timeout, retransmit using the same backoff mechanism.

25
2) Controlled Access Protocol
A protocol used in shared channels to avoid collisions by regulating device transmission access.

famousrajbhatt@[Link]
i. Reservation
• Devices reserve future time slots by marking their requests in a reservation frame
(containing mini-slots for requests). This is done in a centralized or distributed manner
depending on the system.
• Only the devices that reserved slots are allowed to transmit, eliminating collisions.
• Image shows: Stations mark their slots in the reservation frame and send data
accordingly in turn.

ii. Polling
• A central controller (primary) polls devices one at a time to
check if they have data to send.
• Only the polled device can respond, preventing data collisions.
• Image shows: Primary polls Station A and B; A responds with
no data (NULL or silence), B sends DATA and receives ACK.

iii. Token Passing

• A token circulates in a logical ring; only the holder
transmits. If lost, the protocol regenerates the token.
• Ensures orderly and collision-free data transmission.
• Image shows: The token being passed from Station 1 to
2, 3, and then 4 in a circular manner.

3) Channelization protocols
Channelization protocols divide the shared communication medium so that multiple
users can transmit data simultaneously without interference.

FDMA TDMA CDMA

Frequency Division Multiple Time Division Multiple Code Division Multiple

Access Access Access

Users get separate Users get separate time Users share frequency &
frequency bands slots time with unique codes

Transmit simultaneously on Transmit one after another Transmit simultaneously

different frequencies in sequence using different codes

Divides total bandwidth into Highly efficient due to

TDMA is more bandwidth-
fixed frequency bands, shared frequency/time;
efficient than FDMA in
which can lead to supports encryption for
digital systems.
underutilization. security.

Used in analog systems Used in GSM (2G) Used in 3G networks (e.g.,

(e.g., radio) networks Qualcomm)

26
 famousrajbhatt@[Link]
Ethernet Frame Format
• LAN technology Protocol
• Used by CSMA/CD

i. Preamble (7 Bytes)
• Used for bit synchronization.
• Receiver gets alert before the actual data starts flowing.

ii. Start Frame Delimiter - SFD (1 Byte)

• Marks start of frame (before destination address).
• Sometimes considered part of Preamble (making 8 Bytes total).

iii. Destination Address (6 Bytes)

• MAC address of the receiver.
• Can be unicast, broadcast, or multicast.

iv. Source Address (6 Bytes)

• MAC address of the sender.
• Always unicast (LSB of first byte is 0).
• Multicast addresses have LSB = 1.

v. Length (2 Bytes)
• Indicates length of the data field (up to 1500 Bytes).
• Value range can go up to 65535 in IEEE 802.3 but Ethernet payload limit is 1500 Bytes.

vi. Data / Payload (46-1500 Bytes)

• Contains actual IP header + user data.
• Padding is added if total frame (excluding preamble / SFD) is less than 64 Bytes
(14-byte header + 4-byte CRC + 46 – 1500 Bytes data).

vii. Cyclic Redundancy Check - CRC (4 Bytes)

• 32-bit error-detecting hash.
• Calculated on: Destination Address + Source Address + Length/Type + Data + Padding
(if present).
• Used to detect corruption at receiver side.

27
 Network Layer

famousrajbhatt@[Link]
1. Where is it located in the OSI Model?
It is 3rd layer from bottom; 5th from top.

2. What is the data format? → Packets

3. What are the services offered by this layer?

• Routing
• Logical Addressing
• Packet Switching
• Forwarding
• Fragmentation & Reassembly
• Subnetting
• NAT (Network Address Translation)

4. How does the Network Layer work?

i. Each device gets a unique IP address for identification.
ii. Data is split into packets with source & destination IPs.
iii. Routers choose the best path to reach the destination.
iv. Packets move hop-by-hop through routers.
v. Large Packets are broken into smaller packets.
vi. At the destination, packets are combined into original data.
vii. If delivery fails, error messages (like ICMP) are sent back.

5. What is an IP address?
• An IP address can be 32-bit (IPv4) or 128-bit (IPv6), used to identify devices on a network.
• Helps locate and communicate with devices, especially outside your local network (LAN).
• 32 bits → each 8 bits are separated by dots(.)
• Types of IP address
i. Public
ii. Private
• We have two different addressing types
o Classful addressing
o Classless addressing (CIDR)

6. What is Classful addressing

• IPv4 addresses divided into 5 classes: A, B, C, D, E.
• Each class has a fixed range and a class-specific subnet mask.
• Determines how the IP is split into network and host parts.

28
 First Number of
Leading Network Host Default
Class Octet Purpose Networks
Bits Bits Bits Subnet Mask
Range Calculation

famousrajbhatt@[Link]
Large
A 0-127 0 8 24 [Link] 27=128
networks

Medium/large
B 128-191 10 16 16 [Link] 214=16,384
networks

Small
C 192-223 110 24 8 [Link] 221=2,097,152
networks

D 224-239 1110 N/A N/A Not defined Multicasting N/A

Research /
E 240-255 1111 N/A N/A Not defined N/A
Experimental

First octet range

The first octet range shows the first number in the IP address.
Ex. 0-127 = [Link] to [Link] (similarly for B-E)

Subnet Mask
32-bit number used to determine the network portion of an IP address via bitwise AND.

