TCP

TCP stands for Transmission Control Protocol.
It is a connection-oriented protocol that allows devices to communicate reliably over a network.

Key Features of TCP

• Ensures data reaches the destination correctly

• Maintains correct order of data

• Detects and retransmits lost packets

• Works well even when the network is slow or unreliable

Used : Web browsing, emails, file transfers.

TCP 4-Layer Model

The TCP/IP model is a 4-layer networking model used in real-world internet communication.

1. Application Layer

Purpose:
Allows users to interact with network-connected applications.

Functions:

• Data formatting

• Application-level communication

Protocols:

• HTTP / HTTPS – Web

• SMTP – Email

• FTP – File transfer

2. Transport Layer

Purpose:
Manages end-to-end communication between devices.

Functions:

• Data segmentation

• Port addressing

• Flow control and reliability

Protocols:

• TCP – Reliable, ordered delivery

• UDP – Fast, best-effort delivery

3. Internet Layer

Purpose:
Responsible for logical addressing and routing across networks.

Functions:

• Assigns IP addresses

• Routes packets between networks

Protocols:

• IP (IPv4 / IPv6)

• ICMP

4. Network Access Layer

Purpose:
Handles physical data transmission over the network.

Functions:

• Frame creation

• MAC addressing

• Signal transmission

Technologies:

• Ethernet

• Wi-Fi

• ARP

Problems in IP Networks & How TCP Solves Them

Packet Loss → Reliable Delivery

Problem:
Some data packets may disappear on the way.

TCP Solution:

• Receiver sends an ACK (acknowledgement) for received data

• If sender doesn’t get an ACK → it resends the data

Result: No data is lost.

Out-of-Order Packets → Sequencing

Problem:
Packets can take different paths and arrive in the wrong order.

TCP Solution:

• Each packet gets a sequence number

• Receiver rearranges packets correctly

Result: Data is delivered in the right order.

Data Corruption → Error Checking

Problem:
Data can get damaged during transmission.

TCP Solution:

• Uses a checksum to detect corruption

• Corrupted data is discarded

• Sender is asked to resend

Result: Only correct data is accepted.

Network Congestion → Congestion Control

Problem:
Sending too much data too fast can overload the network.

TCP Solution:

• Starts sending data slowly (slow start)

• Increases speed if network is fine

• Slows down if packet loss is detected

Result: Network doesn’t get overwhelmed.

Receiver Overload → Flow Control

Problem:
A fast sender can overwhelm a slow receiver.

TCP Solution:

• Receiver tells sender how much data it can handle

• Uses a sliding window mechanism

Result: Sender sends only what receiver can process.

Connection Management → Connection-Oriented

Problem:
Data shouldn’t be sent unless both sides are ready.

TCP Solution:

• Uses 3-way handshake to start communication

  • SYN → SYN-ACK → ACK

• Uses a 4-step process to close connection safely

Result: Both sides stay synchronized.

Duplicate Data → Detection

Problem:
Same packet may arrive more than once.

TCP Solution:

• Uses sequence numbers to identify duplicates

• Discards extra packets

Result: No repeated data.

Why TCP Is Primary Use

Because TCP guarantees accuracy and completeness, it is used where data must not break:

• Web browsing (HTTP / HTTPS)

• Email (SMTP, IMAP, POP3)

• File transfers (FTP, SFTP)

• Remote login (SSH)

TCP 3-Way Handshake

The 3-way handshake is the process used by TCP to establish a connection before data transfer begins.

Step 1: SYN (Synchronize)

• The client wants to establish a connection with the server

• The client sends a TCP segment with the SYN flag set

• This segment includes the initial sequence number chosen by the client

Meaning:
“I want to communicate, and this is the sequence number I will start with.”

Step 2: SYN + ACK (Synchronize Acknowledge)

• The server receives the SYN segment

• The server responds with a segment that has both SYN and ACK flags set

• ACK acknowledges the client’s SYN

• SYN contains the server’s own initial sequence number

Meaning:
“I received your request, I’m ready too, and this is my starting sequence number.”

Step 3: ACK (Acknowledge)

• The client sends an ACK segment back to the server

• This acknowledges the server’s sequence number

Meaning:
“I received your response. Connection is established.”

