Section 5: OSI Layer 4 - The Transport Layer

In the TCP header, which 16-bit field occupies bits 0–15 (blanked out in the diagram)?

Source Port

HTTP uses port [...].

80

SMTP email uses port [...].

25

The combination of [...] and destination port numbers can be used to track sessions.

Source

The most common Layer 4 protocols are TCP and [...].

UDP

The process by which a host is able to support multiple sessions simultaneously and manage the individual traffic streams over a single link. 

Session Multiplexing

The process of adjusting the flow of data from the sender to ensure that the receiving host can handle all of it.

Flow Control

This OSI layer provides transparent transfer of data between hosts and is responsible for end-to-end error recovery and flow control.

Transport Layer

This transport layer protocol carries out sequencing to ensure segments are processed in the correct order and none are missing.

TCP

This transport layer protocol does not carry out sequencing to ensure segments are processed in the correct order and none are missing.

UDP

What does the acronym UDP stand for?

User Datagram Protocol

Which transport layer protocol does not perform flow control? TCP or UDP?

UDP

Which transport layer protocol is connection oriented? TCP or UDP?

TCP

Which transport layer protocol is not reliable, meaning the receiving host does not send acknowledgments back to the sender?

UDP

Which transport layer protocol is reliable where lost segments are resent? TCP or UDP?

TCP

Which transport layer protocol is used by Telnet?

TCP

Which transport layer protocol is used by TFTP?

UDP

Which transport layer protocol is used by Telnet?

TCP

Which transport layer protocol sends traffic best effort? TCP or UDP?

UDP

At the Transport layer, what key piece of information is included that identifies which application traffic belongs to?

The port number

Why does a device check the Layer 4 header (port number) after already verifying the destination IP at Layer 3?

To determine which specific application the traffic is intended for (e.g., a server running both a web service on port 80 and an email service on port 25).

What's the key tradeoff between using TCP vs. UDP at the Transport layer?

TCP provides reliability; UDP prioritizes speed over reliability. TCP is used when reliable delivery matters; UDP is used for traffic like voice/video where speed matters more.

Why does the Transport layer break large files into smaller segments?

Smaller segments are less likely to incur transmission problems than one large block of data.

What are the three steps of the TCP three-way handshake?

SYN (sender initiates) → SYN-ACK (receiver acknowledges and synchronizes) → ACK (sender acknowledges) — connection is then established.

In a typical client-to-server connection, what source port does the client use, and what determines the destination port?

The client uses a random source port above 1024; the destination port is the standard port for the service being accessed (e.g., port 80 for HTTP).

How does a stateful firewall use port numbers to allow return traffic while blocking unsolicited inbound traffic?

It tracks the source/destination port combination of outbound connections it allowed, and permits matching return traffic back through — but blocks traffic that wasn't initiated from the trusted side.

What fields are included in the TCP header that are NOT present in the UDP header? (name at least 3)

Sequence number, acknowledgment number, header length, reserved field, code bits, window, urgent pointer, options.

What fields make up the UDP header?

Source port, destination port, length, checksum, and data.

Why does UDP have significantly less overhead than TCP?

Its header is much smaller/simpler — it lacks the connection setup, sequencing, acknowledgment, and flow control fields TCP requires.

Why do real-time applications like voice and video typically use UDP instead of TCP?

They are highly sensitive to latency/delay, and TCP's extra overhead (handshake, acknowledgments, retransmission) would degrade real-time quality.

If UDP doesn't provide reliability or error recovery, how can an application using UDP still achieve reliable delivery?

Error detection and recovery must be implemented at the upper layers (application level) instead — UDP itself does not provide it.

What are the standard TCP port numbers for: FTP, SSH, Telnet, HTTP, HTTPS?

FTP – 21, SSH – 22, Telnet – 23, HTTP – 80, HTTPS – 443

What are the standard UDP port numbers for TFTP and SNMP?

TFTP – 69, SNMP – 161

Which well-known application protocol can use both TCP and UDP, and what port does it use?

DNS, port 53