IP & Subnet Mask = Network ID

Example
o [Link] & [Link] → [Link]
o [Link] & [Link] → [Link]
o [Link] & [Link] → [Link]

Number of Networks Calculation

Total network bits – fixed leading bits = bits for networks

Number of networks = 2 Total network bits

Example (Class A):

• Network bits = 8
• Leading bits fixed = 1
• Remaining bits = 8 - 1 = 7
• Networks = 27= 128
Number of Hosts Calculation

Total host bits = bits reserved for hosts in the IP class

Number of hosts per network = 2 Host bits − 2

Example (Class A)
o Host bits = 24
o Actual hosts = 224 - 2 = 16,777,214

29
 Why do we subtract 2?

Type Description Host Bits Purpose

famousrajbhatt@[Link]
First IP address in the
Network ID All host bits = 0 Identifies the network
network

Last IP address in the

Broadcast Address All host bits = 1 Sends data to all hosts
network

Example for Class C

Address Type Host Portion (last 8 bits) IP example

Network ID 00000000 [Link]
Usable hosts 00000001 to 11111110 [Link] to [Link]
Broadcast Address 11111111 [Link]

7. Disadvantages of Classful addressing?

• Class A, B: Blocks are too large for most organizations → many addresses wasted.
• Class C: Blocks can be too small for some organizations → insufficient addresses.
• Class D: Used for multicast groups → addresses not assigned to individual hosts, causing
wastage.
• Class E: Reserved for future use → currently unused, causing wastage.

8. Classless addressing
• Improved IP system replacing classful addressing.
• Efficient IP allocation based on user need.
• CIDR = Classless Inter-Domain Routing.
• IPs are given in format:
CIDR Block
• A CIDR block is a range of IP addresses assigned dynamically.
• It contains the exact number of IPs needed.
Rules to Form CIDR Block
i. All IPs in the block must be contiguous.
ii. Block size must be a power of 2 (e.g., 2¹, 2², 2³, ...).
iii. First IP must be divisible by the block size.
NOTE: An address is divisible by 2ⁿ if its last n bits in binary are all equal to 0.
CIDR Notation (/n)
• n = number of network bits
• 32 - n = number of host bits

CIDR notation → x. y. z. w/n

Example:
[Link]/28
→ 28 bits for network, 4 bits for host
→ 24 = 16 - 2 = 14 usable hosts.

30
9. Subnet
• A subnet is a smaller group inside a larger network.

famousrajbhatt@[Link]
• Allows efficient data routing within smaller, isolated sections.
• Each subnet can operate and be managed independently.
• Subnetting means dividing a large network into smaller networks called subnets.
Benefits of subnetting
• Efficient IP allocation.
• Reduces network congestion.
• Enhances security, scalability and speed.

10. How Subnetting works

• Each subnet has its own IP range; inside which devices can talk directly.
• Routers connect different subnets.
• Borrow bits from host part to create subnets:
• Subnet ID = First IP in subnet (network address).
• Broadcast ID = Last IP in subnet (for sending to all hosts).
• Formula:
Subnets = 2 borrowed_bits

Hosts = 2 host_bits_left – 2

Example (for Class C):

o Borrow 2 bits → 22 = 4 subnets; 8 - 2 = 6 host bits left,
o 26 – 2 = 62 (64 - 2) hosts each
NOTE: More subnets mean fewer hosts per subnet.

11. Subnetting types

FLSM (Fixed Length Subnet VLSM (Variable Length Subnet

Feature
Mask) Mask)
Different masks per subnet based on
Subnet Mask Same mask for all subnets
need

Fixed-size blocks, which may not

Variable-size blocks, matching the host
IP Allocation always match specific host
needs
requirements.

Less efficient; More efficient;

Efficiency
wastes IP addresses reduces IP wastage

Addressing Typically used with classful Typically used with classless

Type addressing addressing (CIDR)

[Link]/24 split into four /26 [Link]/24 split into /25, /26, /27
Example
subnets (64 Ips each) subnets matching host needs

31
12. IPV4 Header

famousrajbhatt@[Link]
• Version (4 bits): IP version number (4 for Ipv4).
• Header Length (HLEN, 4 bits): Length of the header in 32-bit words (min 5, max 15).
• Type of Service (8 bits): Priority flags like Low Delay, High Throughput, Reliability.
• Total Length (16 bits): Size of header + data (20 to 65,535 bytes).
• Identification (16 bits): Unique ID for packet fragments.
• Flags (3 bits):
o Reserved: Always set to 0; not currently used.
o DF (Don’t Fragment): Prevents the packet from being fragmented.
o MF (More Fragments): Indicates that more fragments are coming after this one.
• Fragment Offset (13 bits): Position of fragment in original data (in 8-byte units).
• Time to Live (TTL, 8 bits): Limits packet lifetime (prevents infinite looping).
• Protocol (8 bits): Indicates protocol (TCP, UDP, ICMP) for payload.
• Header Checksum (16 bits): Error-check for header integrity.
• Source IP Address (32 bits): Senderʼs IP.
• Destination IP Address (32 bits): Receiverʼs IP.
• Options (variable): Optional data (e.g., routing info, diagnostics).

32
13. IPV6 Header

• Version (4 bits): IP version

famousrajbhatt@[Link]
number, always 6 (binary 0110).
• Traffic Class (8 bits): Indicates
packet priority/class; helps with
QoS. Similar to Ipv4ʼs Type of
Service.
• Flow Label (20 bits): Identifies
packet flow for special handling
like QoS or real-time service.
• Payload Length (16 bits): Length of data following the header (includes extension headers
+ actual payload). Max = 65,535 bytes.
• Next Header (8 bits): Type of next header (can be an extension header or upper-layer
protocol like TCP/UDP).
• Hop Limit (8 bits): Like TTL in Ipv4; max number of hops before the packet is discarded.
• Source Address (128 bits): Ipv6 address of the packet sender.
• Destination Address (128 bits): Ipv6 address of the final receiver.
• Extension Header
o To replace IPV4 options field.
o Each has a Next Header (8 bits) → points to next header or upper-layer protocol
(TCP/UDP).
o We can have it for Hop-by-Hop options, Destination options, Routing, Fragment,
Authentication, Encapsulating Security Payload.

