RU2118052C1 - Method and device for information transmission in multiple beam channel - Google Patents

Abstract

FIELD: communication, in particular, for multiple-beam channels with fast phase fluctuations and heavy alternating Doppler frequency shifts. SUBSTANCE: method involves use of complex wideband signals by means of carrier phase-shift keying using pseudorandom code and using principle of optimal frequency addition for waves arriving at different routes with different delays detected in course of their matched filtration. Such matching is hard to achieve for fast alternations in signal characteristics. If phase of signal of waves which arrive at different routes with different delays has several non- correlated sign alternations, correlation characteristics of signal are decreased drastically and connection reliability suffers. Method of invention involves additional encoding of code sequence for transmission conforming to condition of relative phase-shift keying, and filling each elementary signal with phase which is increased from signal to signal as quadratic function. When signal is received, it is delayed by duration of single element of complex signal to be transmitted, its frequency is shifted by total value of intermediate frequency and value which is equal to average increment of instantaneous frequency of carrier wave for delay time. This value is multiplied by source wave which has been received and wave at additional frequency is detected. Signal at intermediate frequency is processed as in channel without Doppler shift with known constant phase shifts for waves which arrive with different delays. EFFECT: increased reliability and validity of signal reception. 3 cl, 3 dwg

Description

 The invention relates to communication technology and is intended for noise-free transmission of discrete information over complex multipath channels. Such channels include a high-speed cellular mobile communication channel in urban areas, a long-range short-wave mobile communication channel, a hydro-acoustic communication channel with a mobile underwater vehicle.

 There are a large number of methods and devices in which the problem of error-correcting transmission of discrete information over multi-path communication channels is solved [1, 2]. The patent study revealed the following three development trends for discrete information transmission systems via multipath communication channels: increased noise immunity by using diversity reception methods [3, 4]; increasing the efficiency of communication systems through the use of new methods of adaptive reception [5, 6]; improving the noise immunity of reception due to the use of complex broadband signals and correlation processing methods [7]. The present invention relates to the third development trend of communication systems over multipath channels and proposes a method based on the use of broadband signals with corresponding correlation conversion on the receiving side for building communication systems, which makes the system insensitive to multipath signal propagation and random variable Doppler shifts of the oscillation frequency, coming in different ways.

 A prototype of the invention is the method underlying the rake system. The rake system that implements this method is considered as a prototype of the claimed devices. This classical system is described in a large number of sources [1, 2], most in detail in the monograph [1]. The "Rake" system is built in accordance with a method involving the transmission of discrete information using an ensemble of noise-like signals, and upon reception, the signals received by different beams are separated by the time of their arrival using matched filtering or correlation filter processing, followed by their temporary combination and quasi-coherent association before making a decision. The Rake system uses two noise-like orthogonal signals to transmit information. Separation of oscillations arriving at the receiving point along different paths with different delays is carried out using correlation filter processing. The received oscillation is divided into two reception channels corresponding to two transmitted orthogonal noise-like signals. In each channel, the received oscillation enters the delay line with taps made with a uniform step Δt, which is equal to the time-resolving ability of the noise-like signal. The signal from each tap of the delay line using the mixers is multiplied by the corresponding noise-like signal at the carrier frequency and then by harmonic oscillation with the amplitude and phase equal to the estimates of the amplitude and phase of the oscillation arriving with a given delay. The resulting oscillations are summed up, filtered in a band equal to the reciprocal of the time duration of the transmitted signal, and after detection, they are sent to a comparator, where it is decided that a signal corresponding to a channel with a high detector voltage level is transmitted. Harmonic oscillations with amplitudes and phases, which are estimates of the amplitudes and phases of signals with different delays, are supplied to the multipliers-mixers from their other inputs after frequency offset and separation in narrow-band filters as a result of accumulation over a long period of time significantly exceeding the duration of the transmitted signal period. This part of the circuit can be allocated as a channel parameter estimator.

 The rake system works well in a multipath channel with very slow fading and signal fluctuations. It is assumed that in the interval of the duration of the signal transmitted through the channel, the phase of the oscillations arriving at individual beams remains constant. With an increase in the rate of phase fluctuations and the Doppler frequency shift of the received oscillations, their correlation with the reference signal decreases, the coordinated processing is violated, the estimates of the amplitude and phase of the oscillations made in the previous steps become incorrect, the system stops working.

