RU2610285C1 - Method of detecting low-rate encoding protocols - Google Patents

Abstract

FIELD: physics, acoustics.

SUBSTANCE: invention is intended for detecting low-rate speech encoding (LRSE) protocols . The technical result is achieved by increasing the number of dimensions of a measured vector of redundancy factors _Z to _L, L=Z+2 and accounting for the shifting effect of vector elements _L by forming a quadratic reference matrix _{Lj et} for all J known LRSE protocols, j=1, 2 , J. To this end, a digital stream Y is received during a given time interval T. A rectangular information matrix Y_KL is formed, rows of which are information units arranged in series one below the other. The vector of redundancy factors _L is calculated, followed by element by element comparison of the measured vector _L with rows _{Lj et}(l) of all J quadratic reference matrices _{Lj et}, determining deviation P_j(l) between the measured vector _L and rows of all J reference matrices _{Lj et}, making a decision in favour of the j-th LRSE protocol, for which is ensured minimum deviation P_j(l)_min of the measured vector _L from the l-th row of the j-th quadratic reference matrix _{Lj et}.

EFFECT: method of improving accuracy of detecting LRSE protocols.

3 cl, 9 dwg

Description

The invention relates to the field of communication technology, in particular to the processing of digital information, and can be used in a wide class of data transmission systems for recognizing the low-speed speech coding (NSCR) protocols used by vocoders.

The claimed technical solution expands the arsenal of tools for a similar purpose due to the possibility of recognizing the NSCR protocols with higher accuracy.

State of the art

A known method for the recognition of radio signals (see RF Patent 2551903 / Method for the recognition of radio signals / Dvornikov A.S., Dvornikov S.V. /, published June 10, 2015), in which, at the first stage, matrices are formed from sampled and quantized samples of reference radio signals energy distribution based on their frame wavelet transforms. Then from them, starting from the second line, feature vectors are formed by row-wise concatenation of all wavelet coefficients. Next, the elements of the feature vectors are normalized and their parameters are calculated. Moreover, the average value of the normalized amplitude values of the elements of the feature vectors is determined as parameters, and the decision is made based on the results of calculating the differences in the values of the parameters of the recognized radio signal and the reference radio signals. The recognized radio signal is considered to be incident with the reference radio signal, the module of the difference in the parameters of the feature vectors with which will be minimal.

The disadvantage of this method is the lack of accuracy of recognition of the NSCR protocols.

, z=1, 2, …, Z; Z=N_б-2, содержащих по три соседних столбца прямоугольной информационной матрицы Y_K×L, где текущий номер z прямоугольной матрицы

определяется номером столбца матрицы Y_K×L, являющийся первым столбцом матрицы

, по каждой из матриц

вычисляют коэффициент избыточности ϕ_z, на основе полученного набора коэффициентов избыточности формируют вектор измеренных коэффициентов избыточности ϕ_Z путем последовательного их размещения в соответствии с порядковыми номерами z, формируют эталонные вектора коэффициентов избыточности ϕ_{Zj эт} известных протоколов НСКР, вектор измеренных коэффициентов избыточности ϕ_Z сравнивают с эталонными векторами коэффициентов избыточности ϕ_{Zj эт}, j=1, 2, … J, известных протоколов НСКР.Closest to the claimed one is a method (prototype) for decoding a cyclic error-correcting code (see Timofeev D.I., Tavalinsky D.A., Chubaty D.N. Analysis of parameters of low-speed speech encoders under structural and parametric uncertainty // High-tech technologies, No. 8, 2011. P. 4-9.), Consisting in the fact that the digital information stream Y is received during the time interval ΔT, based on Y, a normalized autocorrelation function A is formed, according to regular extrema of the autocorrelation functions at regular intervals Δt and A decide about the presence of the block structure in the data stream Y, in intervals between the extrema of the autocorrelation function of A divided information Y flux on information blocks in volume N _b bits each, sequentially assigned information units sequence numbers k = 1, 2, ..., K, from the first information unit is formed rectangular information matrix Y _{k × L, L} = N _b whose rows are sequentially arranged one below the other information blocks in accordance with their sequence numbers k = 1, 2, ..., k, forms ruyut sequence of rectangular matrices

, z = 1, 2, ..., Z; Z = N _b -2, containing three adjacent columns of a rectangular information matrix Y _{K × L} , where the current number z of the rectangular matrix

determined by the column number of the matrix Y _{K × L} , which is the first column of the matrix

, for each of the matrices

calculate the redundancy coefficient ϕ _z , on the basis of the obtained set of redundancy coefficients form a vector of the measured redundancy coefficients ϕ _Z by sequentially placing them in accordance with serial numbers z, form the standard vectors of the redundancy coefficients ϕ _{Zj et} known NSCR protocols, compare the vector of the measured redundancy coefficients ϕ _Z with reference vectors of redundancy coefficients ϕ _{Zj et} , j = 1, 2, ... J, of the well-known NSCR protocols.

