L4 Transport Layer Flashcards
What are the main services provided by the transport layer?
Logical communication between processes; Multiplexing and demultiplexing; Integrity checking; Connection management; Acknowledgments and retransmissions; Flow control
How does the transport layer differ from the network layer?
Transport layer provides process-to-process communication, whereas network layer provides host-to-host communication.
What are the two main transport protocols in the Internet?
Transmission Control Protocol (TCP) and User Datagram Protocol (UDP)
What are the characteristics of UDP?
Connectionless, unreliable, unordered delivery, minimal error checking
What are the characteristics of TCP?
Reliable delivery, in-order delivery, flow control, congestion control, connection establishment required
What is multiplexing in the context of the transport layer?
Gathering data from multiple application processes and sending them as transport segments.
What is demultiplexing in the context of the transport layer?
Delivering incoming transport segments to the correct application process.
How are UDP sockets identified?
By a 2-tuple: (destination IP address, destination port number)
How are TCP sockets identified?
By a 4-tuple: (source IP, source port, destination IP, destination port)
How do web servers use port numbers?
Clients send segments to destination port 80; server distinguishes clients using source IP and source port.
What happens with persistent versus non-persistent HTTP connections?
Persistent HTTP uses one TCP connection for multiple requests; non-persistent HTTP opens/closes a connection for each request/response.
What information does the UDP segment structure include?
Source port number, Destination port number, Length, Checksum
How is checksum used in UDP?
Sender calculates checksum over segment contents; receiver verifies checksum to detect errors.
What is depicted in the diagram showing hosts A, B, and C communicating with a server?
Transport layer multiplexing and demultiplexing using TCP 4-tuples.
How does the server (IP B) distinguish between different client connections?
By using the source IP address and source port number along with destination IP and port.
What are the components of the TCP 4-tuple used for socket demultiplexing?
Source IP, Source Port, Destination IP, Destination Port
Why do different client segments, even if targeting the same port 80, get directed to different sockets?
Because they have different source IPs and/or source ports.
What role does the transport layer play in relation to the network layer in this image?
It provides logical process-to-process communication, while network layer provides host-to-host delivery.
In the diagram, what application processes (P2, P3, P4, P5, P6) are involved?
They represent different server processes or threads managing client sessions.
What key concept is demonstrated by the note ‘three segments, all destined to IP address B, dest port 80 are demultiplexed to different sockets’?
Connection-oriented demultiplexing in TCP using 4-tuples.
What does the TCP/UDP segment format diagram illustrate?
The segment structure including source port, destination port, header fields, and payload.
What is the size of the combined source and destination port fields in a TCP/UDP segment?
32 bits total (16 bits each)
Where are the source and destination ports located in the TCP/UDP segment?
At the beginning of the transport segment header.
What follows the source and destination port numbers in a TCP/UDP segment?
Other header fields and then the application data.