After the Handshake

• A reliable TCP connection is now established

• Both sides know each other’s sequence numbers

• Actual data transfer can begin

Steps of TCP Data Transfer

Step 1: Segment Generation

• Data moves from the application layer to TCP

• TCP adds a header (ports, sequence numbers)

• Data is split into segments

Step 2: IP Encapsulation

• TCP segments are passed to the IP layer

• IP wraps them into packets

• Packets are routed across the network

Step 3: Reception and Acknowledgment

• Receiver collects incoming segments

• Reorders them using sequence numbers

• Sends ACKs to confirm successful delivery

Step 4: Connection Termination

When communication is complete:

• One side sends a FIN signal

• The other side acknowledges it

• Both sides close the connection gracefully

How TCP Ensures Reliability, Correctness, and Order

TCP achieves this through connection management, sequence numbers, acknowledgments, error checking, and control mechanisms.

1. Reliability (Ensuring Data Is Delivered)

TCP guarantees that no data is lost during transmission.

How it works:

• TCP uses positive acknowledgment with retransmission (PAR).

• Every piece of data sent must be acknowledged (ACKed) by the receiver.

• If an acknowledgment is not received, TCP assumes the data was lost and resends it.

Key mechanisms:

• Sequence numbers
  Each byte of data is numbered. This helps TCP track what has been sent and what has been received.

• Acknowledgments (ACKs)
  The receiver sends back an ACK indicating the next byte it expects.

• Timeout and retransmission
  If the sender does not receive an ACK within a certain time, it retransmits the data.

• Fast retransmit
  If the sender receives multiple duplicate ACKs, it quickly resends the missing data without waiting for a timeout.

• Three-way handshake
  TCP first establishes a connection (SYN, SYN-ACK, ACK) to ensure both sides are ready before sending data.

Result:
All data sent by the sender eventually reaches the receiver.

2. Correctness (Ensuring Data Is Not Corrupted)

TCP ensures that data is received exactly as it was sent, without errors.

How it works:

• Each TCP segment includes a checksum.

• The sender calculates the checksum before sending the data.

• The receiver recalculates the checksum after receiving it.

What happens if data is corrupted:

• If the checksum does not match, the packet is considered corrupted.

• The receiver discards the packet and does not send an ACK.

• The sender notices the missing ACK and retransmits the data.

Result:
Only correct, error-free data is delivered to the application.

3. Order (Ensuring Data Is Delivered in the Correct Sequence)

Packets on the internet may arrive out of order because they can take different routes.

How TCP fixes this:

• Each segment carries a sequence number.

• The receiver uses these numbers to reassemble data in the correct order.

• Out-of-order packets are temporarily stored in a buffer until missing packets arrive.

Result:
The application always receives data in the exact order it was sent.

4. Flow Control and Congestion Control (Preventing Overload)

TCP also prevents systems and networks from being overwhelmed.

Flow Control (Receiver Protection)

• The receiver tells the sender how much data it can handle using a window size.

• The sender limits how much data it sends based on this value.

This prevents a fast sender from overwhelming a slow receiver.

Congestion Control

• TCP monitors network conditions.

• If packet loss occurs, TCP assumes congestion.

• It reduces the sending rate using mechanisms like Slow Start and Congestion Avoidance.

This keeps the network stable and avoids congestion collapse.

How a TCP Connection Is Closed

A TCP connection is closed using a 4-way handshake, which allows both sides to finish sending data and agree to terminate the connection safely. Because TCP is full-duplex, each direction of communication is closed independently.

1. FIN — Close Request

One side (client or server) sends a FIN (Finish) segment.

• Meaning: “I have no more data to send.”

• This closes only the sender’s outgoing data stream.

2. ACK — Acknowledgment

The other side replies with an ACK.

• Meaning: “I received your FIN.”

• The connection is now half-closed.

• The receiver can still send remaining data.

3. FIN — Close Response

When the second side finishes sending its data, it sends its own FIN.

• Meaning: “I am also done sending data.”

4. ACK — Final Acknowledgment

The original sender responds with a final ACK.

• The logical connection is now closed.

• The sender enters the TIME-WAIT state.

Important Concepts

Half-Close

• Each FIN closes one direction of the connection.

• Data can still flow in the opposite direction until its FIN is sent.

TIME-WAIT State

After sending the final ACK, TCP waits for a short period:

• Ensures the last ACK is received

• Handles delayed or duplicate packets

• Prevents old packets from interfering with new connections

Abrupt Close (RST)

• If a connection must be terminated immediately, a RST (Reset) is sent.

• No handshake, no graceful shutdown.

• Used for errors or unexpected conditions.