This method potentially has a sufficiently high recognition accuracy of the NSCR protocols.

However, the disadvantage of the prototype is the relatively low recognition accuracy of the NSCR protocols in the absence of N _p characters in the first block of the digital information stream Y. The probability of the absence of characters in the first block of the digital information stream Y is determined by an expression of the form:

.

.

, что вызывает сдвиг значений вектора коэффициентов избыточности ϕ_Z на величину

. В этом случае в векторе ϕ_Z будут отсутствовать значения коэффициента избыточности ϕ_z с номерами z=N_п, N_п+1, а также добавляются значения коэффициента избыточности ϕ_z с номерами z=N_б-1, N_б. Значения N_б цифровых потоков, формируемых вокодерами, которые разработаны для радиостанций диапазона ВЧ, могут составлять 45, 48, 54, 60, 64, 66, 81 или 96 бит. Таким образом, высока вероятность искажения в векторе коэффициентов избыточности ϕ_Z, что приводит к ошибочным решениям в способе-прототипе. Кроме того, точностные характеристики прототипа снижаются из-за некорректного сокращения длины измеренного и эталонного векторов ϕ_Z и ϕ_{Zj эт} соответственно на два элемента. Последнее приводит к неполному использованию информации о структуре принятого сигнала и, как следствие, к снижению точности распознавания протоколов НСКР.When the condition 0≤N _p ≤N _{b is} satisfied, in the information matrix Y _{K × L} there is a shift of the columns to the left by the amount

, which causes a shift in the values of the vector of redundancy coefficients ϕ _Z by

. In this case, the redundancy coefficient ϕ _z with numbers z = N _p , N _p +1 will be absent in the vector ϕ _Z , and the redundancy coefficient ϕ _z with numbers z = N _b -1, N _b will be added. The values of N _b digital streams generated by vocoders, which are designed for radio stations in the high frequency range, can be 45, 48, 54, 60, 64, 66, 81 or 96 bits. Thus, the probability of distortion in the vector of redundancy coefficients ϕ _Z is high, which leads to erroneous decisions in the prototype method. In addition, the accuracy characteristics of the prototype are reduced due to incorrect reduction in the length of the measured and reference vectors ϕ _Z and ϕ _{Zj et} respectively by two elements. The latter leads to the incomplete use of information about the structure of the received signal and, as a result, to a decrease in the recognition accuracy of the NSCR protocols.

The aim of the invention is to develop a method for recognizing the NSCR protocols, providing increased accuracy of recognition of the NSCR protocol.

, z=1, 2, …, Z; Z=N_б-2, содержащих по три соседних столбца прямоугольной информационной матрицы Y_K×L, где текущий номер z прямоугольной матрицы

определяется номером столбца прямоугольной информационной матрицы Y_K×L, являющийся первым столбцом прямоугольной матрицы

, вычисление по каждой из прямоугольных матриц

коэффициента избыточности ϕ_Z, формирование измеренного вектора коэффициентов избыточности ϕ_z на основе полученного набора значений коэффициента избыточности путем последовательного их размещения в соответствии с порядковыми номерами z, формирование эталонных векторов коэффициентов избыточности ϕ_{Zj эт} известных протоколов НСКР, сравнение измеренного вектора коэффициентов избыточности ϕ_Z с эталонными векторами коэффициентов избыточности ϕ_{Zj эт}, j=1, 2, …, J, известных протоколов НСКР. Увеличивают до значения L размерность эталонных векторов коэффициентов избыточности ϕ_L. Увеличивают до значения L размерность измеренного вектора коэффициентов избыточности ϕ_{Lj эт} для известных протоколов НСКР. Для каждого j-го известного протокола НСКР на основе соответствующего эталонного вектора ϕ_{Lj эт} формируют квадратную эталонную матрицу Φ_{Lj эт}. Поэлементно сравнивают измеренный вектор ϕ_L со строками всех квадратных эталонных матриц Φ_{Lj эт}. Принимают решение в пользу j-го протокола НСКР, для которого обеспечивается минимальное и меньше заданного порога R отклонение

