from Wikipedia, the free encyclopedia

Multiplex processes (lat. Multiplex "multiple, diverse") are methods of signal and message transmission in which several signals are combined (bundled) and transmitted simultaneously via one medium ( line , cable or radio link ). Multiplex processes are often combined in order to achieve an even higher level of utilization. The bundling takes place after the user data has been modulated onto a carrier signal . They are correspondingly demodulated at the receiver after unbundling ( demultiplexing ).

Illustration of an LED matrix control with synchronous line multiplexing

In particular, the synchronous time division multiplex method (see below) is also used for serial digital data transmission , the simultaneous data acquisition of several analog channels and for the control of displays (LED and fluorescent displays , LCD , plasma and CRT screens and displays).

The intentions for the use of multiplexing are partly different for wired and radio transmission. In the case of wired transmission, the signals from several sources are bundled by a so-called multiplexer and transmitted together over one instead of several parallel paths. The demultiplexer then unbundles the signals again. The aim here is to keep the costs for the transmission link as low as possible. Radio technology is particularly suitable for being able to connect several participants, who are usually also spatially distributed, to a central radio station at the same time. With directional radio as a point-to-point connection, there are also multiplex technologies. Here, the high-frequency transmission channel is used jointly for a large number of data streams.


Multiplex processes were developed in order to achieve optimal utilization of the lines and frequencies that are available as transmission paths in electronics and communication technology .
This reduces costs and increases reliability, since, for example, fewer connecting and connecting lines are required. Some technical solutions can only be implemented with multiplex signal transmission (e.g. reading and writing to individual pixels from digital cameras or flat screens ).

In the multiplex process, several different signals are bundled or temporally interleaved in order to be able to transmit them simultaneously and jointly without influencing one another.

In communication technology and signal transmission, a distinction is made between the following multiplex processes:

Space division multiplexing
Transmission channels (lines, radio links) are bundled for parallel but exclusive use by several transmitters and receivers.
Frequency or wavelength division multiplexing
In the case of line transmission, several signals are transmitted separately in different frequency ranges; With radio and fiber optic transmission , different signals are assigned different wavelengths. Since frequency and wavelength are firmly linked to one another via the speed of propagation , frequency and wavelength multiplex are synonymous processes.
Time division multiplexing
Several signals are transmitted with a time delay. They are nested in one another in time. The time windows can be synchronized and of the same length or asynchronous and depending on requirements.
Code division multiple access
This method is used in radio technology and in data buses . Different signal sequences are transmitted via a line or a radio frequency and can be assigned based on their different coding ; only the appropriately coded signal is recognized and evaluated in the receiver.

Analogies for illustration

Space division multiplexing
People talk to each other in different places. If the distance is large enough, the conversations do not interfere with one another.
Frequency division multiplexing
A dog whistle or bats generate signals that are inaudible to humans. Communication is possible in parallel.
Time division multiplexing
In a school class or with CB radio , usually only one speaker has the word at the same time (asynchronous time division multiplexing), in parliaments each speaker has a speaking time of a defined length (synchronous time division multiplexing)
Code division multiple access
If you have many languages ​​spoken at the same time, you can hear your mother tongue. You can recognize well-known people by the sound of their voice.

Difference between the terms "multiplexing" and "multiple access"

In connection with the herein described multiplexing method is part of Multiplexing (often in conjunction with hardware) and partly by Multiple Access (often as a software solution) spoken.

To Multiplexing in the narrower sense is if at the beginning of a transmission path, a multiplexer multiple signals bundles and an end demultiplexer these separating again. In the field of audio / video processing, one speaks of multiplexing when a video track and an audio track are combined into a single file (e.g. * .mpeg) using the appropriate software.

The term multiple access is used more when there are several transmitter-receiver pairs (e.g. mobile radio subscribers) that use a transmission medium (usually a radio interface - i.e. the air in the vicinity of a base station or a radio router) independently split up. This is done either with a central entity (e.g. the base station for mobile telephony ) that assigns the channels, or the subscriber devices work with collision detection .

