Types of multiplexing. Frequency multiplexing. Railway data networks

The multiplexing process is a procedure for compressing data to transmit such a stream over a common communication line in order to significantly increase bandwidth information channel (Fig. 1). A switching device capable of producing compressed data streams is called a multiplexer (MUX). Reverse process, i.e. Decompression of data is called demultiplexing. A device with such switching is called a demultiplexer (DEMUX).

Figure 1. Data multiplexing and demultiplexing

There are three main multiplexing methods:
- frequency multiplexing (FDM, Frequency Division Multiplexing) or, more precisely, frequency division multiplexing
- time multiplexing (TDM, Time Division Multiplexing) or time division multiplexing
- wave multiplexing (WDM, Wavelength Division Multiplexing) or wavelength division multiplexing.

Frequency Division Multiplexing (FDM).
Frequency division multiplexing (Figure 2) is used in telephone networks for organizing voice signal transmission, and can also be used in cable television.

Figure 2. Frequency multiplexing and demultiplexing

The basic idea of frequency multiplexing is as follows. At the first stage, the process of dividing the general broadband communication channel into separate frequency bands (subchannels) is carried out, on which subscriber frequencies are superimposed frequency ranges. At the second stage, in order to avoid the mutual influence of compressed user ranges, an insurance frequency uninformative band, the so-called defiltering band, is added to each subchannel. The speech harmonic spectrum includes a frequency width from 300 Hz to 3400 Hz. Thus, the size of each subchannel is 4 kHz, where 3.1 kHz is the voice information range + 0.9 kHz is the filtering band. The frequency multiplexing method provides three standardized levels of hierarchy of multiplexed subscriber subchannels:
1st level, basic group - 12 subscriber subchannels in a 48 kHz band from 60 kHz to 108 kHz. This standard is the most common.
2nd level, supergroup - 5 basic groups (60 subscriber subchannels) in a 240 kHz band from 312 kHz to 552 kHz.
3rd level, main group - 10 supergroups (600 subscriber subchannels) in a 2520 kHz band from 564 kHz to 2048 kHz.
It must be said that a certain paradox has arisen in using the frequency division multiplexing method. On the one hand, this technique of analog data multiplexing (FDM) began to be inferior to that of digital data multiplexing (TDM) due to its significant drawback - the appearance of noise when increasing the amplification of the voice signal. On the other hand, with the use of optical fiber as a new data transmission medium, the method of wavelength multiplexing of light radiation (WDM) was born (return to the topic!). And wave and frequency, as is known, are inversely proportional parameters. Therefore, frequency multiplexing logically “merged” into wave multiplexing. Your status has increased!

Time Division Multiplexing (TDM).
Time division multiplexing (Figure 3) is widely used in network technologies PDH, SDH/SONET, ATM, Ethernet, PON.

Figure 3. Time multiplexing and demultiplexing

The essence of this time division multiplexing method is as follows: using a TDM multiplexer, the input subscriber channels are sequentially connected to a common communication channel for a certain time interval, the so-called time slot, and on the receiving side, a demultiplexer decompresses the general stream into separate samples and distributes them via the corresponding receiving subscriber channels.

Wavelength Division Multiplexing (WDM)
Wavelength division multiplexing appeared with the advent of optical fiber. Wave multiplexing is a procedure for multiplexing the spectrum of optical infrared waves, using the unique property of optical fiber for WDM multiplexing (Figure 4). The essence of this phenomenon is as follows: on one optical fiber, using a wave optical multiplexer, it became possible to compress the entire spectrum of carrier laser waves and, accordingly, at the reception stage, to decompress this light flux into individual waves using an optical demultiplexer. This feature significantly increases the throughput of fiber-optic communication lines (FOCL).