,

, измеренного вектора ϕ_L от одной из строк j-й квадратной эталонной матрицы Φ_{Lj эт}.This goal is achieved by the fact that in the known method for recognizing NSCR protocols, which includes receiving a binary digital information stream Y over a time interval ΔT, generating, based on Y, a normalized autocorrelation function A, deciding whether there is a block structure in the information stream Y at regular intervals with equal intervals Δτ extrema autocorrelation function a, the division of the information stream on the information blocks Y volume N _b bits each by the autocorrelation interval between extrema Fung tion A, the assignment of successive information blocks of sequence numbers k = 1, 2, ..., K, from the first information unit, forming rectangular information matrix Y _{K × L, L} = N _b; the lines of which are successively placed under each other information blocks in accordance with their serial numbers k = 1, 2, ..., K, the formation of a sequence of rectangular matrices

, z = 1, 2, ..., Z; Z = N _b -2, containing three adjacent columns of a rectangular information matrix Y _{K × L} , where the current number z of the rectangular matrix

determined by the column number of the rectangular information matrix Y _{K × L} , which is the first column of the rectangular matrix

calculating for each of the rectangular matrices

the redundancy coefficient ϕ _Z , the formation of the measured vector of the redundancy coefficients ϕ _z based on the obtained set of values of the redundancy coefficient by sequentially placing them in accordance with serial numbers z, the formation of the reference vectors of the redundancy coefficients ϕ _{Zj et} known protocols NSCR, the comparison of the measured vector of the redundancy coefficients ϕ _Z with the reference vectors of the redundancy coefficients ϕ _{Zj et} , j = 1, 2, ..., J, of the well-known NSCR protocols. The dimension of the reference vectors of redundancy coefficients ϕ _{L is} increased to _L. The dimension of the measured vector of redundancy coefficients ϕ _{Lj et is} increased to the value L for known NSCR protocols. For each j-th known NSCR protocol, on the basis of the corresponding reference vector ϕ _{Lj et} form a square reference matrix Φ _{Lj et} . The measured vector ϕ _{L is} compared elementwise with the rows of all square reference matrices Φ _{Lj et} . Decide in favor of the j-th NSCR protocol, for which a minimum and less than a given threshold R deviation is provided

,

of the measured vector ϕ _L from one of the rows of the jth square reference matrix Φ _{Lj et} .

In this case, an increase in the dimension of the measured vector of redundancy coefficients to the value L is achieved by increasing the dimension of the rectangular information matrix Y _{K × M} , M = N _b +2, by adding to the original matrix Y _{K × L} its first and second columns as the penultimate and last columns, respectively .

,

содержат значения вектора ϕ_{Lj эт}, последовательно сдвинутые в каждой строке относительно первой строки влево на величину

.A square reference matrix Φ _{Lj et is} formed (see FIG. 6) using the reference vector ϕ _{Lj et} as the first row ϕ _{Li et} (0) and the rest of its rows

,

contain the values of the vector ϕ _{Lj et} , successively shifted in each row relative to the first row to the left by the amount

.

Thanks to the new set of essential features in the claimed method, an increase in the recognition accuracy of the NSCR protocols is achieved by increasing the dimension of the measured vector of redundancy coefficients ϕ _L , from Z = N _b -2 to L = N _b taking into account the shift effect of the elements of the vector of redundancy coefficients ϕ _L by forming a square the reference matrix Φ _{Lj et} for all J known NSCR protocols.

The claimed method is illustrated by drawings, which show:

in FIG. 1 - recognition algorithm protocols NSCR;

in FIG. 2 - autocorrelation function of the information flow Y;

in FIG. 3 - the order of formation of the information matrix Y _{K × L} ;

in FIG. 4 - the order of formation of the information matrix Y _{K × M} ;

;in FIG. 5 - matrix formation order

;

in FIG. 6 - the order of formation of the square reference matrix Φ _{Lj et} ;

in FIG. 7 - the results of simulation;

in FIG. 8 - the effect of the loss of symbols in the CPU on the vector of measured redundancy coefficients ϕ _Z :

a) the vector of measured redundancy coefficients ϕ _Z , formed in the absence of losses in the first block of the digital information stream Y;

b) the vector of measured redundancy coefficients ϕ _Z , formed when the first four bits (N _p = 4) are lost in the first block of the digital information stream Y;

in FIG. 9 is a generalized block diagram of an NSCR protocol recognition device.