(modL), где суммирование производится по модулю L; x_j-j-й символ исходной кодовой комбинации;

- символы результирующего кода. Сложный передаваемый сигнал формируют как последовательность элементарных (простых) сигналов, имеющих начальную фазу:

, где j - номер элементарного сигнала; C - константа, коэффициент квадратичного закона изменения начальной фазы несущей. При приеме сигнал задерживают на длительность одного элементарного сигнала τ_s, смещают его частоту на величину

равную суммарному значению промежуточной частоты f_in и значению среднего приращения мгновенной частоты несущего колебания за время задержки τ_s, умножают результирующий сигнал на исходное принимаемое колебание и выделяют сигнал на частоте f_in. Затем этот сигнал обрабатывают также, как в канале без доплеровского смещения на частоте f_in с известными постоянными фазовыми сдвигами.The problem to which this invention is directed is to maximize the reliability and reliability of signal reception in a multipath channel with fast phase fluctuations and strong variable Doppler frequency shifts. This problem is solved by using complex broadband signals based on the manipulation of the carrier phase with a pseudo-random code and using the principle of optimal addition of oscillations arriving along different propagation paths with different delays, taking into account the estimation of channel parameters using correlation-filter processing with subsequent decision making. The proposed method differs from the prototype in that the original code sequence x _{j of the} L-digit code is transformed in accordance with the law of relative phase modulation:

(modL), where the summation is done modulo L; x _{j is the jth} character of the source code combination;

- characters of the resulting code. A complex transmitted signal is formed as a sequence of elementary (simple) signals having an initial phase:

where j is the number of the elementary signal; C is a constant, the coefficient of the quadratic law of change of the initial phase of the carrier. When receiving a signal, it is delayed by the duration of one elementary signal τ _s , its frequency is shifted by

equal to the total value of the intermediate frequency f _in and the value of the average increment of the instantaneous frequency of the carrier oscillation for the delay time τ _s , multiply the resulting signal by the initial received oscillation and select the signal at the frequency f _in . Then this signal is processed in the same way as in a channel without Doppler shift at a frequency f _in with known constant phase shifts.

 Schematic diagrams of systems that implement the proposed method of transmitting information are presented in figures 1, 2, 3.

; 10 - умножитель; 11 - полосовой фильтр с полосой пропускания, равной ширине спектра сигнала, и центральной частотой f₀ + f_in; 12 - полосовой фильтр с центральной частотой, равной промежуточной частоте f_in, и с полосой, равной ширине спектра сигнала; 13 - обычный приемник сигналов, не имеющих фазовых флуктуаций и доплеровского смещения частоты. На фиг. 2 приводится схема устройства (системы передачи информации СПИ-1), реализующая предлагаемый способ, на основе шумоподобного фазоманипулированного сигнала, где 14 - линия задержки с равномерными отводами через интервал, равный величине обратной ширине полосы сигнала; 15, 16, 17, 18 - умножители на опорный (ожидаемый) сигнал, соответствующий данному каналу, на промежуточной частоте f_in с начальной фазой

; 19, 20, 21, 22 - умножители на значения H(k), которые пропорциональны квадратам коэффициентов передачи канала по отдельным лучевым путям распространения с соответствующими задержками t_k; 23 - сумматор; 24 - фильтр нижних частот с полосой, равной ширине спектра сигнала; 25, 26 - соответственно второй и последний каналы приемного устройства; 27 - схема сравнения и принятия решения. На фиг. 3 приводится схема простейшей двоичной системы связи (СПИ-2), реализующей рассматриваемый способ, где 9 - умножитель на сигнал W(t) = sin((2πf_in+2C/τ_s)t+φ₀); 27 - схема сравнения с порогом и принятия решения; 24 - фильтр нижних частот с шириной полосы, равной обратной величине длительности сигнала (интегратор со сбросом); 28 - умножитель на сигнал

.In FIG. 1 is a flowchart showing the general principle of the proposed method, where 1 is a source of digital messages; 2 - encoder; 3 - transcoding device; 4 - modulator; 5 - controlled phase shifter (controlled delay element); 6 - transmitter; 7 - input selective circuit of the receiver; 8 - element delay for the duration of the elementary signal τ _s ; 9 - multiplier (mixer) for the signal

; 10 - multiplier; 11 - bandpass filter with a passband equal to the width of the spectrum of the signal, and the center frequency f ₀ + f _in ; 12 is a band-pass filter with a center frequency equal to the intermediate frequency f _in , and with a band equal to the width of the signal spectrum; 13 is a conventional receiver of signals without phase fluctuations and Doppler frequency shift. In FIG. 2 is a diagram of a device (information transmission system SPI-1) that implements the proposed method, based on a noise-like phase-shifted signal, where 14 is a delay line with uniform taps at an interval equal to the reciprocal of the signal bandwidth; 15, 16, 17, 18 - multipliers by the reference (expected) signal corresponding to this channel at an intermediate frequency f _in with the initial phase

; 19, 20, 21, 22 - multipliers by values of H (k), which are proportional to the squares of the transmission coefficients of the channel along individual radial propagation paths with corresponding delays t _k ; 23 - adder; 24 - low-pass filter with a band equal to the width of the spectrum of the signal; 25, 26 - respectively, the second and last channels of the receiving device; 27 is a diagram for comparing and making decisions. In FIG. 3 is a diagram of the simplest binary communication system (SPI-2) that implements the considered method, where 9 is the multiplier by the signal W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ); 27 is a comparison diagram with a threshold and decision making; 24 - low-pass filter with a bandwidth equal to the reciprocal of the signal duration (integrator with reset); 28 - signal multiplier

.

