Illustrated TCP/IP by Matthew G. Naugle Wiley Computer Publishing, John Wiley & Sons, Inc. ISBN: 0471196568   Pub Date: 11/01/98

Chapter 188
Multiplexing and Demultiplexing

UDP accepts data from the application layer, formats it (UDP header) with its information, and presents it to the IP layer for network delivery. UDP will also accept data from the IP layer and, depending on the port value, present it to the appropriate application. As shown in the slide, UDP is responsible for directing the rest of the packet (after stripping off its headers) to the correct process according to the port number assigned in the UDP header. This process is called demultiplexing. There are many different types of port numbers to indicate any application running on the network station. UDP reads the Destination Port field of the UDP header (demultiplex) and gives the data to the application. When the application (identified by the port number) initializes, the station’s operating system works in conjunction with it and provides a buffer area in which information may be stored. UDP will place the data in this area for retrieval by the application. UDP does provide one error mechanism for ports that are not valid. It can generate an ICMP Port Unreachable message to be sent to the originator of the packet.

Since the TCP/IP protocol suite includes applications that are specifically written to it (TFTP, Domain Name Service, etc.), there are statically assigned port numbers that identify these applications. Certain port numbers are reserved and cannot be used by any unknown application. The reserved port numbers are specified in RFC 1700.

Multiplexing and Demultiplexing