An analysis of the use of low-speed speech encoders in communication systems indicates that despite the desire of the international telecommunication union to switch to the use of standard encoders, a new compression algorithm and the corresponding NSCR protocol are often developed, which leads to a variety of low-speed speech encoders (see Bondarenko M.F. ., Rabotyagov A.V., Shchepkovsky S.V. Modern methods of encoding a speech signal.- Kharkov: KhNURE Bionics of intelligence No. 2 (69), 2008). Under the conditions of structural and parametric uncertainty, in connection with an increase in the number of NSCR protocols, the task of recognizing them becomes more complicated, the solution of which is necessary for subsequent decoding of speech signals.

Due to the fact that different speech signal encoders form different structures of CPU blocks, the analysis of CPU parameters consists in determining the number of bits in one CPU block and identifying the structure of the CPU block based on the analysis of conditional probabilities of transitions of parameter values from block to block over the entire length of the linear prediction window .

In the high frequency range vocoders used in foreign high frequency range radio stations, the coding method (MK) with ten linear prediction coefficients (LP) LPC-10 (Linear Predictive Coding), MK with LP and mixed excitation MELP (Mixed Excitation Linear Prediction) is implemented, MK with mixed multiband excitation MMBE (Mixed-Multiband Excitation). The parameters that characterize vocoders are: CPU speed B _wok (600, 800, 1200 or 2400 bps), the number of N _b bits in one information block (length of the information block) and the duration t _{b of} its transmission. The latter is asked at the stage of developing a vocoder (see Bykov S.F., Zhuravlev V.I., Shalimov I.A. Digital telephony. Textbook for universities. - M .: Radio and communications, 2003. - 144 p.). The speed V _{wok of the} output CPU is related to the parameters N _b and t _{b by the} expression V _wok = N _b / t _b , which implies N _b = B _wok ⋅t _b (see Aladinsky V.A., Kuzminsky S.V. Analysis of digital streams at the outputs of vocoders used on foreign radio lines of the high frequency range. - M .: Successes of modern radio electronics. No. 7, 2015. P. 73-76). The duration of one voice message is determined by the formula ΔT = K⋅t _b .

A positive effect in the proposed method is achieved by increasing the dimension of the measured and reference vectors of redundancy coefficients ϕ _Z and ϕ _{Zj et,} respectively, and taking into account the shift in the rows of the information matrix Y _{K × L} (due to the absence of N _p characters in the first block of the digital information stream Y) by forming square reference matrices Φ _{Lj et} on the basis of standard vectors of redundancy coefficients of increased dimension ϕ _{Lj et} .

Implementation of the claimed method can be carried out as follows (see Fig. 1). During the preparation stage it is expedient to carry out the calculation of the coefficients of reference vectors φ _{Lj et} redundancy, but for a better understanding of this operation is discussed after the description of the order of forming the measured redundancy vector coefficients φ _L.

In the process, take a binary digital information stream Y of length N _cp bits, the number of which is determined based on the expression:

After that, the normalized autocorrelation function A is calculated, which characterizes the linear dependence of the original sequence Y on the sequence Y (τ), shifted circularly by τ = 1, 2, ..., N _cp -1 bit with respect to Y. The normalized autocorrelation function of a real speech signal, the view of which is shown in FIG. 2 has distinct local maxima located at periodic intervals Δτ, and Δτ is equal to the length of the information block. The presence of local maxima of the autocorrelation function located with a period Δτ is explained by the fact that the parameters of the speech signal are located inside the information block strictly in a certain place and when the shifts Δτ are multiples of the length of the information block, the ACF value will be the largest. The desired autocorrelation function (see Fig. 2) is considered a set of values of the form:

where a (x) = c (τ) / D (Y) is the correlation coefficient; c (τ) is the covariance coefficient; D (Y) is the dispersion of the digital information stream Y.

The result of the correlation analysis is the detection of a periodic component in the digital information stream Y. In this case, the frequency value is equal to the length of the information block of the speech signal.

By regularly at equal intervals Δτ = N _b extrema normalized autocorrelation function A decide a presence information stream block structure extrema Y. absence normalized autocorrelation function A at equal intervals Δτ indicates that the digital information stream Y does not contain a voice message, and its analysis is completed.

If there are extrema of the normalized autocorrelation function A with equal intervals Δτ, the number of information blocks contained in the digital information stream Y is calculated as follows:

At the next stage, at intervals Δτ = N _b between the extrema of the normalized autocorrelation function A, the information flow Y is divided into information blocks of N _b bits each (see Fig. 3). Sequentially assign information blocks to serial numbers k = 1, 2, ..., K, starting from the first information block. A rectangular information matrix Y _{K × L} , L = N _{b is formed} (see Fig. 3), the rows of which are information blocks sequentially placed under each other in accordance with their serial numbers k = 1, 2, ..., K.

