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

Chapter 187
User Datagram Protocol (UDP)

A transport layer allows communication to exist between network stations. Data is handed down to this layer from an upper-level application. The transport layer then envelopes the data with its headers and gives it to the IP layer for transmission onto the network. There are two transport-layer protocols in TCP/IP: UDP and TCP.

The functionality of UDP should sound familiar. It is a connectionless, unreliable transport service. It does not issue an acknowledgment to the sender upon the receipt of data. It does not provide order to the incoming packets, and may lose packets or duplicate them without issuing an error message to the sender. This should sound like the IP protocol. The only offering that UDP has is the assignment and management of port numbers to uniquely identify the individual applications that run on a network station and a checksum for simplex error detection. UDP tends to run faster than TCP, for it has low overhead (8 bytes in its header compared to TCP’s typical 40 bytes). It is used for applications that do not need a reliable transport. Some examples are network management, name server, or applications that have built-in reliability.

Any application program that incorporates the use of UDP as its transport-level service must provide an acknowledgment and sequence system to ensure that packets arrive, and that they arrive in the same order as they were sent.

As shown in the slide, an application’s data is encapsulated in a UDP header. The transport layer has its own header, independent of all other layers, that it prefaces to the data handed to it from its upper-layer protocol. The UDP header and its data are then encapsulated in an IP header. The IP protocol would then send the datagram to the data-link layer, which would then encapsulate the datagram with its headers (and/or trailers) and send the data to the physical layer for actual transmission.

Upon receipt of the packet, the datalink would interpret the address as its own, strip off its header (and/or trailers), and submit the packet to the IP layer. IP would accept the packet based on the correct IP address in the IP header, strip off its header, and submit the packet to the UDP-layer software. The UDP layer accepts the packet and now has to demultiplex the packet based on the port number in the UDP header.

Looking at the slide, the packet header for UDP is small (the minimum packet size is 8 bytes), but functional. The message length indicates the size of the UDP header and its data in bytes. The checksum is used to check for the validity of the UDP header and data. It does not have to be implemented and would be set to 0 if not implemented. UDP is primarily used by applications to simply provide for an application MUX at the transport layer. This is described next.

User Datagram Protocol (UDP)