Figure 4. Wave multiplexing and demultiplexing

The path of development of wavelength division multiplexing methods followed the following scheme: WDM → DWDM → HDWDM → CWDM, where
1st stage: 2 and 3 channel multiplexing (WDM)
Stage 2: Dense Multiplexing (DWDM) up to 88 channels
Stage 3 - high-density multiplexing (HDWDM) up to 256 channels
4th stage - sparse multiplexing (CWDM) up to 16 channels.
Historically, the first to appear were two-wave WDM splitters operating in duplex mode at wavelengths from the second and third transparency windows of optical fiber at 1310 nm and 1550 nm (see in the section “ Helpful information” of our website entitled “Transparency windows and spectral ranges of optical fiber”). Later, a third wave at 1490 nm was added to the WDM splitter. Due to their ease of installation and connection, such inexpensive multiplexers are indispensable in PON-type optical networks. The 1310/1490 nm wave pair, operating in an interactive mode, is used in the Internet and IP telephony. And the 1550 nm wave is intended for cable television. The emergence of dense DWDM multiplexing (Dense WDM) was associated with the need to increase the capacity of optical networks (PON) and fiber-optic communication lines (FOCL). But only then did DWDM become real and effective when optical erbium amplifiers (EFDA) began to be introduced into fiber optic links. And then there was a need to select and standardize the three main determining parameters for this method of optical signal compression: the reference wave, the operating frequency range and the pitch between channels. The choice fell on the 1550 nm wave from the second transparency window of the optical fiber. Setting the spectral range and distance between channels determines the so-called frequency plan or frequency grid. For DWDM, according to the ITU-T G.694.1 recommendation, a frequency plan is defined in the wave range 1528.77 – 1568.77 nm with a step of 0.8 nm or in the frequency dimension in the range 196.1 – 191.1 THz with a step of 100 GHz. Currently, a frequency plan with a reduced step of 50 GHz (0.4 nm) has been developed for high-density HDWDM multiplexing (High Dense WDM), and experimental systems already offer HDWDM with frequency grids with channel spacing of 25 GHz (0.2 nm) and 12.5 GHz ( 0.1 nm)!
However, the process of high-density multiplexing cannot be endless and certainly has its logical limit. And the main criterion for the demand for HDWDM-type systems is, first of all, the price of the fiber-optic line transceiver components. For relatively inexpensive and high-quality wave multiplexing, sparse CWDM (Coarse WDM) multiplexing with a frequency plan (ITU-T G.694.2) was developed in the range 1270 - 1610 nm with a step of 20 nm. This frequency grid specifies 18 frequencies for multiplexing from 4 to 16 channels.
The widespread use of optical fiber in PON and FTTH technologies using wave spectral multiplexing in fiber-optic communication channels has led to a fundamental breakthrough in the field of building modern data transmission networks with highest speed and unprecedented throughput.

Now about multiple access. Multiple access is a way of sharing a common communication channel resource between participants information exchange. At the same time, the effectiveness and sufficiency of multiple access as such and as a procedure for collective interaction of users can only occur if there is a technology that significantly increases the capacity of the communication channel. In this context, multiple access, depending on which scheme works to increase the throughput of the communication channel, is divided into the following types:

Using multiplexing techniques
- frequency division multiple access (FDMA)
- time division multiple access (TDMA)
- Wavelength Division Multiple Access (WDMA).

Using other methods
- multiple access with transfer of authority or token (TPMA)
- Carrier Sense Multiple Access and Collision Detection (CSMA/CD).

TPMA multiple access uses a deterministic token method of data transmission, sometimes this method is called relay, since the right to transmit is relayed from subscriber to subscriber. This method necessarily assumes a ring topology for the location of subscribers, with two rings being built: one ring is a backup ring in case of emergencies or failures. The essence of the method is this. A token, a special control package, continuously rotates around the ring. Hence another name for the method – token! So, if the token is free, it gives the subscriber the right to transmit. The subscriber who has received a free token makes the token busy, attaches his packet of information to it and sends such a package around. The remaining subscribers in the ring analyze this package for the addressee. If the package is not addressed to the subscriber, he passes it around. If the subscriber finds his address in the package, he accepts the information, marks the marker as accepted and sends the package back through the ring. The transmitting subscriber, having received his parcel back with a mark of acceptance, deletes his information packet, marks the token as free and sends a clean token further along the ring. Everything repeats itself again. Multiple access with token passing has been successfully used in Token technologies Ring and FDDI.