означает «присоединение». Увеличение размерности прямоугольной информационной матрицы Y_K×L до значения Y_K×M необходимо для последующего увеличения размерности вектора измеренных коэффициентов избыточности ϕ_L.After that, a new information matrix Y _{K × M} , M = N _b +2 is formed by adding to the matrix Y _{K × L} its first and second columns as the penultimate and last columns (see Fig. 4). Here is the sign

means joining. An increase in the dimension of the rectangular information matrix Y _{K × L} to the value of Y _{K × M is} necessary for the subsequent increase in the dimension of the vector of the measured redundancy coefficients ϕ _L.

, где K - количество строк, а 3 - количество столбцов в окне информационной матрицы Y_K×M.It is known that the matrix of probabilistic transitions gives a complete probabilistic description of the redundancy of the representation of the NSCR parameter, and also qualitatively and quantitatively shows the difference between the speech signal processed by the encoder. The conditional probabilities of the transition from the ith to the jth state are calculated by the “sliding window” method, which is a rectangular matrix

, where K is the number of rows, and 3 is the number of columns in the information matrix window Y _{K × M.}

, z=1, 2, …, L; L=N_б (см. фиг. 5), содержащих по три соседних столбца информационной матрицы Y_K×M. Текущий номер z прямоугольной матрицы

определяется номером столбца матрицы Y_K×M, являющимся первым столбцом прямоугольной матрицы

по каждой из прямоугольных матриц

.Based on the information matrix Y _{K × M} , a sequence of rectangular matrices is formed

, z = 1, 2, ..., L; L = N _b (see Fig. 5), each containing three adjacent columns of the information matrix Y _{K × M.} The current z number of the rectangular matrix

determined by the column number of the matrix Y _{K × M} , which is the first column of the rectangular matrix

for each of the rectangular matrices

.

The values of the zth redundancy coefficient ϕ _{z are} calculated using the well-known expression (see Lebedev A.N., Kupriyanov M.S., Nedosekin D.D., Chernyavsky E.A. Probabilistic methods in engineering problems: Reference book. - St. Petersburg. : Energoatomizdat, St. Petersburg Branch, 2000. - 333 p.):

;

;Where

;

;

;

- текущее значение энтропии прямоугольной матрицы

; Н₀(Y_K×3) - максимальное значение энтропии матрицы размерностью K×3; s=2³ - объем алфавита 3-битных символов; p_z(y_i) - вероятность появления 3-битного символа y с i-м состоянием в прямоугольной матрице

; p_z(y_j/y_i) - условная вероятность перехода 3-битного символа y из i-го в j-е состояние в прямоугольной матрице

.

;

- the current value of the entropy of the rectangular matrix

; H ₀ (Y _{K × 3} ) - the maximum value of the entropy of the matrix dimension K × 3; s = 2 ³ - the volume of the alphabet of 3-bit characters; p _z (y _i ) is the probability of the appearance of a 3-bit symbol y with the i-th state in a rectangular matrix

; p _z (y _j / y _i ) - conditional probability of the transition of a 3-bit symbol y from the i-th to the j-th state in a rectangular matrix

.

Based on the calculated values of the redundancy coefficients ϕ _z , a vector of measured redundancy coefficients ϕ _L = (ϕ ₁ , ϕ ₂ , ..., ϕ _z , ..., ϕ _L ) is formed. It should be noted that due to the increase in the dimension of the information matrix Y _{K × M} (more complete use of information about the received signal), the dimension of the vector of the measured redundancy coefficients ϕ _L increased by two elements.

,

квадратных эталонных матриц Φ_{Lj эт} (см. фиг. 6), определяют отклонения между измеренным вектором ϕ_L и строками (эталонными векторами) всех J квадратных эталонных матриц Φ_{Lj эт} (см. Ту Дж., Гонсалес Р. Принципы распознавания образов / Пер. с англ. - М.: Мир, 1978. - 411 с.):Next, the vector ϕ _{L is} compared elementwise with the rows

,

square reference matrices Φ _{Lj et} (see Fig. 6), determine the deviations between the measured vector ϕ _L and the rows (reference vectors) of all J square reference matrices Φ _{Lj et} (see Tu J., González R. Principles of pattern recognition / Per . from English .-- M .: Mir, 1978. - 411 p.):

выбирают минимальное отклонение

для

. Решение о наиболее вероятному j-м протоколе НСКР принимают при выполнении:From the resulting set of values

choose the minimum deviation

for

. The decision on the most probable jth protocol of the NSCR is made when:

where R is the threshold value due to the number of information blocks K in the digital information stream Y.