Important difference

Feature Ipv4 Header Ipv6 Header

Header Size Variable (20-60 bytes) Fixed (40 bytes)

Version 4 6

Address Length 32-bit 128-bit

Options/Extensions Options in header (variable length) Uses separate Extension Headers

Not present (removed to speed up

Checksum Present (Header Checksum)
processing)

Done only by source using

Fragmentation Done by routers and hosts
extension header

QoS Support ToS (Type of Service) field Traffic Class field

Flow Label Not available Present (used for flow identification)

Identification, Flags, Fragment Simplified; adds Flow Label,

Other Fields
Offset, etc. removes some fields

TTL/Hop Limit Time to Live (TTL) Renamed as Hop Limit

33
14. Routing
Routing is the process of selecting the best path for data

famousrajbhatt@[Link]
to travel from a source to a destination across networks.
How does it work?
i. A device sends a request to another device (e.g., a web
server).
ii. Data is split into smaller packets. Each packet gets the
destination IP address.
iii. The router checks its routing table to find the best path
to the destination.
iv. Packets travel through multiple routers (hops) to
reach the destination.
v. Packets are collected and reassembled at the
destination. Errors are checked and corrected.

Routing table
A routing table shows the best paths to reach different networks.

Feature Static Routing Dynamic Routing

Automatically updated by routing
Configuration Manually set by the network admin
protocols

Changes only when manually Adjusts automatically to network

Updates
reconfigured changes

Ideal for large, complex, or changing

Use Case Best for small and stable networks
networks

Resource Low resource usage, but less Higher resource usage, but more
Use flexible flexible

Example Small office router with fixed paths Protocols like RIP, OSPF, EIGRP

15. Types Routing Protocols [used for unicast communication]

• Intra-domain: Routing within a single organization or autonomous system (AS).

• Inter-domain: Routing between different organizations or ASes.

34
 For the solved example have a look on the video.

i. Distance Vector Routing (RIP)

famousrajbhatt@[Link]
• Each router maintains a table listing the best-known distance and next hop to each
destination.
• Routers periodically share only their distance vectors (not the entire routing table)
with immediate neighbours.
• Upon receiving a neighbourʼs distance vector, a router updates its own vector by
selecting routes with the lowest distance metric.
• Shares routing info only with neighbour routers.

RIP: Routing Information Protocol

RIP is a Distance Vector protocol using hop count as its metric.
Advantages:
• Simple setup and operation.
• Low resource usage.
• Auto-updates with network changes.
• Suitable for small networks.
Disadvantages:
• Slow convergence.
• Poor scalability for large networks.
• High bandwidth usage due to periodic updates.
• May choose suboptimal routes.
Count-to-Infinity Problem
• Occurs when a network route goes down or becomes unreachable.
• Routers keep increasing the hop count for that route in their updates (e.g., 14 → 15 →
16), slowly “counting to infinityˮ.
• This leads to routing loops and slow convergence, where routers mistakenly believe
the route is still reachable.
ii. Link state Protocol
• Also called Open Shortest Path First (OSPF) protocols.
• Use Hello messages (keep-alive) for discovering and maintaining neighbour routers.
• Use triggered updates for routers to send updates only when network topology
changes.
• Sends updates when topology changes occur.
• Tables used? Neighbour Table, Topology Table, Routing Table
Advantages
• Fast updates, quickly adapts to network changes.
• Accurate routing with full network visibility.
• Works well in large networks.
• Prevents routing loops.
• More reliable and stable routing.

35
 Disadvantages
• Requires more memory and CPU power.

famousrajbhatt@[Link]
• More complex to configure and maintain.
• Uses more bandwidth for routing updates.
• Not suitable for small networks due to overhead.
iii. Path Vector Protocol
• Dynamic exterior gateway protocol for routing between autonomous systems (AS).
• Uses path-vector mechanism to maintain and update route information dynamically.
• Selects the most efficient and reliable path for data transmission across networks.
Chooses the best path for reliable data transfer.
Advantages:
• Scalable for large, complex networks like the Internet.
• Supports multi-homing for better redundancy and reliability.
• Provides fine control over routing policies and traffic management.
• Quickly adapts to network failures and topology changes.
• Enables inter-domain routing across diverse networks.
Disadvantages
• Complex to configure and manage compared to interior protocols.
• Requires more CPU and memory resources on routers.
• Can be slower to converge during topology changes.
• Vulnerable to misconfigurations which can affect large parts of the Internet.

16. Address Resolution Protocol (ARP)

• Purpose: Maps an IP address to its corresponding MAC (hardware) address in a Local
Area Network (LAN).
• Role: Enables devices to communicate within the same network by resolving IP addresses
to MAC addresses.
• Operates at Network Layer (part of TCP/IP suite).
.

How it works?
i. Device wants to send data but only knows the IP.
ii. Checks ARP Cache for MAC address.
iii. If not found,
broadcasts ARP
Request asking who
has the IP.
iv. Target device replies
with its MAC address.
v. Sender saves MAC in
ARP Cache.
vi. Data is sent using the
MAC address.

36
17. NAT
• Purpose: Allows multiple private IPs to share a single public IP for internet access.

famousrajbhatt@[Link]
• Reason: IPv4 addresses
are limited. NAT allows
devices to share one public
IP by using private IPs,
helping avoid address
shortage.
• How it works:
o NAT runs on a
router/firewall at the
network border.
o Does both types of translation:
▪ Private IP → Public IP
▪ Public IP → Private IP
o Translates IP and port numbers using a NAT table.
• If no free address: Packets dropped, ICMP “host unreachableˮ message is sent.