. Выходы умножителей 15, 16, 17, 18 через умножители 19, 20, 21, 22 на значения H(k), пропорциональные квадратам коэффициентов передачи по отдельным путям распространения сигналов, характеризующиеся задержками t_k, подключены к сумматору 23. Выходы сумматоров отдельных каналов через фильтры 24 нижних частот подключены к схеме 27 принятия решения, которая сравнивает уровни сигналов разных каналов и выдает на выход номер сигнала, соответствующего каналу с максимальным уровнем.SPI-1 consists of transmitting and receiving parts. The transmitting part contains in series: a source of digital messages, an encoder 2, a phase difference modulation transcoding device 3, a phase modulator 4, a controlled phase shifter (controlled delay element) 5 and a transmitter 6. The receiving device contains a selective amplifier 7 with a central frequency equal to the carrier the frequency of the signal f ₀ connected to it through the first input of the multiplier 10, sequentially connected between the output of the amplifier 7 and the second input of the multiplier 10, a delay line 8 for the duration of the electric of a single signal τ _s , the second multiplier (mixer) 9 for harmonic oscillation W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ) and a band-pass filter 11 with a central frequency equal to the sum of the frequencies of the carrier oscillation f ₀ and intermediate frequency f _in . The output of the multiplier 10 is connected to the input of the bandpass filter 12, tuned to the intermediate frequency f _in . The rest of the circuit, connected to the output of this band-pass filter, is a signal receiving device with a known frequency and constant phase based on a transfer filter. It consists of M parallel channels 25, 26. Each channel contains a delay line 14 with K uniform taps that are connected to the multipliers 15, 16, 17, 18 with a reference (expected) signal corresponding to this channel and given delay t _k at the intermediate frequency f _in and with the initial phase

. The outputs of the multipliers 15, 16, 17, 18 through the multipliers 19, 20, 21, 22 by the values of H (k), which are proportional to the squares of the transmission coefficients for individual signal propagation paths, characterized by delays t _k , are connected to the adder 23. The outputs of the adders of individual channels through low-pass filters 24 are connected to a decision circuit 27, which compares the signal levels of different channels and outputs the signal number corresponding to the channel with the maximum level.

SPI-2 consists of transmitting and receiving parts. The transmitting part contains sequentially connected: a source of digital messages 1, a modulator 4, a controlled phase shifter (controlled delay element) 5 and a transmitter 6. The receiving device contains a selective amplifier 7 with a central frequency equal to the carrier frequency of the signal, the first multiplier 10 connected to it through the first input and a series connection between the outlet of the selective input amplifier 7 and a second input of the first multiplier 10, a delay line 8 at a time chip duration τ _k, a second multiplier (CME rer) 9 harmonic oscillation W (t) = sin (( 2πf in + 2C / τ s) t + φ ₀₎ and a bandpass filter 11 with a center frequency equal to the sum of the carrier wave frequency f ₀ and the intermediate frequency f _in. The output of the first multiplier 10 is connected to the input of the band-pass filter 12, tuned to the intermediate frequency f _in . The rest of the circuit, connected to the output of this floor filter, is a coherent receiver of signals with amplitude manipulation with a known frequency and constant phase. The output of the bandpass filter 12 is connected to the mixer multiplier 28 by the oscillation z (t) = sin (2πf _in t + C-2πf ₀ τ _s + φ ₀ ). The output of the multiplier is connected to the input of the low-pass filter 24 with a band equal to the reciprocal of the signal duration, and the signal from the output of the filter 24 goes to a decision circuit 27, which compares its level with a threshold and, when the threshold is exceeded, decides that one was transmitted, otherwise it is believed that zero was transmitted.

(mod 2), где суммирование производится по модулю два; x_j-j-й символ исходной кодовой комбинации;

- символы результирующего кода. Получаемая в результате кодовая последовательность управляет фазовым модулятором 4, где код преобразуется в фазоманипулированный сигнал в соответствии с правилом:

где
σ(t) - прямоугольный импульс единичной амплитуды с длительностью τ_s; A - амплитуда формируемого сигнала. Далее с помощью управляемого фазовращателя 5 к фазе каждого элементарного сигнала добавляется дополнительный фазовый сдвиг, увеличивающийся от сигнала к сигналу по квадратичному закону φ_j = Cj². В результате получаем сигнал:

Данный сигнал можно приближенно описать в более удобном для дальнейших преобразований виде. Поскольку j равна частному от деления нацело t на τ_s, то

и можно предложить более удобное описание сигнала s(t):

где

,
S^m(t) - отгибающая псевдошумового фазоманипулированного сигнала, соответствующая коду с номером m. Данный сигнал подается на выходные каскады передатчика и передается по каналу. На вход приемника сигнал приходит по нескольким лучам с разными задержками:

,
где
K - число лучей; h_k - амплитуда k-го луча или коэффициент передачи канала по k-му лучу; t_k - задержка k-го луча; φ_k (t) - изменяющаяся во времени по произвольному закону фаза k-го луча; n(t) - аддитивный белый шум.SPI-1 works as follows. Data from block 1 of digital messages is sent to encoder 2 in blocks of l characters. Each such block of l binary units is encoded in the form of one of M = 2 ¹ block binary code sequences. Next, the code is encoded in accordance with the principle of relative modulation, in which the output binary symbol changes its value when it enters the transcoding device 1 and retains its value when it arrives at 0 in accordance with the rule:

(mod 2), where the summation is done modulo two; x _{j is the jth} character of the source code combination;

- characters of the resulting code. The resulting code sequence controls the phase modulator 4, where the code is converted into a phase-shifted signal in accordance with the rule:

Where
σ (t) is a rectangular pulse of unit amplitude with a duration of τ _s ; A is the amplitude of the generated signal. Then, using a controlled phase shifter 5, an additional phase shift is added to the phase of each elementary signal, increasing from signal to signal according to the quadratic law φ _j = Cj ² . As a result, we get the signal:

This signal can be approximately described in a more convenient form for further transformations. Since j is equal to the quotient of dividing t completely by τ _s , then

and we can offer a more convenient description of the signal s (t):

Where

,
S ^m (t) is the bending pseudo-noise phase-manipulated signal corresponding to the code with the number m. This signal is fed to the output stages of the transmitter and transmitted over the channel. The signal arrives at the input of the receiver along several beams with different delays:

,
Where
K is the number of rays; h _k is the amplitude of the kth beam or the transmission coefficient of the channel along the kth beam; t _k is the delay of the k-th ray; φ _k (t) is the phase of the k-th ray, which varies in time according to an arbitrary law; n (t) is the additive white noise.

После полосовой фильтрации 11 сигнал принимает форму:

После умножения полученного таким образом сигнала на исходный и фильтрации в полосовом фильтре 12 получаем:

Прервем описание операций, производимых над сигналом, и проанализируем полученное выражение.At the receiver, the received signal is amplified, filtered in the frequency band of its spectrum, and subjected to a correlation transformation, which is further discussed in more detail. The received oscillation is delayed by the time duration of the elementary signal τ _s and is multiplied in mixer 9 by the signal W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ). A signal is generated at the output of the multiplier (noise conversions are omitted):

After bandpass filtering 11, the signal takes the form:

After multiplying the signal thus obtained by the original and filtering in the band-pass filter 12, we obtain:

We interrupt the description of operations performed on the signal and analyze the resulting expression.

Suppose that the phase for each ray in the interval of the duration of the elementary signal τ _s can be considered constant and the following relation holds: φ _k (t-τ _k ) ≈ φ _k (t).

The product of two noise-like signals with a time shift relative to each other can be considered as a noise-like signal S _rk (t) with parameters r and k: S _rk (t) = S ^m (tt _r ) S ^m (tt _k - τ _s ).

Преобразования используют равенство:

где
x_j - исходная кодовая последовательность сигналов, а суммирование производится по модулю два. При этом

- исходный сигнал, соответствующий исходному коду, передаваемому по каналу. В результате получаем:

где

Полученный в результате описанных преобразований сигнал состоит из суммы двух выделенных частей. Первая часть - сигнал точно на промежуточной частоте без фазовых флуктуаций колебаний отдельных лучей. Вторая часть y₂(t) представляет собой сумму шумоподобных сигналов на несущей частоте, не совпадающей с промежуточной частотой f_in. Выражение для второй части y₂(t) можно преобразовать к следующему виду:

При этом использовалось упрощенное выражение для изменения фаз отдельных лучей:
φ_k(t-τ_s) ≈ φ_k(0)+ω_dk(t-τ_s);
φ_r(t) ≈ φ_r(0)+ω_drt,
где
ω_dk,ω_dr - доплеровские сдвиги частот сигналов разных лучей.Opening the expression for the signal:

Transformations use equality:

Where
x _j is the source code sequence of the signals, and the summation is done modulo two. Wherein

- the source signal corresponding to the source code transmitted over the channel. As a result, we get:

Where

The signal obtained as a result of the described transformations consists of the sum of two selected parts. The first part is a signal exactly at an intermediate frequency without phase fluctuations in the oscillations of individual rays. The second part y ₂ (t) is the sum of noise-like signals at the carrier frequency that does not coincide with the intermediate frequency f _in . The expression for the second part y ₂ (t) can be converted to the following form:

In this case, a simplified expression was used to change the phases of individual rays:
φ _k (t-τ _s ) ≈ φ _k (0) + ω _dk (t-τ _s );
φ _r (t) ≈ φ _r (0) + ω _dr t,
Where
ω _dk , ω _dr - Doppler frequency shifts of signals of different rays.

 Note that such a representation is used only in this part of the calculations, where the properties of interference of signal origin are essentially substantiated, and therefore it is not fundamental.