CSMA/CD Multiple Access uses a multiple access technique with carrier sensing and collision detection. Such a multiple access method does not allow creating a collision, i.e. the situation of simultaneous data transmission over a common channel of several users. The information unit is a frame superimposed (modulated) on a carrier frequency (5-10 MHz). The frame header contains the sender and recipient addresses of the frame. The principle of operation of such access is based on two fundamental points: first, each subscriber determines the situation when he can transmit a frame, second, how the transmitting subscriber should behave in the event of the simultaneous start of frame transmission by another subscriber. The situation - whether the communication channel is free or not is determined by the passage of the information-carrying frequency (frame) through the channel. To capture the essence of the algorithm this method, consider the work of subscriber No. 1. So, subscriber No. 1, who needs to transfer data, has determined that there is a carrier in the communication channel, i.e. The communication channel is busy: our subscriber takes a technological pause (9.6 μs). After a pause, he again switches to the mode of listening to the communication channel for the carrier frequency, Carrier - yes! Someone is broadcasting, subscriber No. 1 takes a technological break again. There is no carrier! The communication channel is free, frame transmission begins. Immediately, transmitting subscriber No. 1 monitors the state of the communication channel for the fact of detecting a collision. All users participate in listening to the channel, analyzing the incoming frame, and the subscriber whose address is written in the frame header begins receiving the frame. Other subscribers ignore the “alien” frame. If subscriber No. 1 does not detect a collision during the entire frame transmission, the process of receiving and transmitting information ends correctly. If subscriber No. 1 detects a collision, frame transmission stops, subscriber No. 1 throws a special signal into the communication channel, upon receiving which the simultaneously triggered user stops his transmission, and our subscriber takes a random pause, after which he tries to continue transmitting the current frame according to the above algorithm. The algorithm contains 16 attempts to correctly complete the reception and transmission of the current frame. If, due to collisions, it is still not possible to complete the transmission of the current frame within of this algorithm, such an unfortunate frame is simply discarded by transmitting subscriber No. 1 and his receiving counterpart. Next – who’s first again! Multiple carrier sensing and collision detection has proven itself in Ethernet networks.

We do not affect other multiple access schemes that are used in wireless communication such as: CDMA (Code Division Multiplexing), OFDMA (Orthogonal Frequency Division Division), SDMA (Spatial Division Multiplexing). This is the topic of another article.

In long-distance data networks, the capacity of trunk communication lines usually significantly exceeds the transmission capacity of individual applications. This is done for the purpose of transmitting many of these applications at the same time. In order to increase the efficiency of the transmission medium and adapt it to many heterogeneous applications, the simultaneous transmission of several information signals in one medium is used - m multiplexing. In other words.:d To use the high-speed characteristics provided by broadband channels, the so-called multiplexer connection is widely used.

General purpose multiplexer - matching a large number of low-speed channels with a smaller (usually one) number of high-speed ones.

С1 Own Сi-channel capacity.

1) If In = Out, then the system is called a multiplexer.

2) If In > Out, then it is a static multiplexer or concentrator.

3)If Svh<Свых- то коммутатор.

Frequency division multiplexing (FDM) when using GTS channels has good characteristics in terms of communication range, being inferior to time multiplexing in terms of data transmission speed.

An example of a multiplexer with frequency division multiplexing is an MFM.

When using this MCU, it is possible to use the frequency band for transmitting data and voice messages, as well as for transmitting telemetry information.

A type of FDM is WDM wave multiplexing, which is also used in fiber optic systems. The spectral region from l=1.3 nm (230 THz) to 1.6 nm (188 THz) is predominantly used. For dense wave multiplexing, the spectrum region of 1530-1560 nm is used.

WDM is widely used in so-called inverse multiplexing, when a broadband signal during transmission is placed in several channels with a smaller bandwidth.

An example of using inverse WDM multiplexing in a long-distance LAN.

A multiplex signal represented by multiple wavelengths better resists the influence of dispersion and introduced noise from EDFA optical amplifiers in a long fiber line.

To obtain high data transfer rates, multiplexers with time multiplexing are used, character-by-character or bit-by-bit synchronization is used. Let's consider, using the example of a SVU multiplexer that combines four directions A, B, C, D, how one and the other synchronization method is implemented.

Output package from the MVU in a high-speed channel for the MVU.

A) with character-by-character (byte-by-byte, object) synchronization.

Time slots

B) with bit synchronization.

B) With asynchronous (static) time division of channels.

With character-by-character synchronization, greater compression is possible, because starting wall signals can be excluded, but with bit synchronization this is impossible and, in addition, with bit synchronization it can lead to the loss of address information and the loss of the entire sequence due to transmission to incorrect addresses; bit synchronization codes are simpler to implement. By placing a simple MVU in a synchronous modem, you can get a multi-input (multi-channel) modem. To analog TLF Modem with MVU (modem / muldex) up to 19200 bps. Analog broadband IDP channel with alternating bits and bytes with statistical compaction up to 64 kbit/sec.