37
 Transport Layer

famousrajbhatt@[Link]
1. Where it is located in OSI Model?
It is 4th layer from bottom; 3rd from top.

2. What is the data format? → Segments

3. What are the services offered by this layer?

• End-to-End control (port-to-port/ process-to-process)
• Reliable Data Delivery
• Error Control
• Flow Control
• Congestion Control
• Multiplexing and Demultiplexing

4. How does a transport layer work?

• Each application is assigned a port number for identification.
• Data from applications is split into segments with source & destination ports.
• Multiplexing allows many apps to send data simultaneously.
• The layer ensures reliable delivery using acknowledgments (TCP) or sends data quickly
without (UDP).
• Flow control manages the data rate to avoid overload.
• Error detection checks for damaged or lost segments and requests retransmission (TCP).
• At the receiver, demultiplexing delivers data to the correct application.

5. What is a Socket Address?

A socket address is a unique identifier for a network
connection.
Format:

Socket Address = IP Address + Port Number

This combination ensures that data is delivered to the correct application on the correct device.

i. Why is it required?
• The IP address identifies the device (e.g., a
computer or smartphone).
• The port number identifies the specific
application or service running on that device.
Together, they allow accurate and complete
routing of data from one application to another
across a network.

38
 ii. Why IP Address or Port alone is not enough?
• If Only IP is used: Cannot identify which application should receive the data.

famousrajbhatt@[Link]
• If Only Port is used: Cannot identify which device should receive the data.

Example:
• [Link]:80 ⇒ Web server
• [Link]:21 ⇒ FTP server
Same IP; different ports — different applications.

6. TCP
Transmission Control Protocol (TCP) is a connection-oriented protocol for communications
that helps in the exchange of messages between different devices over a network.
Features
i. Reliable Data Transfer
Ensures error-free, in-order delivery using acknowledgments and retransmissions.

ii. Connection-Oriented
Establishes a connection using a three-way handshake (SYN, SYN-ACK, ACK) and
closes it with a three-step process (FIN, FIN-ACK, ACK).

iii. Byte Streaming

• Data from the application layer is treated as a continuous stream of bytes.
• The transport layer divides this byte stream into segments, where one segment can
contain multiple bytes.

iv. Full Duplex Communication

TCP Supports simultaneous two-way data transmission between sender and receiver.

v. Piggybacking
When sending an acknowledgment (ACK), if there is also outgoing data, both are sent
together in the same segment to optimize efficiency. This is known as piggybacking.

vi. Flow Control

Manages the data rate using the receiverʼs buffer size to prevent overflow.

vii. Congestion Control

Uses algorithms like Slow Start, Congestion Avoidance, Fast Retransmit, and Fast
Recovery to manage network traffic.

viii. Error Detection

Uses a checksum in the TCP header to detect data corruption and request retransmission
if needed.

39
7. TCP Header
• Source Port (16 bits): Port number of the

famousrajbhatt@[Link]
sending application.
• Destination Port (16 bits): Port number of
the receiving application.
• Sequence Number (32 bits): Byte number
of the first byte in this segment; used for
reassembling out-of-order segments.
• Acknowledgement Number (32 bits):
Byte number which the receiver expects
next; acknowledges successful receipt of
previous bytes.
• Header Length (HLEN) (4 bits): TCP header size in 4 byte words; ranges from 5 (20 bytes)
to 15 (60 bytes).
To get the actual length, multiply by 4.
• Control Flags (6 bits): Manage connection and data flow:
o URG: Urgent pointer valid.
o ACK: Acknowledgement number valid.
o PSH: Push data immediately.
o RST: Reset connection.
o SYN: Synchronize sequence numbers (connection setup).
o FIN: Finish/Terminate connection.
• Window Size: Size of the senderʼs receive window (in bytes) for flow control.
• Checksum: Error-checking field (mandatory in TCP).
• Urgent Pointer: Points to urgent data byte (valid if URG flag set).
• Options & Padding: Optional extra data fields; used to send additional info like Maximum
Segment Size (MSS).
o MSS for Ethernet is typically 1460 bytes (1500 – 20 IP header – 20 TCP header).

8. Flags used in Communication

• SYN (Synchronize): Used to initiate a new TCP connection between two hosts.
• ACK (Acknowledge): Confirms that data or a connection request has been received.
• FIN (Finish): Gracefully closes a connection when one side has no more data to send.
• RST (Reset): Immediately terminates a connection due to errors or unexpected packets.

9. TCP Connection Establishment

i. Client sends SYN to request connection.
ii. Server responds with SYN and ACK to
acknowledge and agree.
iii. Client sends ACK to confirm connection
established.

40
 i. Client sends SYN
• Picks a random sequence number (e.g., 9000).

famousrajbhatt@[Link]
• Sets SYN flag = 1 to request
synchronization.
• Uses any source port (A) and destination
port 80 (HTTP).
• Announces window size (e.g., 12000 bytes).
• MSS (e.g., 1200 bytes) in options.
• Sends SYN segment to server.

ii. Server responds with SYN + ACK

• Picks its own random sequence number
(e.g., 3000).
• Acknowledges Clientʼs SYN by setting.
ACK = Clientʼs Seq + 1 (9001)
• Sets SYN = 1 to synchronize back.
• Sets ACK = 1 to acknowledge.
• Uses source port 80 and destination port A.
• Announces its window size (e.g., 18000
bytes) and MSS (e.g., 600 bytes).
• Sends SYN + ACK segment to client.

iii. Client sends final ACK

• Sends ACK with sequence number = Clientʼs Initial Seq + 1 (9001).
• Sets ACK number = serverʼs Seq + 1 (3001) to acknowledge server.
• Sets ACK flag = 1.
• Connection established and ready for data transfer.