 Otherwise:

It is believed that a new (previously unknown) NSCR protocol was applied to form the digital information stream Y.

A known method of forming standards, each of which includes a reference vector of features, is described in (see Tu J., Gonzalez R. Principles of pattern recognition / Transl. From English. - M .: Mir, 1978. - 411 pp.). In this case, the reference feature vector ϕ _{Lj et} has the form:

- z-й элемент эталонного вектора ϕ_{Lj эт} j-го класса, ϕ_z,j,w - z-й элемент w-й реализации обучающей выборки j-го класса, w=1, 2, …, W - порядковый номер реализации обучающей выборки; W - объем обучающей выборки, т.е. Where

is the zth element of the reference vector ϕ _{Lj at the} jth class, ϕ _{z, j, w} is the zth element of the wth implementation of the training sample of the jth class, w = 1, 2, ..., W is the serial number of the implementation training sample; W is the volume of the training sample, i.e.

, ϕ_Lj,w - w-я реализации обучающей выборка j-го класса.

, ϕ _{Lj, w} is the wth implementation of the training sample of the jth class.

, принадлежащей к заведомо известному j-му классу протоколов НСКР. Достаточным считают объем j-й обучающей выборки W≥200, (см. Математический энциклопедический словарь. / Гл. ред. Прохоров Ю.В. - М.: Большая российская энциклопедия, 1995. Репринтное издание - М.: Сов. энциклопедия, 1988. - 847 с.).Each wth implementation of the jth class ϕ _{Lj, w is} formed by computing the vector ϕ _Lj based on the information matrix

belonging to the well-known j-th class of protocols of the NSCR. Sufficient consider the volume of the j-th training sample W≥200, (see. Mathematical Encyclopedic Dictionary. / Gl. Red. Prokhorov Yu.V. - M .: Big Russian Encyclopedia, 1995. Reprint edition - M: Sov. Encyclopedia, 1988 . - 847 p.).

,

квадратной эталонной матрицы Φ_{Lj эт} содержат значения вектора ϕ_{Lj эт}, последовательно сдвинутые в каждой строке относительно первой строки влево на величину

.Then, on the basis of the vector ϕ _{Lj et} , a square reference matrix Φ _{Lj et is formed} (see Fig. 6), for which the first row ϕ _{Lj et} (0) contains the reference vector ϕ _{Lj et} , the remaining rows

,

the square reference matrix Φ _{Lj et} contain the values of the vector ϕ _{Lj et} sequentially shifted in each row relative to the first row to the left by the amount

.

Simulation of the claimed method for recognizing NSCR protocols using digital information streams received from vocoders IPC-10-2400 {STANAG 4197) and MELPe-2400 {STA-NAG 4591) showed the following. In the absence of symbol losses in the first block of the digital information stream Y, the use of the redundancy coefficients ϕ as a measured vector_L due to the increase in its dimension, it provides recognition accuracy D 3 ... 5% higher (see Fig. 7, graph 1) than when using redundancy factors as a measured vector ϕ_Z smaller dimension (see Fig. 7, graph 2). Under conditions of loss of characters in the first block of the digital information stream Y, the recognition accuracy of the claimed method does not change (see Fig. 7, graph 1). At the same time, the recognition accuracy of the NSCR protocols using the measured vector of redundancy coefficients ϕ_Z does not exceed 16.3% (see Fig. 7, graph 3) due to the displacements of the elements of the measured vector of redundancy coefficients ϕ_Z. For example, in FIG. 8a shows the measured vector of redundancy coefficients ϕ_Zformed in the absence of symbol loss in the first block of the digital information stream Y (see Fig. 8a). The loss of the first four bits in the first block of the digital information stream Y leads to a shift of all subsequent elements of the measured vector of redundancy coefficients ϕ_Z (see Fig. 8b). Therefore, a recognition method using a measured vector of redundancy coefficients ϕ_L, has better accuracy compared to the recognition method using the measured vector of redundancy coefficients ϕ_Z, by increasing the dimension to L and eliminating the influence of the loss of characters in the first block of the digital information stream Y.

