Frequency division of communication. Signal separation Signal separation

Frequency separation of signals. Functional diagram the simplest system Multichannel communication with frequency division is shown in Fig. 9.2.

Let us trace the main stages of signal formation, as well as the changes in these signals during the transmission process. First, in accordance with the transmitted messages, primary (individual) signals having energy spectra G 1 (ω), G 2 (ω),..., G N (ω) modulate the subcarrier frequencies of each channel. This operation is performed by modulators M 1, M 2, ....., M N channel transmitters. Received at the output frequency filtersΦ 1, Φ 2, ..., Φ N spectra g k (ω) channel signals occupy respectively the frequency bands Δω 1, Δω 2,..., Δω N (Fig. 9.3), which in general case may differ in width from the message spectra Ω 1, Ω 2,..., Ω N. With broadband types of modulation, for example FM, the spectrum width Δω k ≈2(β + 1)Ω k, with OM Δω k = Ω k, i.e. in the general case Δω k ≥ Ω k For simplicity, we will assume that OM is used (as is customary in cable systems of multi-channel communication with frequency division), i.e.

Δω k = Ω and Δω = NΩ. (9.11)

We will assume that the spectra of individual signals are finite. Then you can select the subcarrier frequencies ω k so that the bands Δω 1 ,..., Δω 1 do not overlap in pairs. Under this condition, the signals s k (t) (k = 1,..., N) are mutually orthogonal. Then the spectra g 1 (ω), g 2 (ω),...,g N (ω) are summed (SU) and their totality g(ω) is fed to the group modulator (M). Here, the spectrum g(ω) with the help of oscillations of the carrier frequency ω 0 is transferred to the frequency region allocated for the transmission of a given group of channels, i.e. the group signal s(t) is converted into line signal s L (t) In this case, any type of modulation can be used.

At the receiving end, the linear signal is fed to a group demodulator (receiver Π), which converts the spectrum of the linear signal into the spectrum of the group signal g(ω). The spectrum of the group signal is then, using frequency filters Φ 1, Φ 2,..., Φ N, again divided into separate bands Δω k corresponding to individual channels. Finally, channel demodulators D convert the spectra of signals g k (ω) into spectra of messages G k (ω) intended for recipients.

From the above explanations it is easy to understand the meaning of the frequency method of channel separation. Since any real communication line has a limited bandwidth, during multi-channel transmission each individual channel is allocated a certain part of the total bandwidth.

On the receiving side, signals from all channels operate simultaneously, differing in the position of their frequency spectra on the frequency scale. In order to separate such signals without mutual interference, receiving devices must contain frequency filters. Each of the filters Φ l must pass without attenuation only those frequencies ω∈Δω k that belong to the signal of a given channel; the frequencies of the signals of all other channels ω∉Δω k the filter must suppress.

Mathematically, the frequency separation of signals by ideal bandpass filters can be represented as follows:

where g k (t) is the impulse response of an ideal bandpass filter that passes the frequency band Δω k without distortion. Expression (9.12) coincides with (9.6) with the weight function η k (t, τ) = g k (t-τ). In the spectral domain, transformation (9.12) corresponds to multiplying the spectrum of the group signal by a Π-shaped transfer function (see Fig. 9.3).

So, from the point of view of the possibility of completely separating signals from different channels, it is necessary to have such filters Φ k, the bandwidth of which fully corresponds to the width of the spectrum of the signal Δω k; the filter Φ k should not respond to harmonic components outside the band Δω k. This means that the energy of the signals s k is completely concentrated within the limited band Δω k allocated to the k-th channel. If both of these conditions were satisfied, then using frequency filters it would be possible to separate the signals of different channels without mutual interference. However, none of these conditions are fundamentally feasible. The result is mutual interference between channels. They arise both due to the incomplete concentration of the signal energy of the k-th channel within a given frequency band Δω k, and due to the imperfection of real bandpass filters. In real conditions, it is also necessary to take into account mutual interference of nonlinear origin, for example due to the nonlinearity of the characteristics of the group channel.