10. TCP Data Transfer

i. Client to Server (Data Segment):
• Sending data bytes: 21-30 (10 bytes).
• Sequence Number: 21.
• ACK flag: 1 (acknowledging earlier segment).
• Acknowledgment Number: 71 (expecting byte
71 from server).

ii. Server to Client (Data Segment):

• Sending data bytes: 71-80 (10 bytes).
• Sequence Number: 71.
• ACK flag: 1.
• Acknowledgment Number: 31 (expecting byte
31 from client).

iii. Rule:
Acknowledgment Number always indicates the next byte the receiver expects.

41
 iv. TCP Acknowledgment
a. Pure Acknowledgment (No Data):
If no data is to be sent, a segment can carry only the acknowledgment:

famousrajbhatt@[Link]
• Only ACK flag is set.
• Prevents receiver from waiting.
• Does not consume a new sequence number.
• No payload.
• Indicates successful receipt of data.

b. Data + ACK Combined:

If sending data and acknowledging at the same time:
• Sequence Number: starts with the next byte.
• ACK flag: 1
• Acknowledgment Number: updated based on last data received (e.g., 81 if last received
byte was 80).
• This is called as piggybacking.

11. Number of Segments

• Window Size: 18000 Bytes
→ Total data that can be sent without waiting for ACK.
→ Reflects receiverʼs buffer capacity.
• MSS (Max Segment Size): 1800 Bytes
→ Max data per TCP segment (excluding header).
• Segments Needed:

Segments Needed = Window size / MSS

• In ideal conditions, Segments Needed = 18000 / 1800 = 10 segments (excluding header

overhead).

12. TCP Connection Termination

3 WAY
i. Client sends FIN = 1
• Indicates client wants to close its side of the
connection.
• Data may or may not be sent along with FIN.

ii. Server releases client-related resources

• Client → Server communication is closed (half-
close).
• Client cannot send data anymore, but can still send
ACKs without resources.

iii. Server responds with FIN = 1 and ACK = 1

• Server signals it also wants to close connection.
• Acknowledges clientʼs FIN.

42
iv. Client releases server-related resources
• Server → Client data sending stops.
• Server can still send ACKs without resources.

famousrajbhatt@[Link]
v. Client sends final ACK (ACK = 1)
• Confirms serverʼs FIN

vi. Server receives final ACK and closes the connection fully
• Both sides have released resources.
• Connection termination complete.

4 WAY
i. Client sends FIN = 1 to close its side.
ii. Server sends ACK = 1 to acknowledge FIN (connection
half closed).
iii. Server waits (may send/receive data) before closing.
iv. Server sends FIN = 1 when ready to close.
v. Client sends final ACK = 1 to acknowledge serverʼs FIN.
vi. Connection fully closed after final ACK.

TCP Connection termination example

Initial State: Ongoing Data Transfer
• Client → Server:
SEQ = 21, ACK = 1, ACK No = 81
(Client sending data starting from byte 21, acknowledging server's data up to byte 80)
• Server → Client (ACK Segment):
SEQ = 81, ACK = 1, ACK No = 31
(Server acknowledges receipt of data up to byte 30)
Step 1: Client initiates Connection Termination
• Client → Server (FIN Segment):
SEQ = 31, FIN = 1, ACK = 1, ACK No = 91
(Client wants to close the connection; acknowledges
Server data up to byte 90)

Step 2: Server responds with FIN

• Server → Client (FIN + ACK Segment):
SEQ = 91, FIN = 1, ACK = 1, ACK No = 32
(Server acknowledges Clientʼs FIN, and also sends its
own FIN)

Step 3: Client sends Final ACK

• Client → Server (Final ACK):
SEQ = 32, ACK = 1, ACK No = 92
(Client acknowledges server’s FIN)

43
13. User Datagram Protocol
• Connectionless & Unreliable.

famousrajbhatt@[Link]
• No handshake, no delivery/order/error guarantee.
• Lightweight & fast, low overhead.
• No flow or congestion control.
• Best for: Time-sensitive, loss-tolerant applications.
• Since UDP relies on IP, segments may follow different paths based on IP routing.

14. UDP Header

• Fixed size: 8 bytes (simple & efficient).
• No options field (unlike TCP).
UDP Header Fields (Each field is 2 bytes (16 bits)):
i. Source Port:
• Identifies the sender's port.
• Optional (can be 0 if not used).
ii. Destination Port:
• Identifies the receiver’s port.
• Required for delivery.
iii. Length:
• Total length of UDP header + Data.
• Minimum value: 8 bytes.
iv. Checksum:
• Used for error detection.
• Includes UDP header, data, and parts of the IP header (pseudo header).
• UDP checksum is optional in IPv4 but mandatory in IPv6.

15. Difference between TCP and UDP

TCP (Transmission Control Protocol) UDP (User Datagram Protocol)

Connection-oriented protocol Connectionless protocol

Reliable Not reliable

Supports robust error-checking mechanisms Basic error-checking using checksums only

Acknowledgment mechanism is present No acknowledgment segment

Slower compared to UDP Faster, simpler, and more efficient

Retransmission of lost packets is supported No retransmission of lost packets

Variable header length (20-60 bytes) Fixed header length (8 bytes)

Used in: Web browsing, email, file transfer Used in: Video calls, online games, streaming

Ex. HTTP, HTTPS, FTP, SMTP Ex. YouTube, Zoom, WhatsApp Calls, DNS

44
 Session Layer

famousrajbhatt@[Link]
1. Where it is located in OSI Model?
It is the 5th layer from bottom; 3rd from top.

2. What is the main purpose of this layer?

It is responsible for establishing, managing, and
terminating communication sessions.