(строками всех квадратных эталонных матриц Φ_{Lj эт}) с целью определения отклонения

в соответствии с (5), и блок принятия решения 14, предназначенный для определения принадлежности анализируемого цифрового информационного потока Y к наиболее вероятному j-му протоколу НСКР в соответствии с правилами (6) и (7). Кроме того, устройство содержит первый и второй блоки памяти 9 и 13 соответственно и генератор тактовых импульсов 10.The proposed method for recognizing NSCR protocols can be implemented using a device whose generalized structural diagram is shown in FIG. 9. The device contains a series-connected input selector 4, intended for registration (storing a predetermined volume of the input digital information stream Y), a first calculator 5, intended for calculating the normalized autocorrelation function A in accordance with (2) and based on it determining the values of N _b and K (expressions (1) and (3), respectively), the first driver 6, designed to generate the information matrix of the received signal Y _{K × M} , the second calculator 7, designed to determine the coefficient in redundancy ϕ _z in accordance with (4), the second shaper 8, designed to generate a vector of measured redundancy coefficients ϕ _L , the third calculator 11, designed for elementwise comparison of the measured vector of redundancy coefficients ϕ _L with reference vectors

(by the rows of all square reference matrices Φ _{Lj et} ) in order to determine the deviation

in accordance with (5), and decision block 14, designed to determine whether the analyzed digital information stream Y belongs to the most probable j-th protocol of the NSCR in accordance with rules (6) and (7). In addition, the device contains first and second memory blocks 9 and 13, respectively, and a clock generator 10.

At the preparatory stage, the required minimum and maximum signal sample sizes N _{cp min} and N _{cp max are} set on the first installation bus 1. The second installation bus 2 receives square reference matrices Φ _{Lj et} for all J known NSCR protocols, the calculation of which is carried out in accordance with (8) at the preparatory stage and is then stored in block 9. On the third installation bus 12, the operator sets the threshold level R, which recorded in the second memory block 13.

In the process of operation of the device (see Fig. 9), the digital information stream Y enters the input of block 4 through the input bus 3. The functions of block 4 include its storage in the amount of N _{cp max} , but not less than N _{cp min} . These values are set on bus 1. Otherwise, a signal is generated on the second group of its outputs (first output bus of device 15) about the end of the device and the need to proceed to the analysis of another information sequence.

The memorized sample of the digital information stream Y under the influence of the clock pulses of block 10 from the output of block 4 goes to the group of information inputs of the first calculator 5. The functions of block 5 include the calculation of the normalized autocorrelation function A, its analysis to detect regular extrema of the autocorrelation function A with equal intervals Δτ , splitting the received digital information flow to Y blocks of N information bits _used by each of the number in the intervals between the extrema of the normalized autocorrelation A. the first group of information outputs of block 5 block at a N _b bit information sequence Y is supplied to a group of information inputs of the first generator 6. Otherwise, when a signal is generated in block 5 are detected regularly at equal intervals Δτ extrema of the function A, the second group of information outputs its the completion of the analysis of the received information sequence, which is fed to the first output bus 15 of the device.

(см. фиг. 5) с последующим вычислением значений ϕ_z, z=1, 2, …, L в соответствии с (4).In block 6, information matrix Y _{K × M is formed} by sequential block recording of digital information stream Y in accordance with FIG. 4 and the additions of two columns in accordance with FIG. 5. The value of the information matrix Y _{K × M} in the parallel code is supplied to the group of information inputs of the second calculator 7. In block 7, the redundancy coefficients are determined by dividing the information matrix Y _{K × M} by a sequence of L rectangular matrices

(see Fig. 5) with the subsequent calculation of the values of ϕ _z , z = 1, 2, ..., L in accordance with (4).

The obtained values of ϕ _{z are} sequentially fed to the group of information inputs of the second shaper 8.

In block 8, they are sequentially recorded (the formation of the vector of the measured redundancy coefficients ϕ _L ) upon receipt.

;

, и наименьшего значения

. Данная величина поступает на первую группу информационных входов блока принятия решения 14. На второй группе его информационных входов присутствует значение порога R, поступающее с группы информационных выходов второго блока памяти 13. Блок 14 выполняет функцию сравнения значений

и при выполнении условия (6) на его группе информационных выходов (одновременно являющихся второй выходной шиной устройства 16) формирует значение j, соответствующее идентифицированному протоколу НСКР. В противном случае выдается информация об обнаружении нового, ранее неизвестного протокола НСКР. Синхронизация работы элементов устройства обеспечивается импульсами генератора тактовых импульсов 10.The generated value of the vector of measured redundancy coefficients ϕ _L goes to the first group of information inputs of the third calculator 11. The second group of its information inputs contains the values of square reference matrices Φ _{Lj et} coming from the group of information outputs of the first memory unit 9. In block 11, the element-by-vector comparison is performed measured redundancy coefficients ϕ _L with rows (reference vectors) of all J square reference matrices Φ _{Lj et} , determination of their mutual deviations

;

, and the smallest value

. This value goes to the first group of information inputs of decision block 14. At the second group of its information inputs there is a threshold value R coming from the group of information outputs of the second memory block 13. Block 14 performs the function of comparing values

and when condition (6) is fulfilled, on its group of information outputs (simultaneously being the second output bus of device 16) generates a value j corresponding to the identified NSCR protocol. Otherwise, information is issued about the discovery of a new, previously unknown NSCR protocol. The synchronization of the operation of the elements of the device is provided by the pulses of the clock generator 10.