With error control HDLC protocol.

The transmission speed through the internal common interface between the MVU and the modem is equal to the total capacity of synchronous data channels, for example 9600 bits/sec = 4 x 2400 bits/sec.

MSVU (statistical time division multiplexers) – dynamically distribute the bandwidth of a common channel. A fixed frame format is not used in modern systems. Frames of the common channel can vary both in length and in the composition of data from the data channels. Each position in a frame is allocated to a data channel only when it has data to transmit. If at any time only one data channel is active. All positions in the frame can be allocated to this channel. If all channels are active at the same time, a priority system comes into play, preventing any one channel from occupying all positions in the frame. More data channels can be connected to the MVD than to a synchronous multiplexed fixed frame MVD because the time spent sending null characters is reduced. IDPs with alternating characters and a fixed frame have actually been discontinued. Currently, there is a wide range of MVVU with the ability to connect from 4/8 to several hundred channels.

One of the advantages of MSVU is the use of protocols such as ARQ (ARQ - automatic reguest for retransmission) - an automatic request for retransmission in a common channel: any data block with distortion as a result of linear interference must be retransmitted. Typically, duplex protocols such as HDLC are used, used, for example, at the data link layer of the X25 protocol stack.

Manufacturers of MRVs usually specify the number of asynchronous data channels that can be connected at a given channel speed (bits/sec). For an approximate calculation, we can assume that the total speed of asynchronous channels can be 4 times higher than the transmission speed of a synchronous common channel (but none of the data channels should exceed the transmission speed of the common channel!)

An example of an approximate calculation of the number of asynchronous channels. Common channel – 9600 bps. å The speed of asynchronous channels is 4 x 9600 = 38400 bits/sec., this capacity can provide the following number of channels: a) eight asynchronous channels of 4800 bits/sec, b) 16 asynchronous channels of 2400 bits/sec, c) 32 asynchronous channels of 1200 bits/sec. sec.

First, statistical multiplexing was used in networks with X.25 protocols, and later in Frame Relay and ATM networks, which will be discussed when considering wide area network (WAN) technologies.

The table below shows the comparative characteristics of synchronous and statistical multiplexing.

As can be seen from the table, the advantages of one method can be considered to some extent as the disadvantages of another.

Data transmission over broadband multiplexed channels over long distances is carried out in accordance with standards regulating data transfer rates.

In the USA and countries that adhere to similar standards, the T1 (1.544 Mbit/s) and T3 (45 Mbit/s) systems are used. In Europe, the analogues of the T1 and T3 systems are E1 (2 Mbit/s) and E3 (34 Mbit/s). The main reason for the popularity of digital lines is that they provide high-speed data transmission that is almost 99% error-free. Digital lines are available in various forms, including DDS standards (point-to-point synchronous connection 2.4; 4.8; 9.6; 56 kbit/s).

T1 is the most common type of digital line. Uses two pairs of wires with data transfer speed of 1.544 Mbps. T1 divides the channel into 24 subchannels and polls each one 8000 times per second. Each time the channel is accessed, 8 bits are transmitted, the speed over the subchannel is 64 kbit/sec.

T3-Line allocation T-3 transmit data at speeds from 6 to 45 Mbit/sec. The highest capacity among lines publicly available today. Can replace several T1 lines.

In information technology and communications, multiplexing(English) multiplexing, muxing) - channel compaction, i.e. transmission of several

data streams (channels) with lower speed (bandwidth) over one channel.

IN telecommunications Multiplexing involves transmitting data over several logical communication channels in one physical channel. A physical channel means a real channel with its own capacity - a copper or optical cable, a radio channel.

IN information technology Multiplexing involves combining several data streams (virtual channels) into one. An example would be a video file in which a video stream (channel) is combined with one or more audio channels.

The device or program that performs multiplexing is called multiplexer.

Frequency division multiplexing (FDM).

Frequency division multiplexing of 3 channels.

Technology.

Frequency Division Multiplexing(English) FDM, Frequency Division Multiplexing) involves placing several channels with a smaller width within the channel bandwidth. A clear example is radio broadcasting, where within one channel (radio broadcast) there are many radio channels at different frequencies (in different frequency bands).

Main applications.