To reduce crosstalk to permissible level it is necessary to introduce protective frequency intervals Δω protect (Fig. 9.4). So, for example, in modern systems multichannel telephone communication Each telephone channel is allocated a frequency band of 4 kHz, although the frequency spectrum of transmitted sound signals limited to a band from 300 to 3400 Hz, i.e., the spectrum width is 3.1 kHz. Intervals of 0.9 kHz width are provided between frequency bands of adjacent channels, designed to reduce the level of mutual interference when filtering signals. This means that in multi-channel frequency division communication systems, only about 80% of the communication link bandwidth is effectively used. In addition, it is necessary to ensure a very high degree of linearity of the entire baseband signal path.

Time separation of signals. The principle of time division of signals is very simple and has long been used in telegraphy. It consists in the fact that with the help of a switch K per group path is provided in turn for the transmission of signals from each channel of a multi-channel system *. When transmitting continuous messages, time sampling (pulse modulation) is used for time division. First, the signal (pulse) of the 1st channel is transmitted, then the next channel, etc. until the last channel number N, after which the 1st channel is turned on again and the process is repeated periodically (Fig. 9.5).

* (In modern equipment, mechanical switches are practically not used. Instead, electronic switches are used, made, for example, on shift registers.)

At the receiving end, a similar switch K pr is installed, which connects the group path in turn to the receivers of the corresponding channels. The receiver of each k-ro channel must be connected only for the duration of the k-ro signal transmission and turned off the rest of the time while signals are transmitted in other channels. This means that for normal operation of a multi-channel time division system, synchronous and in-phase operation of the switches on the receiving and transmitting sides is necessary. Often, for this purpose, one of the channels is occupied for the transmission of special synchronization pulses intended for time-coordinated operation of K lane and K lane.

In Fig. Figure 9.6 shows the timing diagrams of a two-channel system with AIM. The message carrier here is a sequence of pulses (with a period T 0 = 1/2F max) arriving at the pulse modulator (PM) from the clock pulse generator (GTI). The group signal (Fig. 9.6, a) is supplied to the switch K pr. The latter plays the role of “temporary” parametric filters or switches, the transfer function of which K k (Fig. 9.6,6) changes synchronously (with period T 0) and in phase with changes transfer function K lane:

This means that the transmission path within each time interval Δt k is connected only k-n pulse ID-k detector. The messages received as a result of detection s k (t) arrive at the message recipient PS-k.

The operator π k, which describes the operation of the key filter, cuts out from the signal s(t) the intervals Δt k following with period T 0 and discards the rest of the signal. It is easy to see that it can be represented in the form (9.6) if

Here, as before, Δt k denotes the interval during which signals from the kth source are transmitted.

In time separation, mutual interference is mainly due to two reasons. The first is that linear distortions, arising due to the limited frequency band and imperfection of the amplitude-frequency and phase-frequency characteristics of any physically feasible communication system, violate the pulsed nature of the signals. Indeed, if, when transmitting modulated pulses of finite duration, we limit the spectrum, then the pulses will “spread out” and instead of pulses of finite duration we will get processes that are infinitely extended in time. When temporally separating signals, this will lead to the fact that the pulses of one channel will overlap with the pulses of other channels (Fig. 9.7). In other words, mutual crosstalk or intersymbol interference occurs between channels. In addition, mutual interference can arise due to imperfect synchronization of clock pulses on the transmitting and receiving sides.

To reduce the level of mutual interference, it is necessary to introduce “protection” time intervals, which corresponds to some expansion of the signal spectrum. Thus, in multi-channel telephony systems, the band of effectively transmitted frequencies is F = 3100 Hz; in accordance with Kotelnikov’s theorem, the minimum value of f 0 = 2F = 6200 Hz. However, in real systems the pulse repetition frequency is chosen with some margin: f 0 = 8 kHz. To transmit such pulses in single-channel mode, a frequency band of at least 4 kHz will be required. With time separation of channels, the signal of each channel occupies the same frequency band, determined under ideal conditions according to Kotelnikov’s theorem from the relation (without taking into account the synchronization channel)

Δt k = T 0 /N = 1/(2F total), (9.15)

where F total = NF, which coincides with the total frequency band of the system with frequency division. Although theoretically, time and frequency division allow one to obtain the same efficiency in using the frequency spectrum, nevertheless, so far, time division systems are inferior to frequency division systems in this indicator.

