RU2090975C1 - Multichannel signal transmission technique - Google Patents

Abstract

FIELD: electric communications including multichannel communication; multichannel analog transmission systems. SUBSTANCE: on sending end, channel signals are multiplied by harmonic carriers of waves, signals obtained are summed up and transmitted through common channel; on receiving end, in order to discriminate channel signals, common signal is multiplied by carriers of waves of same frequencies and filtered off; novelty is that on sending end, each of N = 2ⁿ channel signals is simultaneously multiplied by n+K (n=1, 2.. . , K - 0, 1, 2...) harmonic carriers which have one of their frequencies higher than maximum frequency of channel signals, frequency of each next one being twice as high as that of preceding one, frequency of last carrier is twice or more higher than that of preceding one, and phase of each carrier is of one of two possible values differing by value other than a multiple of 180 deg; for different channel signals of phase, at least K+1 carriers are different; on receiving end, in order to discriminate definite channel signal, common signal is simultaneously multiplied by n+K carriers of same frequencies; while on transmission end, phases of carriers of particular channel signal are of one of two mentioned possible values, on receiving end, phases of carriers differ by 90 deg from other possible value. EFFECT: provision for similar conditions in transmitting all channel signals in common channel; improved frequency characteristics of channels without noticeable extension of spectrum inherent to digital transmission systems, simplified hardware implementing this technique. 2 dwg

Description

 The invention relates to electrical communications, in particular to multi-channel communications.

 A known method of multichannel communication with frequency division of channels (method with PDA) [1] in which each channel signal is multiplied by harmonic oscillation at a frequency determined for each channel, after which the signals of all channels are summed and transmitted to the line (into the common path), and at the reception, to isolate the signal of a certain channel, the total transmitted signal is multiplied by harmonic oscillation of the same frequency and filtered by a low-pass filter.

 However, with this method, different channel signals occupy different frequency ranges in the common path, as a result of which the attenuation and interference for different channels are different, hence the quality of communication over different channels is different. For example, on cable lines in the worst-case conditions are usually the channels transmitted at the top of the frequency range.

 In addition, the spectrum of the channel signal multiplied by the carrier wave is twice as wide as the spectrum of the original channel signal (and, accordingly, twice as wide as the minimum necessary for transmitting the frequency band). Spectrum reduction is achieved by isolating from the modulated signal only one side frequency band using electric filters. But at the same time, firstly, the complexity of the equipment due to the introduction of filters is required, and secondly, the frequency characteristics of communication channels are deteriorating: the ability to transmit low-frequency (below 300 Hz) spectral components of channel signals is lost, significant phase-frequency distortions appear. And although the quality of the transmission of speech signals decreases slightly, the transmission of other signals (digital, fax, etc.) is possible only through the introduction of special devices (modems).

 There is also a known method of multichannel communication with time division of channels (WRC method), in which each signal is sampled by transmission and the thus obtained pulses of samples of channel signals are transmitted to the common path one by one, and at the reception, a time selector is selected to isolate the channel signal from the general pulse stream samples of this channel signal, after which a low-pass filter filters them, restoring signal continuity.

 However, with this method, the overall signal is characterized by a large absolute spectrum width characteristic of pulsed signals, and, even worse, the spectrum starts at zero frequency, so that it has an infinite relative width. Signals with such a spectrum inevitably undergo significant linear distortions in the general transmission path, as a result of which mutual influences arise between the channels. In addition, with such a spectrum of information signals, difficulties arise in transmitting synchronization signals for controlling the temporary selectors of the receiver, and inaccurate temporal selection further increases the mutual influences.

 The foregoing leads to the fact that in the time division it becomes impossible to transmit analog speech signals, but only digital can be transmitted. Therefore, samples are encoded in the transmission after sampling, converting the signals into digital ones, and, in addition, additional transcoding (linear coding) is performed to eliminate the low-frequency components of the total stream of digital pulses, and the reverse operations are performed for receiving channel signals to recover. All this complicates the equipment and expands the spectrum of the signal transmitted in the common path (with the existing coding standard for each sample with eight digital pulses, the required spectrum width increases by 8 times. The expansion of the spectrum, combined with the increase in the attenuation of the line at high frequencies, leads to a corresponding reduction in the length of the amplification (regenerative ) sections in comparison with the length of amplification sections of systems with frequency division.

 As a prototype of the present invention, the method of frequency division multiplexing (PDK) is selected as the closest in terms of signal conversion operations on transmission and reception.

 The objective of the invention is to provide the same conditions for the transmission of all channel signals in a common path and to improve the frequency characteristics of the channels without significantly expanding the spectrum inherent in digital transmission systems, as well as simplifying the apparatus implementing this method.