3. What are the services offered by this layer?

i. Session Establishment
• Sets up and manages sessions between
communicating devices.
• Sessions are generally connection-oriented to
manage reliable dialog.
• Maps each session to the appropriate
transport layer connection.

ii. Communication Synchronization

• Ensures reliable communication using
checkpoints or synchronization bits in the
data stream.
• Facilitates recovery after errors or failures
by rolling back to the last known good state.

iii. Dialog Management

• Controls the direction of communication (who sends and who receives).
• Supports:
o Half-duplex mode: Uses a token mechanism—only the token holder can transmit.
o Full-duplex mode: Allows simultaneous two-way communication.

iv. Data Transfer

• Manages and monitors the exchange of data between systems within the session.
• Ensures orderly flow of information.

v. Resynchronization
• Restores session state after an interruption by resetting tokens and session
parameters.
• Three resynchronization modes:
o Set: Sets new synchronization points.
o Abandon: Discards current session progress.
o Restart: Restarts session from a previously saved state.

45
 Presentation Layer

famousrajbhatt@[Link]
1. Where it is located in OSI Model?
It is 6th layer from bottom; 2nd from top.

2. What is the main purpose of this layer?

Presents the data in a format that can be properly
interpreted by the sender and receiver.

3. What are the services offered by this layer?

i. Data Translation: Translates data between different
formats to ensure compatibility between systems.
It is also known as Translation Layer.

ii. Data Compression: Reduces data size to minimize the

number of bytes sent over the network.

Types
• Lossy Compression: Reduces file size by
permanently removing some data. Original file cannot
be restored.
• Lossless Compression: Reduces file size without any loss of data. Original file can
be restored.

iii. Encryption & Decryption: Converts data into an unreadable

format (encryption) and back to readable format (decryption)

iv. Syntax Semantics Management: Agrees on a common data syntax and semantics for
smooth communication between devices.

46
 Application Layer

famousrajbhatt@[Link]
1. Where it is located in OSI Model?
It is 7th layer from bottom; 1st from top.

2. What is the main purpose of this layer?

The application layer provides the functionality to send and
receive data from users. It acts as the interface between the user
and the application.

3. Most Imp Protocols

Protocol (Port) Full Form Purpose

Hyper Text Transfer Core protocol of the Web; transmits HTML;

HTTP (80)
Protocol client-server and stateless.

Translates domain names to IP addresses

DNS (53) Domain Name System
(e.g., [Link] → [Link]).

Allows remote command line access to

Telecommunications
TELNET (23) servers; allows clients to access Telnet
Network
server resources.

Dynamic Host Configuration Provides IP addresses and configuration

DHCP (67, 68)
Protocol information to hosts.

FTP (20 data, 21 Transfers files between devices; supports

File Transfer Protocol
control) reliable, efficient data sharing.

Simple Mail Transfer Pushes emails from sender to recipient

SMTP (25, 587)
Protocol (Push) mail server.

Pulls emails from mail server to client for

POP (110) Post Office Protocol (Pull)
offline access.

Allows remote hosts to mount and interact

NFS (2049) Network File System
with file systems as if local.

47
 Real-life examples

Protocol Real-World Working Example

famousrajbhatt@[Link]
When you open a browser and type [Link], your browser sends an
HTTP HTTP request to the web server, which responds with the webpage content (HTML,
images).

When you type [Link], your computer asks a DNS server "What's the IP
DNS for [Link]?" The DNS server replies with the IP, so your computer can
connect.

A network admin remotely logs into a server using Telnet to configure settings or
TELNET
check system status over a text-based session.

When you connect your laptop to Wi-Fi at a cafe, the DHCP server automatically
DHCP
assigns your device an IP address so it can communicate on the network.

You upload your website files to a web hosting server using an FTP client,
FTP transferring your HTML, images, and scripts. For secure transfer, FTPS or SFTP
should be used.

When you send an email from Outlook or Gmail, your email client uses SMTP to
SMTP
push the message to your mail server, which forwards it to the recipient's server.

You use an email client (like Thunderbird) to download emails from your mail server
POP
via POP, allowing you to read emails offline.

A developer accesses shared project files on a remote Linux server as if they were
NFS
on their own computer, using NFS to mount the remote directory locally.

4. HTTP
• Port Number: 80
• Uses TCP because HTTP itself is not reliable.
• In-band protocol (commands and data sent over the same connection).
• Each request is independent; the server does not retain client session information. Hence,
it is stateless.
HTTP Method Description
HEAD Retrieves only the headers of a resource.
GET Requests data from the server.
POST Submits data to be processed to the server.
PUT Updates or replaces a resource on the server.
DELETE Deletes a resource from the server.
CONNECT Establishes a tunnel, often for SSL connections.

5. HTTPS (Hyper Text Transfer Protocol Secure)

• HTTPS uses TLS (Transport Layer Security) to encrypt communication between client and
server.
• It typically runs over port 443 and provides data confidentiality, integrity, and
authentication.

48
6. DNS
• The Domain Name System (DNS) acts like the internet’s phone book.

famousrajbhatt@[Link]
• It maps domain names to their IP addresses.
• Some IP addresses are dynamic; DNS maps domain names to IPs regardless of type.
• DNS lets you change servers without changing the domain, keeping everything smooth
for users.
Example: You host [Link] on AWS. DNS maps the domain to its IP address.
Users type the domain, not the IP. Later, you switch to a cheaper host. The IP changes,
but you just update the DNS. Users still access [Link], no change on their end.