At the same time, time division systems have an undeniable advantage due to the fact that, due to the different timing of signal transmission from different channels, there is no transient interference of nonlinear origin. In addition, time division equipment is much simpler than in frequency division, where each individual channel requires appropriate bandpass filters, which are difficult to implement using microelectronics. An important advantage of time division systems is a significantly lower crest factor. Time division is widely used in the transmission of continuous messages with analog pulse modulation, and especially in digital systems PCM.

Note also that the total power P total of the received signal s(t) necessary to ensure a given fidelity in the presence of fluctuation interference, both with frequency and time division (as well as with other systems with linear division discussed below) in the ideal case in N times greater than the power P for single-channel transmission with the same type of modulation P total = NP. This is easy to understand because when independent signals are added, their powers add up. In fact, due to crosstalk, the reception fidelity in a multi-channel system when this condition is met is slightly lower than in a single-channel system. By increasing the signal power in a multi-channel system, it is impossible to reduce the impact of transient interference, since at the same time the power of the latter also increases, and in the case of interference of non-linear origin it grows even faster than the signal power.

Phase separation of signals. Let us now consider a set of sinusoidal signals:

Here, the information to be transmitted is contained in changes in the amplitude A k (amplitude modulation), the carrier frequency of the signals ω 0 is the same, and the signals differ in the initial phases φ k .

Among the set N of signals (9.16), only any two signals are linearly independent; any n>2 signals are linearly dependent. This means that at one carrier frequency ω 0 with arbitrary values ​​of amplitudes A i and A k and phases φ i and φ k, only two-channel transmission * can be provided.

* (Signal separation at fixed values ​​of amplitudes A i and phases φ i is discussed in § 9.5.)

In practice, the value φ 2 - φ 1 = π/2 is predominantly used:

s 1 (t) = A 1 sin ω 0 t; s 2 (t) = A 2 sin (ω 0 t+π/t) = A 2 cos ω 0 t, (9.17)

In this case, the signals s 1 (t) and s 2 (t) are orthogonal, which facilitates the implementation of the system and improves its energy performance.

The communication line is the most expensive element of the communication system. Therefore, it is advisable to carry out multi-channel information transmission over it, since as the number of channels N increases, its throughput increases. S. Poichem. the following condition must be met:

N K - productivity of the k-th channel.

The main problem of multi-channel transmission is the separation of channel signals at the receiving side. Let us formulate the conditions for this separation.

Let it be necessary to organize the simultaneous transmission of several messages over a common (group) channel, each of which is described by the expression

(7.1.1)

Taking into account formula (7.1.1.) we obtain:

In other words, the receiver has selective properties with respect to the signal Sk(t).

When considering the issue of signal separation, a distinction is made between frequency, phase, and time separation of channels, as well as separation of signals by shape and other characteristics.

Second study question

Frequency division

The block diagram of a multichannel communication system (MCS) with frequency division of channels (FDC) is shown in Fig. 7.1.1, where it is indicated: IS - signal source, Mi - modulator, Fi - filter of the i-th channel, Σ - signal adder, GN - carrier generator, PRD - transmitter, LS - communication line, IP - source of interference, PRM - receiver, D - detector, PS - message recipient.

Fig.7.1.1. Block diagram of a multi-channel communication system

In FDM, the carrier signals have different frequencies fi (subcarriers) and are spaced at an interval greater than or equal to the spectral width of the modulated channel signal. Therefore, the modulated channel signals occupy non-overlapping frequency bands and are orthogonal to each other. The latter are summed (frequency compressed) in block Σ to form a group signal, which modulates the oscillation of the main carrier frequency fн in block M.