However, the procedures for sharing the transmission medium are the same for multiplexing and multiple access and are therefore described together in this article.

Space division multiplex (SDMA)

With spatial multiplexing (Engl. Space Division Multiplex, SDM or Space Division Multiple Access, SDMA ) is referred to in the Telecommunications transmitting multiple messages over installed in parallel transmission paths that are provided to the individual transmitters and receivers each for exclusive use.

A distinction is made between two different variants:

  • wired space division multiplex
  • wireless room multiplexing

Wired space division multiplexing

Schematic representation of a crossbar distribution

The wired room multiplex method is the simplest and oldest multiplex method. Several lines are installed in parallel to support individual connections at the same time. These parallel lines are also called trunk (Engl. Trunk ), respectively.

The simplest application of this process is the multi-core cable that was used in the early days of telecommunications and is still used today. Another method of space division multiplexing is cross bar switching , which is also referred to as a switching network. This is a matrix of several lines with many switches. This shows one of the advantages of the room multiplex method: This matrix ensures that every transmitter can reach every receiver, provided the line is free and the switch is active.

Wireless room division multiplexing

With the wireless room multiplex method, a separate area or a separate radio link is used for each group of connections. Such a radio link or an area is normally used multiple times with the aid of the frequency division multiplex method or the time division multiplex method or a combination of both.

Space division multiplexing becomes necessary when the number of connections to be transmitted increases and there is also a shortage of frequencies. Then the same frequency is used several times with sufficient spatial distance. Sufficient spatial distance is necessary to avoid disruptive interference between the various transmitters with the same transmission frequency . This method is used, among other things, in directional radio networks and for frequency allocation in radio , television and cellular mobile communications .

A special case is multiple input multiple output , in which a logical signal is transmitted through several cooperating antennas in order to significantly improve the quality (low bit error rate) and data rate of a wireless connection.

Frequency division multiplexing (FDMA)

The frequency division multiplex method , in English Frequency Division Multiplex (FDM) or Frequency Division Multiple Access (FDMA), can be used in both wired and wireless communication systems. A first proposal for the multiple use of lines by frequency division multiplexing was made in 1886 for telegraphy by Elisha Gray . Probably the best-known application is the stereo sound transmission in FM radio.

Frequency division multiplex on lines

Schematic representation of the frequency division multiplex method

Several signals are modulated onto their own carrier frequency . Since two identical sidebands arise during the modulation, one sideband and the carrier frequency itself are suppressed in the carrier frequency systems of communications engineering. The resulting, narrow frequency bands are combined into a broadband signal and then transmitted together. The signals are transmitted simultaneously and independently of one another.

The receiver has to replenish the carrier frequencies and can use filters to separate the signals and then convert them back to their original frequency position by demodulation . To avoid interference and to achieve better separation of the signals in the receiver filter, unused, so-called guard bands, are left free between the individual frequency bands.

Frequency division multiplex was mainly used on the (copper) long-distance connection lines in the telephone network. Time division multiplexing or wavelength division multiplexing, which use the fiber optic network, is now mostly used in the telephone network.

Today frequency division multiplex is still used for the transmission of information via broadband distribution networks , such as cable television . In the case of radio and television signals, too, only one sideband is transmitted, but the carrier frequency is not suppressed, as this simplifies the receiver.

Frequency division multiplex for radio transmission

Each signal is also modulated to its own carrier frequency, the signal bands are then combined and broadcast. The carrier frequency is the center frequency of the radiated radio wave . In the receiver, the frequency bands coming from the antenna are split up by filtering or the superheterodyne method and then demodulated.

Classic examples are terrestrial broadcasting and satellite television .

Today's areas of application are directional and mobile radio technology in telecommunications . The Russian satellite navigation system GLONASS also uses the frequency division multiplex method. In addition, this method can be combined with the time division multiplex method, for example with the Global System for Mobile Communications (GSM), with Digital Enhanced Cordless Telecommunications (DECT) or with Bluetooth .