где
ΔF - ширина спектра применяемого шумоподобного сигнала; T = Jτ_s - длительность сигнала. Рекомендуемое значение C = 2π/J. Шумоподобные сигналы из второй части полученного выражения

имеют частоту несущего колебания, более чем на l/T, отличающуюся от несущей частоты сигнала

, равной промежуточной частоте f_in. Таким образом несущая частота шумоподобных сигналов из второй части выражения отличается от несущей частоты сигнала первой части больше, чем на величину разрешающей способности по частоте. При этом в результате согласованной фильтрации сигнала

шумоподобные сигналы

будут восприниматься только через боковые лепестки взаимной функции корреляции и могут рассматриваться как шумовая помеха.Parameter C is selected from the condition:

Where
ΔF is the width of the spectrum of the applied noise-like signal; T = Jτ _s is the signal duration. The recommended value is C = 2π / J. Noise-like signals from the second part of the resulting expression

have a carrier frequency of more than l / T different from the carrier frequency of the signal

equal to the intermediate frequency f _in . Thus, the carrier frequency of noise-like signals from the second part of the expression differs from the carrier frequency of the signal of the first part by more than the value of the frequency resolution. In this case, as a result of consistent filtering of the signal

noise-like signals

will be perceived only through the side lobes of the mutual correlation function and can be considered as noise interference.

 We will consider the first part of the obtained expression to be a useful signal, and the second part - interference of signal origin.

. При этом фазовые флуктуации исходного сигнала могут быть разными для колебаний разных лучей и достаточно быстро меняющимися. Единственным условием, которому они должны удовлетворять, является:
φ_k(t-τ_s) ≈ φ_k(t).
Поскольку фазы колебаний, приходящих по разным лучам, после проведенных преобразований приведены к известному значению, представляется возможным при дальнейших преобразованиях осуществить когерентное накопление этих колебаний и последующее их когерентное сложение перед принятием решения о номере передаваемой кодовой комбинации. Вторую часть полученного выражения, как уже отмечалось выше, можно считать шумом. В соответствии с этим запишем:

Итак, после проведенных преобразований задача свелась к когерентному выделению сигнала из шума в многолучевом канале с известными фазами и частотами колебаний, приходящих по отдельным лучам с разными задержками.Note that, as a result of transformations, a signal that has arbitrary phase fluctuations of vibrations arriving along individual beams is reduced to a form in which these vibrations have the same intermediate frequency and a known constant phase

. In this case, phase fluctuations of the initial signal can be different for oscillations of different rays and can change quite quickly. The only condition that they must satisfy is:
φ _k (t-τ _s ) ≈ φ _k (t).
Since the phases of vibrations arriving at different beams, after the transformations are carried out, are brought to a known value, it is possible with further transformations to coherently accumulate these vibrations and their subsequent coherent addition before deciding on the number of the transmitted code combination. The second part of the expression obtained, as noted above, can be considered noise. In accordance with this, we write:

So, after the transformations, the problem was reduced to the coherent extraction of a signal from noise in a multipath channel with known phases and frequencies of vibrations arriving along individual beams with different delays.

.We continue the description of further transformations of the signal, which are carried out in accordance with the well-known scheme for receiving noise-like signals with a large base in the multipath channel without Doppler frequency shift and without phase fluctuations of the signals of individual radial paths [1, 2]. The signal is fed to M channels matched with M different ensemble signals. In each channel, the signal enters the delay line 14 with taps at equal intervals equal to

.

С выхода смесителя 15 сигнал подается на умножитель 19, реализуемый в виде усилителя с управляемым коэффициентом усиления, где он умножается на значение H(k), пропорциональное квадрату коэффициента передачи канала h $\genfrac{}{}{0ex}{}{\text{2}}{\text{k/}}$ с задержкой t_k. Эта операция подчеркивает значимость каналов приема, отличающихся задержками, с большим уровнем сигнала. Если значение h $\genfrac{}{}{0ex}{}{\text{2}}{\text{k}}$ мало, то полагают H(k) = 0, и данный канал исключается из процесса обработки сигнала. Если значения H(k) заранее не известны, они определяются в процессе приема реального сигнала, как величины, пропорциональные среднему уровню напряжений сигналов, согласованных с задержкой t_k.The signal from each tap of this delay line is fed to a multiplier-mixer 15, where it is multiplied by a reference (expected) signal S (k, t) at an intermediate frequency f _in with a delay t _k and with an initial phase

From the output of the mixer 15, the signal is supplied to the multiplier 19, implemented as an amplifier with a controlled gain, where it is multiplied by a value H (k) proportional to the square of the channel transmission coefficient h $\genfrac{}{}{0ex}{}{\text{2}}{\text{k /}}$ with a delay t _k . This operation emphasizes the importance of receiving channels, characterized by delays, with a large signal level. If the value of h $\genfrac{}{}{0ex}{}{\text{2}}{\text{k}}$ is small, then it is assumed that H (k) = 0, and this channel is excluded from the signal processing. If the values of H (k) are not known in advance, they are determined in the process of receiving a real signal, as values proportional to the average level of signal voltages matched with a delay t _k .

 The signals from the outputs of the multipliers 19, 20, 21, 22 are summed up, accumulated in the low-pass filters (integrators with reset) 24 and fed to the decision circuit 27, which compares the signal levels of different channels and outputs a code combination corresponding to the channel with the maximum level .

 As a result of the proposed transformations, it is possible to bring all the signals arriving at the receiving point by different beams with different delays and phases, which change over time according to an arbitrary law, to one phase and realize their coherent addition before deciding on the number of the transmitted signal. In this case, the goal is achieved, namely, the maximum increase in the reliability and reliability of signal reception in the multipath channel with fast phase fluctuations and strong variable Doppler frequency shifts.