All known methods can be used to modulate channel transporters. But the communication line frequency band is used more economically with single-sideband modulation (SBP AM), since in this case the spectrum width of the modulated signal is minimal and equal to the spectrum width of the transmitted message. In the second stage of modulation (with a group signal), AM OBP is also more often used in wired communication channels.

Such a double-modulated signal, after amplification in the PRD block, is transmitted via a communication line to the PRM receiver, where it is subjected to the reverse conversion process, i.e. demodulation of the signal along the carrier in block D to obtain a group signal, separating channel signals from it using bandpass filters Fi and demodulation of the latter in blocks Di. The central frequencies of bandpass filters Фi are equal to the frequencies of channel carriers, and their transparency bands are equal to the width of the spectrum of modulated signals. Deviation real characteristics ideal bandpass filters should not affect the quality of signal separation, so frequency guard intervals between channels are used. Each of the reception filters must pass without attenuation only those frequencies that belong to the signal of a given channel. The filter must suppress the signal frequencies of all other channels.

The frequency separation of signals by ideal bandpass filters can be represented mathematically as follows:

where g k is the impulse response of an ideal bandpass filter that passes the frequency band of the kth channel without distortion.

Main advantages of CRC: simplicity of technical implementation, high noise immunity, the ability to organize any number of channels. Flaws: inevitable expansion of the used frequency band with an increase in the number of channels, relatively low efficiency of using the communication line frequency band due to filtering losses; bulkiness and high price equipment, due mainly a large number filters (the cost of filters reaches 40% of the cost of a system with FDM). In railway transport, an MKS with a K-24T type PRK has been developed, which uses small-sized electromechanical filters.

Third study question

Frequency channel division: The essence of the frequency channel division method is as follows. Since any real signal must contain the overwhelming majority of its energy within a frequency spectrum limited in width, when organizing multi-channel communication, a certain portion of the total frequency band transmitted by the line is allocated for transmitting signals from each individual channel. Thus, the transmitting device of each sender is obliged to send signals into the line, the frequency spectrum of which completely fits into the allotted this channel frequency band. At the receiving end of each communication channel, a set of voltages of all frequencies is created, forming a linear multichannel communication signal. To isolate frequency voltages that represent a message belonging to a particular sender and suppress voltages of other frequencies, the receiving device must contain frequency filters. The frequency filter of each channel will pass only the frequency spectrum of its channel and will not pass the frequencies of other channels. Separating signals using frequency filters is called frequency division. In the case of frequency division, the condition for the absence of mutual interference between channels is that the signals of different channels must be placed in non-overlapping frequency bands, that is, that none of the frequencies of a given channel fall into the frequency band of other channels. When organizing m-channel communication at the receiving end, it is necessary to have the same number of frequency filters to separate signals from different senders. Conventional oscillating circuits and bandpass filters can be used as frequency filters (for example, in broadcast receivers).

Temporary method of channel division: Pulse transmission methods also allow for the organization of multi-channel communication with time division of channels. In time-sharing systems, the communication line is alternately presented for transmitting signals from different senders using a rotating switch (distributor). The distributor at the receiving end selects signals by time, i.e., separates the signals from different channels. In this case, each channel is allocated a certain part of the total time of line use. To completely separate the signals, it is necessary that switches P and P2 rotate with same speed(synchronously).