A further development of FDM is OFDM ( Orthogonal Frequency Division Multiplexing ) in which a signal is distributed over several (thousand) subcarriers, the frequencies of which are orthogonal to one another, i.e. their angular frequencies are each integral multiples of a fundamental frequency. OFDM is mainly used today for LTE, DVB-T and digital wireless cameras.

Note: The assignment of different frequencies to different (spatially separated) transmitting stations (e.g. mobile radio cells) is usually not referred to as frequency division multiplex. This is called frequency planning or space division multiplex (SDM).

Optical wavelength division multiplexing

The wavelength division multiplexing (engl. Wavelength Division Multiplexing , WDM , or Wavelength Division Multiple Access , WDMA) is an optical frequency division multiplexing method in the transmission of data (signals) over fiber cable ( optical waveguide is used).

Wavelength division multiplexing using arrayed waveguide grating . The light path from (1) to (5) functions as an optical demultiplexer and from (5) to (1) as opt. Multiplexer.

In the wavelength division multiplexing process , light signals consisting of different wavelengths ( see: frequency bands of optical data communication ) are used for transmission in an optical waveguide. Laser diodes (LD) or light-emitting diodes ( LED ) are primarily used as the source for the light signals . Each of these narrowband wavelength ranges generated in this way thus forms its own transmission channel on which the data (signals) of a transmitter can now be modulated. The data (signals) modulated in this way are then bundled by optical coupling elements and transmitted simultaneously and independently of one another. At the destination of this optical multiplex connection, the individual optical transmission channels are separated again by passive optical filters or wavelength-sensitive opto-electrical receiver elements. The method can be combined with other optical multiplex methods.

Several FBG sensors in one glass fiber.png

There are now micro-optical components that can amplify, separate and switch ( route ) individual optical channels without prior conversion to electrical signals . This means that purely optical networks can be implemented. Essential components of this technology are optical multi- and demultiplexers, optical amplifiers and optical crossconnects that separate the channels with filters.

See also: Arrayed Waveguide Grating (right picture).

In wavelength division multiplex technology, a distinction is made today between dense (DWDM), coarse (CWDM) and wide wavelength division multiplex (WWDM).


The so-called density wavelength division multiplexing (Engl. Dense Wavelength Division Multiplexing , DWDM) is considered at the time as the most powerful variant. Here the wavelengths (spectral colors) used for transmission in the fiber optic cable are very close together. The frequency range of the wavelengths is usually in the C or L band with a frequency spacing of 0.4 nm (50 GHz ) to 1.6  nm (200 GHz). These small frequency intervals can only be achieved by using temperature and wavelength stabilized lasers (thermostated DFB laser diodes ) and high quality filters. This results in data transfer rates of around 10–100 Gbit / s per channel with up to 80 channels. By combining the C and L bands, up to 160 channels are possible.
Depending on the manufacturer, network design and type of fiber optic, optical amplifiers are required every 80–200 km and electrical data regeneration every approximately 600–2000 km. For this reason, the main area of ​​application of this variant is use over long distances in wide and global area networks .

The higher the data rate on a channel, the greater the influence of dispersion . At data rates of 10 Gbit / s or more, influences from chromatic dispersion must be expected; at data rates of 40 Gbit / s or more, there are additional effects such as polarization mode dispersion (PMD). These linear distortions occur separately within each channel and can be at least partially compensated for, either by hardware compensators or by fast signal processors. In addition, non-linear distortions can occur in wavelength division multiplex systems, as a result of which the optical power in one channel disrupts the transmission of the neighboring channels. With four-wavelength mixing, a fourth is created from three optical frequencies, which can fall into a different transmission channel. Through the cross-phase modulation, the power of an optical channel changes the refractive index of the fiber and thus the phase of the adjacent channels , which can also lead to depolarization.