 SPI-2 works as follows.

где
σ(t) - прямоугольный импульс единичной амплитуды с длительностью τ_s, A - амплитуда сигнала. В результате при поступлении на вход модулятора дискретного сообщения m = 0, 1 передатчик 6 передает по каналу сигнал вида:

В приемнике сигнал усиливается и фильтруется в полосе частот своего спектра. Далее принимаемое колебание задерживается на время длительности элементарного сигнала τ_s и умножается в смесителе 9 на колебание W(t) = sin((2πf_in+2C/τ_s)t+φ₀). На выходе перемножителя образуется сигнал (преобразования шума опускаются):

После полосового фильтра 11 сигнал принимает следующую форму:

После умножения полученного таким образом сигнала на исходный и фильтрации в полосовом фильтре 12 получаем:

Проведем анализ полученного выражения. Его удобно представить в виде:

где

При этом в выражении для второй части формулы использовалось представление меняющейся во времени фазы через доплеровские частоты
ω_dk,ω_dr.
φ_k(t-τ_s) ≈ φ_k(0)+ω_dk(t-τ_s);
ω_r(t) ≈ φ_r(0)+ω_drt.
Параметр C и длительность сигнала T = m τ_s выбираются из двух условий:

При выполнении этого условия сигналы из второй части полученного выражения имеют частоту несущего колебания, более чем на l/T отличающуюся от несущей частоты сигнала

, равной промежуточной частоте f_in, и поэтому не проникают на выход приемного усилительного тракта после фильтрации в выходном фильтре низких частот. Рекомендуемое значение C = 2 π /J.Binary discrete messages from source 1 are sent to modulator 4, where when 0 appears, the signal is not transmitted, and when 1 arrives, a signal is transmitted, which after a controlled phase shifter has the form:

Where
σ (t) is a rectangular pulse of unit amplitude with duration τ _s , A is the signal amplitude. As a result, when a discrete message m = 0, 1 is received at the input of the modulator, the transmitter 6 transmits a signal of the form:

At the receiver, the signal is amplified and filtered in the frequency band of its spectrum. Next, the received oscillation is delayed by the time duration of the elementary signal τ _s and is multiplied in mixer 9 by the oscillation W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ). A signal is generated at the output of the multiplier (noise conversions are omitted):

After the band-pass filter 11, the signal takes the following form:

After multiplying the signal thus obtained by the original and filtering in the band-pass filter 12, we obtain:

Let us analyze the resulting expression. It is convenient to present it in the form:

Where

Moreover, in the expression for the second part of the formula, we used the representation of the time-varying phase through Doppler frequencies
ω _dk , ω _dr .
φ _k (t-τ _s ) ≈ φ _k (0) + ω _dk (t-τ _s );
ω _r (t) ≈ φ _r (0) + ω _dr t.
The parameter C and the signal duration T = m τ _s are selected from two conditions:

When this condition is fulfilled, the signals from the second part of the resulting expression have a carrier oscillation frequency that is more than l / T different from the carrier frequency of the signal

equal to the intermediate frequency f _in , and therefore do not penetrate the output of the receiving amplifier path after filtering in the output low-pass filter. Recommended value C = 2 π / J.

где
Δ t_m - время многолучевого растяжения сигналов.

Where
Δ t _m is the time of multipath stretching of the signals.

, то есть фаза не зависит от k и все слагаемые вида

из первой части выражения оказываются приблизительно с одинаковыми фазами. Интерференция, вызванная многолучевым распространением сигналов, пропадает, так как происходит когерентное фазирование сигналов, приходящих по разным лучам с разными задержками и с произвольными, меняющимися во времени, фазами.When this condition is true,

, i.e., the phase does not depend on k and all terms of the form

from the first part of the expression are approximately the same phases. The interference caused by the multipath propagation of signals disappears, because there is a coherent phasing of signals arriving on different beams with different delays and with arbitrary, time-varying phases.

При передаче нулевого символа (m=0) по каналу просто ничего не передается. В результате на выходе фильтра 28 нижних частот приемника образуется случайное напряжение, вызванное прохождением по цепям приемника шума. Среднее значение этого напряжения равно нулю, а среднеквадратическое значение определяется в соответствии с формулой:

Напряжение с выхода фильтра нижних частот подается на схему 27 принятия решения, которая сравнивает его уровень с порогом и при превышении порога принимает решение, что передавалась единица, в противном случае считается, что передавался ноль. Уровень порога (в соответствии с критерием Неймана-Пирсона) должен в несколько раз превышать среднеквадратическое значение напряжения, образующегося в результате прохождения шума по цепям приемного устройства при передаче 0. Сравнение формул для уровня полезного сигнала и среднеквадратического значения шума показывает, что увеличение длительности сигнала позволяет неограниченно увеличивать отношение сигнал-шум перед схемой принятия решения, несмотря на произвольный закон изменения фазы сигналов, приходящих по отдельным лучам. Этот эффект служит достижению поставленной цели, а именно максимальному повышению надежности и достоверности приема сигналов в многолучевом канале с быстрыми фазовыми флуктуациями и сильными переменными доплеровскими смещениями частот.After the band-pass filter 12 with a central frequency f _{in, the} signal is multiplied in the multiplier-mixer 28 by a harmonic oscillation of the form:
Z (t) = sin (2πf _in t + C-2πf ₀ τ _s + φ ₀ ),
and passes through a low-pass filter (integrator with reset) 24 with a bandwidth of l / T. The output voltage of the filter is:

When transmitting a null character (m = 0) over a channel, nothing is simply transmitted. As a result, a random voltage is generated at the output of the low-pass filter 28 of the receiver, caused by passing noise through the receiver circuits. The average value of this voltage is zero, and the rms value is determined in accordance with the formula:

The voltage from the output of the low-pass filter is supplied to a decision circuit 27, which compares its level with a threshold and, when the threshold is exceeded, makes a decision that one was transmitted, otherwise it is considered that zero was transmitted. The threshold level (in accordance with the Neumann-Pearson criterion) should be several times higher than the rms value of the voltage generated as a result of the noise passing through the receiver circuits during transmission 0. A comparison of the formulas for the level of the useful signal and the rms value of the noise shows that an increase in the signal duration allows unlimitedly increase the signal-to-noise ratio in front of the decision-making scheme, despite the arbitrary law of the phase change of the signals arriving on individual beams. This effect serves to achieve the goal, namely, to maximize the reliability and reliability of signal reception in the multipath channel with fast phase fluctuations and strong variable Doppler frequency shifts.

 The proposed method is implemented through the described operation of communication devices SPI-1 and SPI-2. Transformation of the transmitted code sequence in accordance with the law of relative phase modulation is implemented in the transcoding device 3. Signal generation based on phase modulation is implemented in modulator 4.

, где f_in - промежуточная частота, реализуется с помощью смесителя 9 и полосового фильтра, настроенного на частоту f₀+f_in. Умножение сигнала, полученного в результате таких преобразований, на исходный сигнал реализуется в перемножителе 10. Выделение необходимой части спектра результата перемножения реализуется в полосовом фильтре 12. Известная структура когерентного приемника в случае СПИ-1 реализуется в виде многоканальной схемы с многоотводными линиями 14 задержки, перемножителями 15 - 22, сумматорами 23 и фильтрами 24 нижних частот, подключенной к схеме принятия решения. В случае СПИ-2 когерентный приемник дискретных сигналов с амплитудной модуляцией реализуется в виде смесителя 28, подключенного к фильтру нижних частот, выход которого подключен к схеме принятия решения. Случай СПИ-2 предельно простой и может рассматриваться как вырожденный. В нем нельзя выделить кодер или устройство перекодирования в соответствии с законом фазоразностной модуляции. В этом случае, при передаче по каналу 1, сигнал формируется из J элементарных сигналов с одной и той же фазой, что соответствует передаче кодовой комбинации из одних нулей. Можно считать, что кодер превышает 1 в кодовую комбинацию, состоящую из одних нулей. Такая кодовая комбинация после перекодирования по закону фазоразностной модуляции не изменяется. После дополнительной фазовой модуляции каждого элементарного сигнала j в соответствии с квадратическим законом φ_j = Cj² с помощью управляемого фазовращателя (управляемого элемента задержки) сигнал становится действительно сложным и широкополосным у него появляется хорошее разрешение по задержкам, которое используется при его приеме. При передаче по каналу 0 просто ничего не передается.Additional phase modulation of each elementary signal j in accordance with the quadratic law φ _j = Cj ² is implemented using a controlled phase shifter or a controlled delay element 5. Pre-amplification and filtering are implemented in a selective amplifier 7. The delay of the received signal for the duration of the duration of the elementary signal τ _s is realized using the delay element 8. Frequency offset by

, where f _in is the intermediate frequency, it is realized using a mixer 9 and a band-pass filter tuned to the frequency f ₀ + f _in . Multiplication of the signal obtained as a result of such transformations by the original signal is implemented in the multiplier 10. The allocation of the necessary part of the spectrum of the multiplication result is implemented in the band-pass filter 12. The well-known coherent receiver structure in the case of SPI-1 is implemented as a multi-channel circuit with multi-tap delay lines 14, multipliers 15 to 22, by adders 23 and low-pass filters 24, connected to a decision circuit. In the case of SPI-2, a coherent receiver of discrete signals with amplitude modulation is implemented as a mixer 28 connected to a low-pass filter, the output of which is connected to a decision-making circuit. The case of SPI-2 is extremely simple and can be regarded as degenerate. It is impossible to distinguish an encoder or transcoding device in accordance with the law of phase difference modulation. In this case, when transmitting on channel 1, the signal is formed from J elementary signals with the same phase, which corresponds to the transmission of a code combination of one zeros. We can assume that the encoder exceeds 1 in a code combination consisting of only zeros. Such a code combination after recoding according to the law of phase difference modulation does not change. After additional phase modulation of each elementary signal j in accordance with the quadratic law φ _j = Cj ² using a controlled phase shifter (controlled delay element), the signal becomes really complex and broadband, it has a good delay resolution, which is used when receiving it. When transmitting on channel 0, nothing is simply transmitted.