In addition, the switches must simultaneously connect either the first pair of correspondents or the second (in phase) to the line. In other words, with time division, signals belonging to a given channel are transmitted in time intervals that are free from signals from other channels. The condition for separability of signals during time division is that the signals of different channels do not overlap in time. Separation by level: It is interesting to consider the case when signals from different channels are not only transmitted simultaneously, but also coincide in form, i.e. their frequency spectra overlap . The signals differ only in magnitude (for example, amplitude). Let there be a three-channel system with signal separation by amplitude. Let us agree that the signals of the first channel are transmitted with (amplitude Si = l, the second channel - with an amplitude s2 = 2, and the signals of the third channel are transmitted in pulses with a height of s3 = 3. It turns out that such a choice of signal levels of different channels does not allow them to be separated in place In fact, if, for example, a signal with level 5 = 3 is received, it is impossible to say whether this corresponds to the transmitted signal of the third channel 5 = s3 = 3 or the sum of the signals of the second and first channels 5 = s2 + S = 2 + 1 = 3. To form separable signals, you need to select the signal levels according to a certain rule. Let us pay attention to the fact that in the simplest case of a two-channel line, the signals of both channels S and s2 can always be separated, if only their amplitudes differ from each other. To isolate the signal of the second channel, the signal of the first channel, amplified by K times, is subtracted from the sum of the signals. Thus, the separating device can be constructed in accordance with the block diagram. Current flows through the diode when the signal voltage does not exceed the voltage value.

The maximum limiter can be implemented according to the diagram. Here, current will flow through the diode only when the signal voltage is greater than the voltage. The resistance R is chosen so large that when current flows through it and the diodes, the voltage drop across the diodes can be neglected. In other words, when current flows through the diode, the output voltage decreases sharply, i.e., it is limited. The choice of limit levels is determined by the required value, as well as the signal levels. By adjusting the voltage, you can set any level of limitation. It should be noted that in this case, signal separation turned out to be possible only with the use of nonlinear elements - limiters. (Under nonlinear elements understand those in which the dependence of current on voltage differs from direct proportionality). Combination separation: Are there any other features that can be used to separate signals? From this point of view, it is useful to become familiar with the so-called combinational separation. Let's start again with the simplest case of a two-channel system. Let both channels operate in binary code with elements 0 and 1; in this case, four different combinations of signals are possible in both channels: if a signal equal to 1 is received, then it is not known which channel it belongs to. However, all four combinations are different from each other. Therefore, instead of the total signal, the combination number can be transmitted, since this number uniquely identifies the signals of each channel. The task comes down to transmitting four numbers, and these numbers can be transmitted in various ways (with any code and modulation). With such transmission, the linear signal is a reflection of a certain combination of signals from different channels.

Separation of signals based on differences in combinations of signals from different channels is called combinational separation. A well-known example of Raman separation is the system of two-channel frequency telegraphy (DCF). Four different frequencies are used to transmit four signal combinations. In the general case of an M-channel system, the base of the code will require transmitting a linear signal consisting of N = nM different combinations. Each combination will correspond to the signal of a specific channel. Both in cases of frequency and time separation, and in cases of separation based on other characteristics, it was assumed that the separating devices completely separate the signals of different channels. However, in real conditions there is always mutual interference between channels. We will now move on to clarifying the nature of these interferences.

Mutual interference between channels: From the point of view of the possibility of completely separating signals by their frequency, the task comes down to creating ideal frequency filters that would respond only to sinusoidal oscillations in a certain frequency band and would not respond at all to oscillations of other frequencies. The frequency method of signal separation was originally based on the phenomenon of harmonic resonance in an oscillating circuit. Resonance is the property oscillatory circuit respond most strongly to harmonic (sinusoidal) vibrations included in a certain band near its resonant frequency, and respond with less intensity to vibrations of other frequencies. The properties of the contour as a selective element are described quite fully by its frequency response, i.e., the dependence of the magnitude of the response, for example, the voltage at the output of the circuit, on the frequency of the applied voltage at the input. The frequencies of adjacent channels passing through the circuit interfere with the reception of the useful signal. Some reduction in the interfering effect is obtained when using a system of coupled circuits as a frequency filter.