7. How it works?
We have a hierarchical structure for query-lookup.

DNS Hierarchy:
• Root Server: Top-level directory that points to
the correct TLD server based on the domain
extension (e.g., .com, .in).
• TLD Server: Manages all domains under that
extension and directs to the authoritative server
for the specific domain.
• Authoritative Server: Holds the actual DNS
records and returns the IP address for the domain.
Example: [Link] → [Link]

8. Types of DNS Query Lookups

A) Recursive
Your computer requests to resolve [Link] in the following steps:
1. It first checks its local cache for the IP address.
o If found, it uses it immediately.
o If not, it sends the query to the DNS resolver.
2. The resolver contacts the Root Server.
3. The Root Server responds with the address of the
.com TLD Server (since [Link] ends
with .com).
4. The resolver queries the .com TLD Server, which returns the address of the
authoritative DNS server for [Link].
5. The resolver contacts the authoritative server for [Link], which returns the
IP address, for example, [Link].
6. The authoritative server responds to the resolver directly.
7. It travels back to DNS Resolver.

8. The resolver caches it inside the local cache for the future requests and then sends the
IP address back to the computer.

49
 B) Iterative
Your computer requests to resolve [Link] in the following steps:

famousrajbhatt@[Link]
1. It checks the local cache for the IP
address.
2. If not found, the DNS resolver contacts
the Root Server.
3. The Root Server replies with a pointer
to the .com Top-Level Domain (TLD)
server.
4. The resolver queries the TLD server.
5. TLD responds with authoritative server.
6. Resolver queries the authoritative server.
7. Authoritative server replies with IP.
8. The resolver caches this IP address.
Finally, the resolver sends the IP address back to your computer.

9. FTP
• File Transfer Protocol
• Used to transfer files.
• It is out-of-band protocol. It operates on separate ports for data and control channels.
• Data connection = non-persistent (20)
• Control connection (port 21) remains open throughout the session; data connection (port
20) is opened as needed.
• Reliable and stateful protocol.

10. Email services

SMTP POP IMAP

Synchronizes emails
Sends or pushes emails Downloads or pulls emails
between server and client

Handles outgoing mail Handles incoming mail Handles incoming mail

SMTP typically works

synchronously to ensure Usually synchronous Usually synchronous
delivery acknowledgment.

Used to send mail from client to Fetches mail from server Accesses mail directly on the
server or between servers to client server

Emails remain on server,

Keeps emails on server Downloads and often
accessible from multiple
(usually) deletes emails from server
devices

50
11. Difference between Persistent HTTP and Non-persistent HTTP

Persistent HTTP (HTTP/1.1) Non-Persistent HTTP (HTTP/1.0)

famousrajbhatt@[Link]
Server keeps the connection open after Server closes the connection after sending
sending response response

Requires 1 RTT to request referenced objects Requires 2 RTTs for each object

More overhead (opens new connection for

Less overhead (reuses same connection)
each object)

Uses TCP connection Uses TCP connection

51
 Delays

famousrajbhatt@[Link]
Delays in computer networks, also called Nodal Delays, refer to the time taken during different
stages of processing and transferring a packet. There are four main types of delays:

1. Transmission Delay (Tt)

Definition: Time required to push all bits of the packet onto the transmission medium.
Formula:

Tt = L/B

Where:
• L = Packet size (in bits).
• B = Bandwidth (in bits per second (bps)).
Example:
If bandwidth B = 1B = 1 bps and Packet size L = 20L = 20 bits, then:
L 20 bits
Tt = = 1 bps = 20 seconds
B

Factors affecting transmission delay:

• Multiple active sessions increase delay.
• Higher bandwidth reduces delay.
• MAC protocols affect delay in shared mediums.
• OS context switching adds to delay.

2. Propagation Delay (Tp)

Definition: Time taken by the last bit to travel from sender to receiver through the medium.
Formula:

Distance
Tp =
Velocity

Factors affecting propagation delay:

• Distance: More distance → More delay.
• Velocity: Higher velocity → Less delay.

3. Queueing Delay (Tq)

Definition: Time a packet waits in a buffer queue before being processed.
NOTE: No exact formula exists, as it varies based on real-time traffic and system
conditions.

Factors affecting queueing delay:

• Longer queues → Higher delay.
• Bursty or high traffic → Increased delay.
• Fewer servers/links → More queueing.

52
4. Processing Delay (Tpro)
Definition: Time taken by routers or destination systems to process the packet (e.g., check

famousrajbhatt@[Link]
headers, update TTL, etc.).
NOTE: No formula, as it depends on processor speed and system load.

Factors affecting processing delay:

• More CPU speed→ less processing delay
• Low System performance → high processing delay

5. Total Delay
Definition: Sum of all delays.
NOTE: It is also called as with negligible Delay.

Ttotal = Tt + Tp + Tpro + Tq

In ideal cases (no processing and queuing delay)

Ttotal = Tt + Tp

53
Interview Questions:
1. Can you name and briefly describe any five common network protocols and their

famousrajbhatt@[Link]
purposes?
2. Which protocol ensures reliable communication between two endpoints on a network?
3. What is the difference between LAN, MAN, and WAN? Give examples.
4. Where would you typically use a PAN or CAN?
5. How does a client-server model differ from a peer-to-peer model?
6. Which architecture is more scalable and why?
7. What are the different types of network topologies and their advantages/disadvantages?
8. Why is star topology more commonly used than bus topology in modern networks?
9. What are the functions of the following devices: Hub, Switch, Router, Modem, Repeater?
10. How is a switch different from a hub in handling network traffic?
11. How does a router determine the best path to route a packet?
12. What’s the difference between a router and a switch in terms of OSI layer and
functionality?
13. What is a firewall and how does it protect a network?
14. Can a firewall block internal threats? Why or why not?
15. Explain the TCP 3-way handshake step by step.
16. What happens if the final ACK is lost during the handshake?
17. What are the four types of delays in a network? Explain each briefly.
18. Which delay is affected by congestion in the network?
19. What are the main differences between IPv4 and IPv6?
20. What is the purpose of private IP addresses?
21. Can a device with a private IP access the internet directly? Why or why not?
22. Describe step-by-step What happens when you hit enter after typing [Link] in
your browser.
23. What is the difference between unicast, multicast, broadcast, and anycast? Give real-life
examples.
24. Why is a switch preferred over a hub in modern LANs?
25. In what form is data represented at each layer of the OSI model (bits, frames, packets,
etc.)?
26. How does subnetting help in efficient IP address management?
27. What is DNS and how does it resolve domain names to IP addresses?
28. What is the difference between recursive and iterative DNS queries?