 Bibliography.

 1. Fink L.M. Theory of discrete message transmission. - M .: Soviet Radio, 1970, p. 728.

 2. Penin P.I.

 3. Channel auto-select device for multi-reception. USSR author's certificate N 886273, cl. H 04 B 7/02, 1980.

 4. A method of compensating for interference and a device for implementing this method. U.S. Patent 4,085,368, cl. H 04 B 7/08, Bell Telephone Laboratory, 80.08.76.

 5. A device for isolating phase-modulated signals against interference. USSR copyright certificate N 613506, cl. H 04 B 1/10, 1976.

 6. Device for adding diversity signals. USSR copyright certificate N 620025, cl. H 04 B 7/04, 1975.

 7. Device for distributing signals of several channels. copyright certificate 1409101, UK, cl. H 04 B 1/00, Western Electric Co. 12/27/71.

Claims (3)

1. The method of transmitting information over a multipath channel, which consists in the fact that the signal carrying information is composed of J elementary signals of the same duration and different in phase, transmit it over the communication line and receive using a multi-channel filter, agreed in each channel with one of the transmitted signals, characterized in that the transmitted code sequence with characters X _j from an alphabet of size L is transformed into a sequence

in accordance with the rule

where the summation is performed modulo L, and the phase of the carrier of each elementary signal with the number

form in the form

where C is a constant, the coefficient of the quadratic law of change in the initial phase of the carrier,
and when receiving the signal is delayed by the duration of the elementary signal τ _s , its frequency is shifted by the value

multiply by the initial received oscillation and process the signal after separation at the intermediate frequency f _in . 2. A system for transmitting discrete information, consisting of a transmitting device including a digital message source, an encoder, a phase modulator, a transmitter, and a signal receiving device with a known frequency and constant phase based on a transverse filter, including a selective input amplifier and circuits of M parallel channels connected to the decision-making circuit, each channel contains a delay line with K uniform taps, which through the multipliers to the reference signal corresponding to this channel y, and multipliers by values proportional to the squares of the transmission coefficients of individual signal propagation paths having corresponding delays are connected to the adder, and the outputs of the adders of individual channels through low-pass filters are connected to a decision circuit that compares the signal levels of different channels and outputs the number signal corresponding to the channel with the maximum level, characterized in that in the transmitting device between the encoder and the phase modulator the transcoding device for The principle of phase difference modulation, in which the output binary symbol changes its value when it enters the transcoding device 1 and retains its value when it arrives at 0, a controlled phase shifter is switched on between the phase modulator and the transmitter, by which a phase increases to the carrier phase of each elementary signal, increasing depending on from the sequence number j of an elementary signal with a duration of τ _s , according to the quadratic law
φ _j = C $\genfrac{}{}{0ex}{}{\text{2}}{\text{j}}$ ,
where C is a constant, the coefficient of the quadratic law of change in the initial phase of the carrier,
and in the receiver, after the input selective amplifier, a circuit is introduced consisting of the first multiplier connected by its first input to the output of the selective amplifier, connected in series between the output of this amplifier and the second input of the first delay line multiplier for the duration of the elementary signal, the second harmonic oscillation multiplier
W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ),
where t is the current time,
and a band-pass filter with a central frequency equal to the sum of the frequencies of the carrier wave and the intermediate frequency f _in , the output of the first multiplier is connected to the input of the band-pass filter tuned to the intermediate frequency, and the output of this filter is connected to the input of the receiver circuit from M channels. 3. A discrete information transmission system consisting of a transmitting device including a digital message source, a modulator and a transmitter, and a signal receiving device with a known frequency and constant phase, including a selective input amplifier, a multiplier-mixer for the carrier wave of the received signal connected to its output of a low-pass filter with a band equal to the reciprocal of the signal duration and connected to the output of this filter decision circuit, characterized in that in The transmitting device between the modulator and the transmitter includes a controlled phase shifter, with which the carrier phase is changed quadratically according to equal time intervals of duration τ _s
φ _j = C $\genfrac{}{}{0ex}{}{\text{2}}{\text{j}}$ ,
where C is a constant, the coefficient of the quadratic law of change of the initial phase of the carrier;
j is the number of the interval,
and in the receiver after the input selective amplifier, a circuit is introduced consisting of the first multiplier connected by its first input to the output of the input selective amplifier, connected in series between the output of this amplifier and the second input of the multiplier, a delay line for the duration of the elementary signal, the second harmonic oscillation multiplier
W (t) = sin ((2πf _in + 2C / τ _s ) t + φ ₀ ),
where t is the current time,
and a band-pass filter with a central frequency equal to the sum of the frequencies of the carrier wave and the intermediate frequency f _in , the output of the first multiplier is connected to the input of the band-pass filter tuned to the intermediate frequency, and the output of this filter is connected to the receiver multiplier-mixer.

Images