In this case, the voltages of the interfering frequencies are weakened more strongly than by a single circuit. However, even here it is not possible to completely suppress the interfering effect. Therefore, when designing multi-channel lines in real conditions, it is necessary to take into account the interfering effect between channels. To reduce mutual interference between channels, so-called guard bands are left. The presence of mutual interference leads to a decrease in the capacity of the communication line, as well as a decrease in the capacity of each channel. Mutual interference between channels also occurs when signals are separated in time. Every communication line, by its physical nature, contains elements capable of accumulating electrical energy. When transmitting signals, this accumulative property of the line manifests itself in its “inertia”. Such inertial elements are, for example, the inductance of wires and the capacitance between them when transmitted over wire communication lines. Let the sum of signals act at the input of a wired two-channel line. Then, due to the presence of inductance and capacitance in the line, the shape of the output signals will be noticeably distorted. The greater the capacitance and inductance of the line, the stronger the distortion. Distortion is caused by the fact that the energy stored in the line from the signal of the first channel is summed up at the line output with the energy of the signal of the second channel. The capacity of a multi-channel link is thus limited even in the absence of any other interference other than inter-channel interference. From the above examples, it becomes clear that when organizing multi-channel communication lines, it is necessary to take additional measures to reduce mutual interference between individual channels.

communication radio wave frequency radio station

Signal separation is the provision of independent transmission and reception of many signals over one communication line or in one frequency band, in which the signals retain their properties and do not distort each other.

With phase separation, several signals are transmitted at one frequency in the form of radio pulses with different initial phases. For this, relative or phase-difference keying is used (conventional phase modulation is used less frequently). Currently, communications equipment has been implemented that allows simultaneous transmission of signals from two and three channels on one carrier frequency. Thus, several channels for transmitting binary signals are created in one frequency channel.

In Fig. 11.3a shows a vector diagram of double phase shift keying (DPSK),

providing transmission of two channels on the same frequency. In the first phase channel, zero (a pulse of negative polarity) is transmitted by currents with a phase of 180°, and one (a pulse of positive polarity) by currents with a phase of 0°. The second phase channel uses currents with phases of 270 and 90°, respectively, i.e., the signals of the second channel move in relation to the signals of the first channel by 90°.

Suppose that it is necessary to transmit code combinations 011 in the first channel (Fig. 11.3, c) and 101 in the second (Fig. 11.3, d) using the DMF method at one frequency. The process of phase manipulation for the first channel is shown by solid lines, and for the second - by dotted lines (Fig. 11.3,6,e)). Thus, each code combination has its own sinusoidal voltage. These sinusoidal oscillations are added and a total sinusoidal oscillation of the same frequency is sent to the communication line, which

indicated by dash-dot in Fig. 11.3, d. It is also shown here that in the interval 0 - t1

zero is transmitted over the first channel and one through the second channel, which corresponds to

transmission of vector A with a phase angle of 135°. In the interval t1 – t2, the transmission of one through the first channel and zero through the second corresponds to vector B with an angle of 315°. and in the interval t2 – t3 - vector C with an angle of 45°, since units are transmitted via the first and second channels.

The block diagram of the device for implementing DMF is shown in Fig. 11.4. The carrier generator Gn has a phase-shifting device FSU to obtain a phase shift of the sinusoidal oscillation by 90° in the second channel. Phase modulators

FM1 and FM2 carry out manipulation in accordance with Fig. 11.3, d), and the adder Σ performs the addition of sinusoidal oscillations. On reception after amplifier

The separation of both channels is carried out in phase detectors - demodulators FDM1 and FDM2, to which a reference carrier voltage is supplied from the Gonn generator,

coinciding in phase with the voltage of this channel. For example, upon admission with

amplifier of the total sinusoidal voltage (vector A in Fig. 11.3, b) on

positive voltage will be allocated to the demodulator of the first channel FDM1,

corresponding to phase 0° (reception of one on the first channel), since the phase of the reference

carrier frequency coincides with the phase of the first channel. Vector A can be decomposed into two

components: Af = 0 and Af = 90. In FDM1, the signal component Af = 0 interacts with

reference voltage applied to this channel, and the Af component will be suppressed

(the signal voltage of the second channel will not appear at the output of FDM1, since the vector

reference frequency is perpendicular to the phase of the voltage vector of the second channel and

the product of these vectors will be equal to zero. At the same time, in FDM2 the arrival

the total sinusoidal voltage (vector A) will create a positive voltage corresponding to a phase of 90° (reception of one in the second channel),

since the phase of the reference frequency is shifted by 90° compared to the reference frequency of the first

channel coincides with the phase of the second channel. Signal voltage of the first channel to the output

FDM2 will not arrive, since the reference frequency vector in this channel is perpendicular

voltage vector of the first channel and the product of these vectors will be equal to zero.