The implementation of the elements of the NSCR protocol recognition device is widely covered in the literature and does not cause difficulties. The input selector 6, the first and second formers 5 and 7, respectively, the first and second memory blocks 9 and 13, respectively, can be implemented on memory chips (see Large integrated circuits of memory devices: Reference / A.Yu. Gordonov et al.: Ed. A.Yu. Gordonov and Yu.N. Dyakonov. - M .: Radio and communications. 1990. - 288 p.). The first, second and third computers 5, 7 and 11 can be implemented on microprocessors with sufficient speed (see Shevkoplyas B.V. Microprocessor structures. Engineering solutions: Handbook. 2nd ed., Rev. And add. - M .: Radio and communications, 1990. - 512 p.). Decision block 14 is a comparison block, which can be implemented on the elementary logic of TTL chips.

Claims (3)

1. A method for recognizing low-speed speech coding (NSCR) protocols implemented in vocoders, which consists in receiving a digital information stream Y over a time interval ΔT, based on the received stream Y, they generate a normalized autocorrelation function A, according to regular extrema at equal intervals Δτ of the autocorrelation function A make a decision about the presence of a block structure in the digital information stream Y, at the intervals between the extrema of the autocorrelation function A divide the digital information current Y to information blocks with a volume of N _b bits each, sequentially assign sequence numbers k = 1, 2, ..., K to information blocks, starting from the first information block, form a rectangular information matrix Y _{K × L} , L = N _b ; the rows of which are successively placed under each other information blocks in accordance with their serial numbers k = 1, 2, ..., K, form a sequence of rectangular matrices

, z = 1, 2, ..., Z; Z = N _b -2, containing three adjacent columns of a rectangular information matrix Y _{K × L} , where the current number z of the rectangular matrix

determined by the column number of the matrix Y _{K × L} , which is the first column of the matrix

, for each of the matrices

calculate the redundancy coefficient ϕ _z , on the basis of the obtained set of redundancy coefficients form the measured vector of the redundancy coefficients ϕ _Z by sequentially placing them in accordance with serial numbers z, form the standard vectors of the redundancy coefficients ϕ _{Zj et the} known NSCR protocols, the measured vector of redundancy coefficients ϕ _{Z are} compared with reference vectors redundancy coefficients φ _{Zj fl,} j = 1, 2, ..., J, known protocols NCIS, characterized in that the increase to the value of L, L = Z + 2, dimensionally be measured vector redundancy coefficients φ _L, is increased to values of L dimension reference vector coefficients redundancy φ _{Lj fl} to known protocols NCIS, based on the reference vectors of coefficients redundancy φ _{Lj et} form square reference matrix Φ _{Lj fl,} j = 1, 2, ..., J, comparing the measured vector of redundancy coefficients ϕ _L with the lines

,

= 0, 1, ..., L-1, of all J square reference matrices Φ _{Lj et} , deviations

between the measured vector of redundancy coefficients ϕ _L and the rows (reference vectors) of all J square reference matrices Φ _{Lj et} , decide in favor of the j-th NSCR protocol for which a minimum and less than a given threshold R deviation is provided

measured vector ϕ _L from

row of the jth square reference matrix Φ _{Lj et} . 2. The method according to p. 1, characterized in that the increase in the dimension of the vector ϕ _Z to the vector ϕ _{L is} achieved by increasing the dimension of the information matrix Y _{K × L} by using its first and second columns as the penultimate and last columns, respectively, of the matrix Y _{K × M} , M = N _b +2, and the subsequent calculation of additional redundancy factors ϕ _L-1 and ϕ _L. 3. The method according to p. 1, characterized in that the elimination of the influence of the left column shift of the information matrix Y _{K × L} in the absence of N _p characters in the first block of the digital information stream Y, N _p <N _b , is carried out by forming square reference matrices Φ _{Lj et} whose first rows ϕ _{Lj et} (0) contain the reference vectors ϕ _{Lj et} , the remaining rows ϕ _{Lj et} (

) the square reference matrices Φ _{Lj et} contain the values of the vectors ϕ _{Lj et} (0) sequentially shifted in each row relative to the first row to the left by the amount

bit,

= 1, 2, ..., L-1.

Images