A lower-cost variant, the coarse wavelength division multiplexing (engl. Coarse Wavelength Division Multiplex , CWDM). In order to transmission of signals are standardized 18 wavelengths with a channel spacing of 20 nm between 1271 nm and 1611 nm. Depending on the fiber type and system manufacturer, not all wavelengths can always be used. This “rough” division of the wavelengths was chosen in order to be able to use cheaper lasers and components. Data transfer rates of up to 10 Gbit / s per channel and cable ranges of up to 70 km are achieved without signal amplification. Network connections in urban areas (so-called Metropolitan Area Network ) are considered areas of application .


The simplest option is WWDM ( wide wavelength division multiplex ). The technique is most often used to transmit the signals from the 1310 nm and 1550 nm windows simultaneously on one fiber.

Polarization multiplex (PM)

With the same data rate, the required bandwidth is halved by modulating the two polarization directions separately. The polarization multiplex method is used for the transmission of 100 Gbps over fiber optic cables in wide area networks. If each polarization direction is modulated with QPSK, the combined signal has a step speed of 25 Gbaud. This allows 100 Gbps to be transmitted over the widespread transmission systems with 50 GHz channel spacing. This modulation format is called DPQPSK (Dual Polarization Quadrature Phase Shift Keying) or PMQPSK. Polarization-selective receivers, which can follow the continuously changing polarization directions, are necessary for reception. In particular, coherent receivers with fast signal processors are used for this purpose.

Time division multiplex (TDMA)

With the time division multiplex method (abbreviation TDM for Time Division Multiplex or TDMA for Time Division Multiple Access ) the data (signals) of different transmitters are transmitted on one channel in certain time segments ( time slots ). The time division multiplex method differentiates between synchronous and asynchronous methods.

Synchronous procedure

synchronous procedure

With the synchronous method (abbreviation STD for Synchronous Time Division ), each transmitter is assigned a fixed time segment by the multiplexer for the transmission of its data (signals) on the transmission channel.

This has the advantage that each connection receives a constant data transmission rate. In addition, a transmitter can be identified at any time by its position on the transmission channel. This simplifies the demultiplexing process required at the destination.

The disadvantage is that if a transmitter does not send any data (signals), the corresponding period of time remains unused. In such a case, the transmission channel is not optimally utilized.

Asynchronous process

asynchronous procedure

The asynchronous method (abbreviation ATD for Asynchronous Time Division ) avoids the disadvantage of the synchronous method, so that unused, assigned time segments can also be occupied by other data streams. This is done by only allowing those transmitters to access the transmission channel that are actually sending data (signals). However, because the unambiguous assignment of time period and data stream is lost, it is necessary to add channel information (other names: header, channel identifier) ​​to each data packet. Using this channel information, the demultiplexer can reassign the data packets to the correct stream at the destination of the transmission channel. This is why the asynchronous method is also sometimes referred to as address multiplexing or label multiplexing . This needs-based allocation of the time segments means that the transmission channel is used very economically. If all transmitters transmit data (signals), all receive a constant data transmission rate. Free time periods due to inactive transmitters are also used by the other transmitters, which increases their data transmission rate. This is then also called dynamic multiplexing . The disadvantage is that the data packets become larger due to the channel information and the effort involved in demultiplexing.

Flexible time division multiplex

The flexible time division multiple access (FTDMA) method is also known as the minislot method. The method is advantageous when there are few requirements in terms of latency. The method is e.g. B. used to transmit dynamic segments in FlexRay .

Time division multiplex in human-machine communication

Since human perception is not able to resolve optical stimuli very highly in terms of time, it is advisable for displays such as screens or LEDs to output the information using time division multiplexing. Since people cannot type as fast as they want, similar considerations apply to inputs, for example on keyboards. With the multiplex method it is possible to reduce the wiring effort compared to individually connected LEDs or buttons (depending on their number considerably). This type of time division multiplex could be classified under “asynchronous”, but because of the different objective compared to message transmission, such a classification is generally not carried out.