Similarly, two messages can be transmitted on the same frequency with

relative phase shift keying (RPKM). Thus, the use of DFM or

DOFM allows you to double the capacity of the communication channel. It is also possible

transmitting three messages on the same frequency using triple relative

If we consider the simplest network consisting of two points A and B, between which N digital channels are organized (it is not specified here how), then independent transmission of signals over these channels is possible if these channels separated between themselves. The following ways of dividing channels between two points are possible:

Space division, using different transmission media to organize channels;

Time division, which transmits digital signals at different time intervals in different channels;

Code division, in which division occurs by applying specific code values ​​to each signal;

Wavelength separation at which digital signals transmitted via digital channels organized at different wavelengths in an optical cable;

Division by fashion when organizing a channel into various types electromagnetic wave (mods) of hollow waveguides and optical cable;

Separation by electromagnetic wave polarization of hollow waveguides and optical cables.

In all cases, the separation of channels between two nodes does not imply the presence of a single propagation medium for the electromagnetic signal. To transmit signals in one propagation medium, channels separated by one or another characteristic (except spatial) are grouped using the operation of combining (multiplexing), forming a digital transmission system (DTS).

In digital switching systems (DSS), such combining and dividing signals most often occurs using time division multiplexing. Time multiplexing is now an important integral part not only the DSP, but also the CSK, and plays a decisive role especially at the junction of these systems. In telephony, time multiplexing is defined as a tool for distributing (dividing and combining) telephone channels over time when transmitted over a single physical communication line. In this case, one of the types of pulse modulation is used. Each pulse corresponds to a signal from one of the channels; signals from different channels are transmitted sequentially.

The principle of temporary signal combining is shown in Fig. 1.8, which shows a rotating commutator TO(center), alternately connected to the outputs of the channel sequence. The switch is connected to the output of channel 1 at the time t, to the output of channel 2 at time t2, to the output of channel N at time tN, after which the process is repeated. The resulting output signal will consist of a sequence of signals from different channels, offset relative to each other for a time At.

The separation of signals on the receiving side will occur in a similar way: a rotating commutator is alternately connected to the channels, transmitting the first signal to channel number 1, the second to channel number 2, etc. Obviously, the operation of the switches on the receiving and transmitting sides must be synchronized in a certain way so that the signals arriving along the line are directed to the required channels. In Fig. 1.9 shows timing diagrams for the case of combining three channels through which amplitude-pulse modulated signals are transmitted.

As mentioned above, DSPs use PCM signals, which are digital code sequences consisting of several bits.

Temporary association several PCM signals is the combination of code sequences coming from different sources for joint transmission over a common line, in which the line at any given time is provided for transmission of only one of the incoming code sequences.

Temporal combining of PCM signals is characterized by a number of parameters. Cycle time combining is a set of consecutive time intervals allocated for the transmission of PCM signals coming from various sources. In the time combining cycle, each PCM signal is allocated a specific time interval, the position of which can be uniquely determined. Since usually each signal corresponds to its own transmission channel, such a time interval allotted for the transmission of one channel is called time slot(CI). There are two types of cycle - basic, whose duration is equal to the signal sampling period, and supercycle - a repeating sequence of successive main cycles, in which the position of each of them is uniquely determined.

Rice. 1.8. Circular interpretation of time multiplexing

Rice. 1.9. Temporary association

When constructing PCM equipment, they use homogeneous temporary association PCM signals, in which the codeword transmission rates of the combined PCM signals are the same. This makes it possible to produce group association PCM signals and build hierarchical systems for transmitting PCM signals on this basis.