This problem is solved by the fact that in the method of multi-channel signal transmission, in which channel signals are multiplied by harmonic carrier oscillations, the received signals are summed and transmitted in the common path, and at the reception, channel signals are multiplied by carrier oscillations of the same frequencies and filter, introduce distinctive features, namely, that each of the N 2 ⁿ channel signals is transmitted simultaneously by multiplying by n + K (n 1,2, K 0, 1, 2,) harmonic carriers, the frequency of one of which is greater than the maximum the frequency of the channel signals, the frequency of each of the next two times the frequency of the previous one, the frequency of the last carrier is twice or more than twice the frequency of the previous one, and the phase of each carrier has one of two possible values that differ by an amount not multiple of 180 ^o , and for different channel signals the phases of at least K + 1 carriers are different, and at the reception, to select a specific channel signal, multiply the total signal simultaneously by n + K carriers of the same frequencies, and if I have one of said two possible values, then the reception carrier phases differ by 90 ^o from other possible values.

 Distinctive features of the method provide a solution to the problem of the invention.

Consider the advantages of this method of generating signals in transmission. The signal in the common path is the sum of N 2 ⁿ channel signals, each of which is multiplied by n + K carriers with frequencies selected according to the indicated rule.

где ω_i=2πf_i,
f_i частота i-й несущей;
Φ_ei= Φ_1i или Φ_2i выбранное на передаче значение фазы этой несущей.

where ω _i = 2πf _i ,
f _{i is the} frequency of the i-th carrier;
Φ _ei = Φ _1i or Φ _2i is the phase value of this carrier selected on the transmission.

максимальная, аналогично, равна f_n+k+(N2-1)f₁, а расстояние между соседними спектральными составляющими равно 2f₁.Find the signal spectrum (1). At first, the spectrum of the product of the carriers. Using trigonometric transformations, this product can be represented as a sum of 2 ^{n + k-1} spectral components with frequencies f _{n + k} ± f _{n + k-1} ± f _{n + k-2} ±. ± f ₁ , and each combination of signs corresponds to the frequency of one of the components. The minimum frequency of the spectral component is

the maximum, likewise, is equal to f _{n + k} + (N2-1) f ₁ , and the distance between adjacent spectral components is 2f ₁ .

Now the spectrum of each modulated signal (of each summand (1) is easy to find as two sidebands corresponding to the spectrum of the original channel signal located near each spectral component of the product of the carriers. Moreover, the minimum frequency of each modulated signal
f _{n + k} - (N2 ^k-1 -1) f ₁ -f _max ,
maximum -
f _{n + k} + (N2 ^k-1 -1) f ₁ + f _max ,
and the absolute width of the spectrum
N2 ^k f ₁ -2 (f ₁ -f _max ) N2 ^k f ₁ ,
where f _{max is the} maximum frequency of the spectrum of the original (unmodulated) channel signal.

FIG. 1 illustrates what has been said for a special case of an 8-channel system (N 8, n 3), at K 1 with channel signals having a frequency of less than 4 kHz, for example, telephone. We select the carrier frequencies in accordance with the rule specified in the method: f ₁ 4 kHz, f ₂ 8 kHz, f ₃ 16 kHz, f ₄ ≥32 kHz, and we will consider several cases of choosing f ₄ , namely 32 kHz, 48 kHz and 64 kHz .

On the first line of FIG. Figure 1 shows the spectrum of the channel signal, on 2–4 lines the spectrum of the product of the carriers with different choices of f ₄ , and on 5–7 lines the spectrum modulated by multiplying by n + k carriers of the channel signal for different f ₄ .

From the result it follows that:
1. The spectra of all modulated signals occupy the same frequency range, so the method provides the same transmission conditions for all channel signals in a common path;
2. The rate of decrease in the amplitude of the spectrum at the upper and lower boundaries is equal to the rate of decrease in the amplitude of the spectrum of the original channel signals, therefore, filters in the common path are not needed. Thus, the method solves the problem of improvement compared with the prototype of the frequency characteristics of the channels and, in addition, provides the transmission of the spectral components of the channel signals starting from zero frequency;
3. If you select f ₁ 4 kHz, then the absolute width of the spectrum with a minimum K (K 0) coincides with the minimum required for the prototype method (method with PDA). An increase in K by one leads to a doubling of the width. At K 3, the width of the spectrum is equal to the minimum required for the digital transmission method (with the accepted standard for the representation of signal samples by 8-bit binary code). It is shown below that an increase in K improves the noise immunity.

Thus, the method allows you to keep the absolute bandwidth the same as in the prototype method, and in addition, allows you to increase the noise immunity by expanding the frequency band;
4. An additional advantage of the method, especially compared to the digital transmission method, is that by selecting a carrier with a maximum frequency, you can place the spectrum of the modulated signal in any frequency range, providing the desired relative width and the ability to use the low-frequency part of the range for transmitting synchronization signals and other overhead signals.