Used in mobile communications networks (see FDMA) for access separation; in fiber-optic communications it is an analogue wavelength division multiplexing(WDM, Wavelength Division Multiplexing) (Where

frequency is the color of the radiation of the emitter), in nature - all types of divisions

by color (frequency of electromagnetic vibrations) and tone (frequency of sound vibrations).

Time division multiplexing (TDM).

Technology.

Time division multiplexing(English) TDM, Time

Division Multiplexing) involves frame data transmission, while

transition from channels of smaller width (bandwidth) to channels with

the larger one frees up the reserve for transmission within one frame of the larger one

the volume of several frames is smaller.

In the figure: A, B and C - multiplexed channels with capacity (width) N and frame duration Δt; E - multiplexed channel with the same duration Δt but with a width M*N, one frame of which ( superframe) carries all 3 frames of input multiplexed signals sequentially, each channel is assigned

part of the superframe time - timeslot, length Δt M =Δt/M

Thus, a channel with a capacity M * N can pass M channels with a capacity N, and while maintaining the channel speed (frames per second), the result of demultiplexing coincides with the original channel stream (A, B or C in the figure) both in phase and in terms of speed, i.e. it proceeds unnoticed by the final recipient.

Main Applications

wireless TDMA networks, Wi-Fi, WiMAX;

channel switching in PDH and SONET/SDH;

packet switching in ATM, Frame Relay, Ethernet, FDDI;

switching in telephone networks;

serial buses: PCIe, USB.

In previous lessons, we looked at typical computer networking equipment such as bridges, switches, and routers. However, due to the increasingly close integration of computer and telephone networks (communication networks in general), knowledge of the general principles of organizing telephone networks is becoming increasingly mandatory for administrators and even users, especially if they work with global networks. Therefore, in this lesson we decided to consider such a technology (more precisely, technologies) as multiplexing.

Laying and operating a low-speed trunk line between two telephone exchanges costs almost the same as a high-speed line, since the main costs are not the purchase of copper or optical cable, but, generally speaking, the digging of a trench for laying the cable. To transmit multiple telephone conversations over one physical line, telephone companies have developed multiplexing, or multiplexing, technologies.

MULTIPLEXING IN TWO WORDS

The principle of operation of the multiplexer is simple: signals arriving via several incoming low-speed lines are transmitted in a frequency range or time interval allocated for each of them via a high-speed outgoing line. At the opposite end of the high-speed line, these signals are isolated, or demultiplexed.

Based on the compaction method, multiplexing technologies can be divided into two main categories: Frequency Division Multiplexing (FDM) and Time Division Multiplexing (TDM). With frequency multiplexing, the frequency spectrum is divided into logical channels, and each user has this channel at his disposal for the duration of the conversation. With time multiplexing, users are periodically allocated the entire bandwidth, but only for a short period of time.

FREQUENCY MULTIPLEXING

As is known, human speech can be adequately transmitted by frequencies in the range from 300 to 3400 Hz, i.e. the required frequency interval is 3100 Hz. However, when multiplexing multiple voice channels, a 4000 Hz range is allocated to each of them so that they do not overlap. The frequency of each channel is increased by its own value, a multiple of 4 kHz, then the channels are combined. As a result, the channels are spread across the entire frequency spectrum of a given line. The channels are separated from each other by so-called guard intervals (see Figure 1).

Picture 1.
With frequency multiplexing, the entire frequency range is divided into several channels. To prevent channels from overlapping, they are separated from each other by guard intervals.

FDM multiplexing schemes are fairly standardized. The most widely used standard is that twelve 4000 Hz wide voice channels are multiplexed in the frequency range from 60 to 108 kHz. Such a block is called a group. The range from 12 to 60 kHz is sometimes used for another group.

A type of frequency multiplexing technology used in the case of optical communication lines is wavelength division multiplexing (WDM). Physically, multiplexing is carried out as follows: several fibers are supplied to a prism (or more often a diffraction grating), light beams are passed through the prism and enter a common fiber. At the opposite end, the beams are separated using another prism. If each input beam is limited to its own frequency range, then they will not overlap. Optical systems are completely passive and, as a result, more reliable.

PULSE CODE MODULATION

The modern world is becoming increasingly computerized and, as a result, digital; Of course, this trend has not bypassed telephone networks. Digital systems are becoming more widespread, and as a result, frequency multiplexing is giving way to time multiplexing. However, before human speech, which is analog in nature, can be transmitted over a digital network, it must be converted into discrete form. This is achieved using pulse-code modulation. Therefore, in modern digital telephone communication networks, time multiplexing is closely related to pulse-code modulation.