Displays and input keys in matrix circuit

Arrangement of 35 light-emitting diodes as a 7 × 5 matrix

Several display elements, for example light-emitting diodes, are interconnected to form a matrix arrangement . Each row or column can be controlled one after the other via electronic switches (transistors, not shown in the picture) connected to the row or column connections. This activation takes place so quickly that all activated elements appear to light up simultaneously to the eye; see the animation in the introduction of the article. In general, care must be taken to ensure that the current only takes the direct current path through the display element in the intersection area of ​​the activated row and column (decoupling). The indirect route via several display elements is undesirable. When using light-emitting diodes, the decoupling takes place because they only conduct electricity in one direction.

In addition to the obvious interconnection of a matrix display to form a matrix circuit , several apparently individual light-emitting diodes can also be interconnected in one device to form a matrix circuit.

Keys or keypads can be interconnected in the same way. The controller activates the columns one after the other while watching for incoming signals at the rows. In order to reliably identify more than two simultaneously pressed keys, diodes must be used to decouple the keys from one another. This effort is z. B. operated on keyboards ; with typewriter-like computer keyboards, on the other hand, one is generally content with ignoring the coupling when more than two keys are pressed.

For a keypad of 64 keys, 8 + 8 = 16 lines are sufficient; 48 signal lines are saved compared to individual interrogation of each key.

Charlieplex process

A variant of the time division multiplex process is the Charlieplex process. You get by with even fewer cables; however, tri-state components are required for this.

For a keypad of 66 keys (without diodes) 12 lines are sufficient, 54 lines are saved compared to the individual query.

Historical solutions and today's applications

The time division multiplex method, like the frequency division multiplex method, can be used both in wired and in wireless communication systems. The first known applications of time division multiplexing go back to the Italian Giovanni Caselli and the French Jean-Maurice-Émile Baudot . Caselli developed the Pantelegraph with which from 1865 two images were multiplexed line by line. Baudot's apparatus, developed in 1874, made it possible to transmit four to six telegraph signals over one line using the synchronous time division multiplex method. Current fields of application are transmission technologies such as Integrated Services Digital Network (ISDN), Asynchronous Transfer Mode (ATM) or U p0 interface . The GSM mobile radio network uses both the time division multiplex method and the frequency division multiplex method and the space division multiplex method.

Code division multiple access (CDMA)

With the code division multiplex method (CDM or CDMA), various signal sequences are transmitted over a line or a radio frequency and recognized and assigned in the receiver based on their coding . The codes used by the participants are chosen in such a way that when a signal is received by the "wrong" receiver at the demodulator output, practically no output signal appears; the codes have a minimal cross-correlation with one another . Depending on the method used, in contrast to the time division multiplex method, there is no need to coordinate the time windows.

Examples are radio remote controls and radio-controlled central locking in motor vehicles .

The UMTS also works with CDMA when differentiating between several participants ; The sending and receiving directions are also distributed over two different frequencies (FDD).

See also


  • Jens R. Ohm, Hans D. Lüke: Signal transmission: Basics of digital and analog communication systems . 8th edition. Springer Berlin, Berlin 2002, ISBN 3-540-67768-2 .
  • Achyut K. Dutta, Niloy K. Dutta, Masahiko Fujiwara: WDM Technologies: Passive Optical Components . Academic Press, San Diego 2003, ISBN 0-12-225262-4 .
  • Nim K. Cheung, Kiyoshi Nosu, Gerhard Winzer: Dense Wavelength Division Multiplexing Techniques for High Capacity and Multiple Access Communication Systems. In: IEEE Journal on Selected Areas in Communications. Vol. 8, No. August 6, 1990.

Web links

Individual evidence

  1. ITU-T G.694.2 (English) ( PDF file (English) ; 195.2 kB)
  2. Charlieplexing at Maxim ICs (Appnote; English)