Now we show that the method allows the selection of carrier phases in the transmission so that at least K + 1 phases are different for different channel signals. To this end, we note that n + K independent carriers with one or the other possible phases can each be considered as a code combination of a binary code in which, for example, 0 indicates the selection of the first phase, and 1 indicates the selection of the second phase. With such a representation, the problem boils down to choosing N 2 ⁿ allowed combinations from all possible 2 ^{n + K} combinations of the block binary code so that the minimum Hamming distance between the allowed combinations (in this case, the number of carriers with different phases) is the maximum possible. In coding theory, this problem is one of the simplest, for example, in [2] it was shown that the Hamming distance for the code (n + K, n) is K + 1, which corresponds to the indicated feature of the invention.

где Φ_ri другая по сравнению с выбраной на передаче для M-го сигнала фаза i й несущей. В результате умножения получим

Раскрывая произведения, входящие в эту сумму, получим
произведения канальных сигналов на сравнительно высокочастотные колебания с частотами не менее удвоенной минимальной частоты несущих. Эти произведения будут подавлены фильтрами нижних частот при последующей фильтрации детектированных сигналов;
пропорциональные канальным сигналам (относительно низкочастотные) колебания вида

При этом для выделяемого μ M-го сигнала для всех несущих v_ri и Φ_ei разные, т. е. если l 1, то r 2, и наоборот. Поэтому произведение не обращается в ноль, и, следовательно, сигнал действительно выделяется. Кстати, отсюда следует, что, во-первых, две возможные фазы каждой несущей не должны отличаться на величину, кратную 180^o, т. к. при этом даже та строка выражения (4), которая соответствует μ M, обратилась бы в ноль, во-вторых, если желательно получить возможно большую интенсивность выделяемого сигнала, то разность двух возможных фаз каждой несущей следует брать равной 90^o или близкой к этому значению. Тогда все сомножители под знаком произведения приобретают максимальное (по модулю) значение, равное 1.Finally, we prove that the reception rule specified in the description of the invention really provides the selection of one of the channel signals while suppressing the rest. As indicated in the description of the method, at the reception, to isolate a certain signal, the transmitted common signal is multiplied by n + K carriers of the same frequencies, and if the phases of the carriers in the transmission have one of two possible values for this channel signal, then at the reception the phases of the carriers should differ from other values by 90 ^o . Following this rule, to isolate, for example, a signal with μ M, we multiply all terms of the common signal

where Φ _{ri is} different from the phase of the ith carrier selected in the transmission for the Mth signal. As a result of multiplication, we obtain

Opening the works included in this amount, we get
products of channel signals with relatively high-frequency oscillations with frequencies of at least twice the minimum carrier frequency. These works will be suppressed by low-pass filters during subsequent filtering of the detected signals;
proportional to channel signals (relatively low-frequency) oscillations of the form

Moreover, for the extracted μ Mth signal for all carriers, v _ri and Φ _{ei are} different, i.e., if l 1, then r 2, and vice versa. Therefore, the product does not vanish, and, therefore, the signal really stands out. By the way, it follows that, firstly, the two possible phases of each carrier should not differ by a multiple of 180 ^o , since even that line of expression (4), which corresponds to μ M, would go to zero, secondly, if it is desirable to obtain the largest possible intensity of the allocated signal, then the difference between the two possible phases of each carrier should be taken equal to 90 ^o or close to this value. Then all the factors under the product sign acquire the maximum (modulo) value equal to 1.

Consider other terms of the sum (4) associated with the signals m ≠ M, i.e. those to be suppressed. In these signals, at least K + 1 carriers have different phases compared to the phases of the carriers of the extracted signal, which means that the phase difference of the corresponding factors sin (Φ _ri -Φ _ei ) vanishes, so that at least K + 1 factors are equal to zero , which ensures the suppression of all signals except the allocated.

, а общая помехозащищенность от влияющего сигнала равна

где P номера несущих, для которых фазы выделяемого и подавляемого (влияющего) сигналов различны.Under real conditions, the phase of each i-th receiving carrier may differ from the required one. If this (relatively small) difference is denoted by δ _i (radian), then the corresponding factor is sinδ _i ≈ δ _i , so that the actual noise immunity from the influencing signal due to the difference in phases of the extracted and suppressed signals on the ith carrier is

, and the total noise immunity from the influence signal is

where P is the number of carriers for which the phases of the allocated and suppressed (influencing) signals are different.

 It follows that an increase in K, although it leads, as shown above, to the expansion of the signal spectrum, but it can dramatically increase the noise immunity from mutual influences between channel signals or, at the same noise immunity, reduce the requirements for the accuracy of the installation of the carrier phases at the reception. Thus, the proposed method allows you to find the optimal solution in one sense or another for specific conditions.