According to Kotelnikov's theorem, the sampling rate must be twice the maximum frequency of the analog signal spectrum to reproduce it correctly, so amplitude measurements must be made 8000 times per second in the case of human speech. The amplitude value approximates an 8-bit binary number, so the baud rate should be 64 kbps. As a consequence, in digital networks, the 64 kbit/s information channel is the base one for calculating the speed of all higher-capacity communication channels.

TIME MULTIPLEXING

With time division multiplexing, each device or incoming channel has the full bandwidth of the line at its disposal, but only for a strictly defined period of time every 125 μs (see Figure 2). The last value corresponds to the sampling cycle, since with PCM every 1/8000th of a second it is necessary to measure the amplitude of the analog signal. The transmission time of an eight-bit instantaneous amplitude value is called a time slot and is equal to the transmission duration of eight pulses (one for each bit). The sequence of time slices following at the above interval forms a time channel. The set of channels in one sampling cycle constitutes a frame.

Figure 2.
In time multiplexing, the entire capacity of an outgoing line is made available for a fixed period of time on an incoming line of smaller capacity.

In Europe, as in the rest of the world, with the exception of the USA and Japan, the standard system is PCM-32/30 (or E-1) with 32 time channels at 64 kbit/s, in which 30 channels are used as data channels for voice transmission , data, etc., and two - as service channels, with one of the service channels intended for signaling (service communication signals), the other for synchronization. As is easy to calculate, the total system capacity is 2.048 Mbit/s.

The E-1 system forms the so-called primary group. The secondary group E-2 consists of 4 E-1 channels with a total capacity of 8.448 Mbit/s, the tertiary E-3 system consists of four E-2 channels (or sixteen E-1 channels) with a total capacity of 34.368 Mbit/s, and the quaternary group consists of four channels E-3 with a total capacity of 139.264 Mbit/s. These systems form the European plesiochronic digital hierarchy.

The principle of sequential channel multiplexing is illustrated in Figure 3. Four E-1 channels are multiplexed into one E-2 channel, and at this and subsequent levels the multiplexing is carried out bitwise, rather than byte-by-byte, as was the case when 30 voice channels were multiplexed into one E-channel -1. The total capacity of the four E-1 channels is 8.192 Mbps, while the total capacity of E-2 is actually 8.448 Mbps. The redundant bits are used for framing and synchronization recovery. The four E-2 channels are then multiplexed into one E-3 channel, and so on.

Figure 3.
Just as small tributaries merge into one large river, low-speed lines are combined into high-speed ones using a hierarchy of multiplexers.

Adopted in North America and Japan, the standard defines the T-1 channel (DS1 frame format). The T-1 channel consists of 24 multiplexed voice channels, with the original intention being that the amplitude of the analog signal would be expressed as a 7-bit binary number, with one bit used for control (signaling) purposes. In addition, in addition to 192 bits, each frame has one more bit for synchronization. Thus, the total capacity of the T-1 channel is 1.544 Mbps. However, eventually all 8 bits were allocated to data, and signaling began to be carried out in one of the following two ways. In common channel signaling, the 193rd bit in each odd-numbered frame is used for synchronization purposes, and in each even-numbered frame for signaling purposes. The essence of another method is that each channel has its own subchannel for transmitting signaling information (one bit in every sixth frame).

SYNCHRONOUS DIGITAL HIERARCHY

The need for a common standard for communications systems in Europe and America, as well as the need for higher maximum transmission rates and integrated communications network management capabilities, led to the development of the synchronous digital hierarchy SDH (unfortunately, the North American version of this standard, called SONET, is somewhat different from the European one, although these differences are not as significant as in the case, for example, of the hierarchy of channels T-1, T-2... and E-1, E-2...).

In SDH, the synchronous transport module (STM-1) forms the lowest level of the hierarchy. It is equivalent to the STS-3c synchronous transport signal in the SONET hierarchy with a capacity of 155.52 Mbit/s. Four STM-1 modules are multiplexed into STM-4 (=STS-12c) with a capacity of 622.08 Mbit/s, and four STM-4 modules are multiplexed into STM-12 (=STS-48c) with a capacity of 2.488 Gbit/s. Hierarchy also defines higher levels.