54
29. Why does a device need both MAC and IP addresses?
30. What is flow control in networking? Name two protocols that provide flow control.

famousrajbhatt@[Link]
31. What are two methods used for error detection and correction?
32. How does checksum differ from CRC in detecting errors?
33. Compare TCP and UDP in terms of reliability, speed, and use cases.
34. Which protocol would you use for live video streaming and why?
35. What are the responsibilities of the session layer in the OSI model?
36. What functions does the presentation layer perform in data communication?
37. Why do we need data encryption?
38. Explain how propagation delay and transmission delay differ.
39. Which delay is dependent on bandwidth and which on distance?
40. What is a VPN and how does it work?
41. What are some advantages and limitations of using a VPN?

MCQs:
1. Which protocol ensures reliable 5. Which architecture offers
communication? centralized control and better
A. UDP security?
B. IP A. Peer-to-peer
C. TCP B. Client-server
D. FTP
6. In peer-to-peer networks, devices
2. Which protocol is used to transfer are both:
files over a network? A. Clients only
A. HTTP B. Servers only
B. FTP C. Clients and Servers
C. DHCP D. None of the above
D. SNMP
7. Which topology connects all nodes
3. Which network is used within a to a central device?
single room or device range? A. Bus
A. PAN B. Ring
B. LAN C. Star
C. MAN D. Mesh
D. WAN
8. Which topology provides full
4. Campus-wide network is an
redundancy?
example of:
A. Mesh
A. WAN
B. Star
B. MAN
C. Bus
C. CAN
D. Ring
D. PAN

55
9. Which device works at the Data Link 17. If the last ACK in handshake is lost,
Layer (Layer 2)? the sender:
A. Router A. Closes connection

famousrajbhatt@[Link]
B. Switch B. Retransmits SYN
C. Modem C. Retransmits ACK
D. Repeater D. Waits or times out
10. What does a router do? 18. Which delay is caused by the time
A. Broadcasts data to push data into the link?
B. Forwards packets between networks A. Propagation
C. Encrypts data B. Queuing
D. Stores web pages C. Transmission
D. Processing
11. Switches are better than hubs
because they: 19. Congestion mainly affects which
A. Use more power type of delay?
B. Flood all ports A. Processing
C. Forward data intelligently B. Propagation
D. Are cheaper C. Queuing
D. Transmission
12. Which device connects two
different networks? 20. IPv6 address size is:
A. Switch A. 32-bit
B. Repeater B. 64-bit
C. Router C. 128-bit
D. Hub D. 256-bit
13. Routers operate on which OSI layer? 21. What is a benefit of using private IPs?
A. Transport A. Global accessibility
B. Network B. Cost saving
C. Session C. Easy routing
D. Data Link D. No need of NAT
14. What is the primary function of a firewall?
22. Can a private IP access the Internet
A. Speed up connection
directly?
B. Filter network traffic
A. Yes
C. Provide IP addresses
B. No
D. Store files
23. What is the first thing that happens
15. Firewalls can block:
when we visit [Link]?
A. Outgoing only
A. TCP connection
B. Incoming only
B. DNS resolution
C. Both incoming and outgoing
C. HTTP request
D. None
D. IP assignment
16. The first step in a TCP 3-way
24. Which is an example of one-to-
handshake is:
many communication?
A. SYN
A. Unicast
B. ACK
B. Multicast
C. SYN-ACK
C. Anycast
D. FIN
D. Broadcast

56
25. Which communication sends data 31. MAC address is unique to:
to all nodes in the network? A. IP
A. Unicast B. User

famousrajbhatt@[Link]
B. Broadcast C. Device
C. Multicast D. Port
D. Anycast
32. IP address is used for:
26. A VPN provides: A. Local communication only
A. Physical security B. Internet routing
B. Encrypted remote access C. Encryption
C. Faster internet D. Hardware control
D. Static IP
33. Flow control prevents:
27. Data at the Network layer is called: A. Packet loss from congestion
A. Frame B. Address resolution
B. Segment C. Network discovery
C. Packet D. IP conflict
D. Bit
34. Which protocol provide no flow
28. Subnetting is used to: control or error correction?
A. Combine Ips A. TCP
B. Allocate bandwidth B. UDP
C. Divide IP space efficiently C. FTP
D. Reduce security D. ICMP
29. DNS converts: 35. Session layer is responsible for:
A. IP to MAC A. Logical addressing
B. Domain to IP B. End-to-end delivery
C. Port to MAC C. Dialog control
D. IP to Domain D. Data encryption
30. Which DNS query type returns the
full IP resolution to the client?
A. Recursive
B. Iterative
C. Redundant
D. Cache-only

ANSWER KEY
1. (C) 2. (B) 3. (A) 4. (C) 5. (B)
6. (C) 7. (C) 8. (A) 9. (B) 10. (B)
11. (C) 12. (C) 13. (B) 14. (B) 15. (C)
16. (A) 17. (D) 18. (C) 19. (C) 20. (C)
21. (B) 22. (B) 23. (B) 24. (B) 25. (B)
26. (B) 27. (C) 28. (C) 29. (B) 30. (A)
31. (C) 32. (B) 33. (A) 34. (B) 35. (C)

57