 A possible implementation of the method is illustrated in FIG. 2, which shows a transmitter and a receiver.

The transmitter comprises a block of multipliers 1, an adder 2 and a block of transmission generating equipment 3 with 2 (n + K) outputs indicated in FIG. 1 as Φ _{1, i} and Φ _{2, i} (i 1,2, n + K), which means the output of the ith carrier with its possible phases 1 and 2, as well as with the output of the clock signal. The multiplier block 1 consists of N (n + K) multipliers, each of which has two inputs for the multiplier signals and one output of the product signal. Multipliers are divided into N groups according to the number of channel signals with n + K cascade-connected multipliers in each group (according to the number of carriers). The first input of the first multiplier of each group is connected to the source of the corresponding channel signal, the outputs of the last multipliers of the groups are connected to the inputs of the adder 2, the output of each is connected to the input of the common path, and the second inputs of the ith multipliers of each groups of block 1 are connected to the corresponding outputs Φ _{1, i} and Φ _{2, i of} block 3 with a selection of carrier phases, as indicated in the method.

 As a result, the output signal of each μth (m 1,2, N) group of multipliers is a channel signal multiplied by n + K harmonic carriers, each of which has one of two possible phases, and the common signal in the common path is the sum of multiplied signals and corresponds to the expression (1).

As multipliers, industrial multiplier chips can be used, for example, K 525 PS 2 A, B [3]
The master (lowest frequency) generator of the generator equipment can be performed in various ways, for example, as in [4] where the generator circuit is shown with two phases of the output signal differing by 90 ^° . Other carrier generators with frequencies, double, quadruple, etc. higher, can be obtained from the master, for example, using the same multiplier K525 PS 2A, B, turned on according to the frequency doubling scheme [3] The receiver contains a block of reception multipliers 4, identical to a block of transmission multipliers, a block of low-pass filters 5, consisting of filters according to the number of channel signals, and the receiving generator equipment block 6 with the synchronization signal input and 2 (n + K) outputs v _{1, i} and Φ _{2, i} (i 1,2, n + K), which means the outputs of the i-th harmonic carrier with two phases different from the phases of the transmission carriers by 90 ^o . The outputs of the first group multipliers of the block of reception multipliers are connected in parallel and connected to the output of the common path, the outputs of each last group multiplier are connected to the inputs of the corresponding filters of block 5, the outputs of which are channel outputs, the second input of the i-th multiplier of each group of block 4 is connected to the output Φ _{1 , i} and Φ _{2, i of} block 6, the phases being selected as indicated in the method.

 The most popular devices in the considered circuit are multipliers. They are required for N (n + K) in transmission and reception, and if you take into account the reverse direction of transmission, then for each two-way channel you need 4 (n + K) multipliers / channel.

 The number of multipliers can be reduced if we take into account that at the last stage of multiplication by transmission, instead of N multipliers, you can do just two, parallelizing all the outputs of the penultimate stage, coming to the input of the multiplier with one or another phase of the last carrier. Similarly, in the penultimate stage, 4 multipliers are sufficient, etc. The same at the reception. With this construction of a circuit for one two-way channel, 4 (K + 2 / N) multipliers are needed.

Sources of information
1. Shuvalov V.P. Katunin G.P. Kruk B.I. and other telecommunication systems. Textbook for high schools / Ed. V.P. Shuvalova. M. Radio and Communications, 1987.

 2. Zyuko A.G. Klovsky D.D. Nazarov M.V. Fink L.M. Theory of signal transmission. Textbook for high schools. M. Communication, 1980.

 3. Ataev D. I. Bolotnikov V.A. Analog integrated circuits for household radio equipment. Directory. M. MEI, 1993.

 4. Horowitz P. Hill W. Art of circuitry. M. Mir, 1983, v. 1.

Claims (1)

A method of multi-channel signal transmission, in which channel signals are multiplied by harmonic carrier oscillations, the received signals are summed and transmitted in the common path, and at the reception, channel signals are multiplied by carrier oscillations of the same frequencies and filtered, characterized in that by transferring each of the n 2 ⁿ multiplied channel signals simultaneously on n + K (1.2 n, K 0,1,2,) harmonic carrier frequency of one of which is greater than the maximum frequency of the channel signals, the frequency of each next twice h simplicity previous frequency last carrier twice or more than twice the frequency of the previous one, and the phase of each carrier has one of two possible values, which differ by the amount nonmultiple 180 ^o, wherein for the different channel signals in phase at least K + 1 carriers are different, and at the reception, to isolate a specific channel signal, the total signal is simultaneously multiplied by n + K carriers of the same frequencies, and if at the phase transmission of the carriers of this channel signal they have one of the two possible values, then the reception of the phase of the carriers subtract by 90 ^o from another possible value.

Images