Multiplexing is byte-by-byte rather than bit-by-bit, i.e., for example, when four STM-1 data streams are combined into STM-4, the multiplexer first sends one byte from the first stream, then one byte from the second, etc. in a circle.

One of the most important differences between a synchronous and a plesiochronous hierarchy is the ability to allocate the desired channel up to the E-1 level without demultiplexing the entire transport signal. This led to the emergence of a fundamentally different type of multiplexers - multiplexers with the addition and allocation of individual channels (in English terminology - add-drop multiplexer, and in Russian technical literature they are briefly called input/output multiplexers).

In addition, many multiplexers began to perform cross-switching functions (however, it may be the other way around, but this is a debate about the chicken and the egg). Cross-connect multiplexors allow concentration and separation of streams (multiplexing and demultiplexing functions) along with switching digital signals from one channel to another in accordance with certain rules (switching functions).

INVERSE MULTIPLEXING

In the case where an organization needs to have a line of a certain capacity, and the offered capacities are either too small (for example, E-1) or too large (say, E-3), then a device called an inverse multiplexer comes in handy. This device allows you to distribute the incoming data stream between several outgoing lines with a smaller capacity than the total amount of data received per unit of time (see Figure 4). Thus, for example, a customer can receive a channel equivalent in capacity to two E-1s. The advantage of this approach compared to independently connecting two E-1 lines is, for example, that an inverse multiplexer allows you to dynamically distribute the load between them.

Figure 4.
Inverse multiplexing brings to mind the flow of a river: going around the islands, it breaks into channels, which then merge again.

CONCLUSION

In this lesson we looked at the basic multiplexing technologies used in telephone networks. Telephony is increasingly intertwined with the world of computers, in any case, more and more often they use the same transport network in both global and local networks, not to mention the fact that such a “hot” ATM technology has appeared as one of broadband digital network options with integration of services. And, by the way, ATM would be more correctly called asynchronous time multiplexing. ATM's predecessor, Asynchronous Time Division (ATD), was developed in France Telecom laboratories as a variation of TDM. Its most important difference from TDM was the dynamic provision of the channel, and not for the entire duration of the connection (telephone conversation); the header made it possible to determine which connection the data belonged to. As a result, available capacity was used more efficiently. Now the successor to ATD claims to be a unified technology for both global and local networks. But this is a topic for another conversation.

Dmitry Ganzha is the executive editor of LAN. He can be contacted at:

multiplexing, muxing) - channel compaction, i.e. transmission of several data streams (channels) at a lower speed (bandwidth) over one channel.

IN telecommunications multiplexing involves transmitting data over several logical communication channels in one physical channel. A physical channel means a real channel with its own capacity - a copper or optical cable, a radio channel.

The device or program that performs multiplexing is called multiplexer.

Principles of Multiplexing

Frequency Division Multiplexing (FDM)

3 Channel Frequency Division Multiplexing

Technology

Frequency Division Multiplexing(English) FDM, Frequency Division Multiplexing) involves placing several channels with a smaller width within the channel bandwidth. A clear example is radio broadcasting, where within one channel (radio broadcast) there are many radio channels at different frequencies (in different frequency bands).

Main Applications

Used in mobile communications networks (see FDMA) for access separation; in fiber-optic communications it is an analogue wavelength division multiplexing(WDM, ) (where frequency is the color of the emitter's radiation), in nature - all types of divisions by color (frequency of electromagnetic oscillations) and tone (frequency of sound oscillations).

Time division multiplexing (TDM)

Technology

Time division multiplexing(English) TDM, Time Division Multiplexing) involves frame-based data transmission, while the transition from channels of smaller width (bandwidth) to channels with larger ones frees up the reserve for transmitting several smaller frames within one frame of a larger volume.

Main Applications

wireless TDMA networks, Wi-Fi, WiMAX;
channel switching in PDH and SONET/SDH;
packet switching in ATM, Frame Relay, Ethernet, FDDI;
switching in telephone networks;
serial buses: PCIe, USB.

Wavelength division multiplexing (WDM)

Technology

Wavelength division multiplexing(English) WDM, Wavelength Division Multiplexing) involves the transmission of channels at different wavelengths over one optical fiber. The technology is based on the fact that waves of different wavelengths propagate independently of each other. There are three main types of wavelength division multiplexing: WDM, CWDM and DWDM.