Statistical time-division multiplexing (STDM) is sharing a communication like TDM but like the TDM the time slots are not allocated to any channel. In the case of TDM, time slots are allocated to channels and reserved even if there is no data to send. This allocation is just wastage of the bandwidth and Statistical time-division multiplexing overcome this inefficiency of standard TDM. Statistical time-division multiplexing achieved this with the use of intelligent devices that are capable of identifying when a terminal is idle.
STDM was specially developed to overcome the inefficiency of TDM. It uses a variable time slot length allowing channels to use any free slot space for sending their information. It uses a buffer memory for temporarily stores the data during periods of peak traffic. STDM requires each transmission to carry identification information or a channel identifier.
Statistical time-division multiplexing is mainly the same as TDM except that every signal is assigned a slot based on priority and demand. Standard TDM and various other circuits switching technique are executed at the physical layer in the OSI and TCP/IP model but Statistical time-division multiplexing is executed at the data link layer and above.
The STDM is used to transport multiple data, audio and video streams with different data rates to be broadcasted across a bandwidth-limited channel. The example is the Frame relay packet-switching and X.25 protocols, in which the packets have different lengths. The STDM is the example of Asynchronous Transfer Mode packet-switched protocol.
In the Asynchronous Transfer Mode, single streams are classified in variable time segments and subsequently transmitted using the asynchronous time-division multiplexing procedure. There no define the order for the bit transmission in this mode. Each time slot carries channel information to separate the packet in the demultiplexer procedure. The figure below illustrates the STDM process.