RU2646867C1 - Method for transmission and reception of digital information in general - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: invention relates to digital information transmission and can be used in receiving devices of synchronous digital communication systems. Method for transmitting and receiving digital information in general comprises, at the transmitting side, converting the information symbol sequence to be transmitted into a stream of information unit elements, converting the information symbol sequence to be transmitted into a stream (collection) of redundant unit elements, merging the streams of information and redundant unit elements into a final stream of unit elements, converting resulting stream of unit elements into oscillations of a hydroacoustic or electromagnetic field, at the receiving side,converting the received oscillations of the hydroacoustic or electromagnetic field into electrical signals, according to the signal observed for the message reception time interval (MRTI) for each r-th information unit element of the message (r = 1…R is the sequence number of the information unit element in the message), the decisive statistics for the element-wise reception of information unit elements are calculated, according to the signal observed at the MRTI with respect to each k-th redundant unit element of the message (k = 1…K is the sequence number of the redundant unit element in the message), the decisive statistics are calculated for the element-wise reception of redundant unit elements, for each of the decisive statistics of element-wise reception of information unit elements, element-wise decision-making is implemented, for each of the decisive statistics of element-wise reception of redundant unit elements, element-wise decision-making is implemented, based on the sets of the decision-making results for each of decisive statistics of element-wise reception of the sequence of information unit elements and the solutions for each of the decisive statistics of element-wise reception of the sequence of redundant unit elements, the number of that set of results of forming decisive statistics of element-wise reception of information unit elements, which does not contain errors, the decision making operation on the fidelity of one of the hypotheses of the sequences of information unit elements is performed on the basis of the number of that set of results of forming decisive statistics of element-wise reception that does not contain errors.

EFFECT: high noise immunity of transmission.

1 cl, 1 dwg

Description

The invention relates to the field of digital information transmission and is intended for use in receiving devices of synchronous digital communication systems. The transmission of information, unless otherwise specified, refers to the totality of the functions of both the actual transmission and reception.

One of the significant factors negatively affecting the efficiency (transmission speed and / or decoding error probability) of digital communication systems is the presence of intersymbol interference (ISI) due to multipath propagation. This provision is due, in particular, to the fact that the presence of ISI entails interference caused by the temporary overlap of transmitted sequentially symbols (hereinafter, we call these symbols elementary premises (EP)). These interference are called interference interference (IP). One way to overcome this problem is to implement the so-called reception as a whole (the definition of this term is given below). However, the implementation of the reception is generally relevant in the absence of MSI. The following are descriptions of the prototype and the proposed method in options involving measures to suppress IP. However, in principle, the implementation of the proposed method is possible and relevant in the absence of MSI (IP).

A known method of detection (as an essential element of the transmission method) of signals (messages) in digital communication systems in general [1, p. 636]. The disadvantage of this analogue is the necessary significant computational resources that implement this method of hardware and software, which is due to the implementation of the set of processing channels provided for by it, each of which is designed to receive the corresponding combination of ES (each possible combination of ES is hereinafter referred to as an alternative to the sequence of ES). In addition, this analogue is not designed specifically for receiving messages in the conditions of MSI.

The specified drawback (the necessary significant computing resources that implement the method of hardware and software) is also characteristic of the method described [2].

A known method of transmitting a digital message, which includes, in particular, a set of reception operations as a whole in the conditions of multipath propagation, described in [3, p. 193, 194]. The basis of this method is the Klovsky-Nikolaev algorithm (AKN). Its disadvantage is the relatively low noise immunity of receiving (decoding) a message, due to the fact that it implements the operation of elementwise decision making.

Explanation of terminology. Reception of the message as a whole with element-wise decision-making - a method of decoding a message that involves making decisions (i.e., determining code combinations carried by ES) separately for each ES of this message, but using the information contained in the signal received over the time interval of the entire message . Receiving a message as a whole with making a decision as a whole (usually referred to simply as receiving a message as a whole without specifying that a decision is also made as a whole) is a method of decoding a message that involves making a unified decision on an alternative to this message a sequence of EPs (code combinations carried by them) using (as in the case of receiving the message as a whole with elementwise decision making) the information contained in the signal received over a time interval of only communication.

A message can be considered as a sequence of its relatively short fragments. Moreover, each of them can be considered as a private message, in relation to which the method of reception as a whole is applicable.

Closest to the technical nature of the claimed is a method of transmitting a digital message, which includes a set of receiving operations as a whole in a multipath environment, described in [4] (prototype); the basis of this set of operations is the modified Klovsky-Nikolaev algorithm (MACN). Due to the brevity of the description of the prototype in [4], when describing it, materials from sources [3 and 5] are given below, supplementing the description given in [4]; all these materials, in particular the totality of operations performed in a communication system on its transmitting side, were implied by the authors of publication [4].

To simplify the description (including the description of the proposed method), it is further assumed that the shape of the impulse response of the channel (IRF) of the multipath propagation at the receiving end of the communication system is known (in order to measure it on the transmitting side at r = 0-th clock cycle of the system is formed and a test pulse is transmitted, the shape of which is known on the receiving side); in addition, to simplify the description, the set of operations ensuring its measurement at the receiving side of the communication system is omitted. A similar situation occurs in the description of the prototype [4].

The principle of operation of the prototype is as follows. On the transmitting side of the communication system, the operations of converting the information to be transmitted into the electronic information stream and their actual broadcasting, the conversion of electrical signals into vibrations, for example, sonar or electromagnetic fields.

On the receiving side, the received vibrations of the hydroacoustic or electromagnetic field are converted into electrical signals.

The assumption of which of the alternatives to the EP sequence is accepted is hereinafter referred to as the hypothesis of the EP sequence. When the set of symbols of the alphabet of transmitted messages is known at the receiving end of the communication system, the form of the signal received or observed at the time interval for receiving the message (IVPS) of the signal through which the entire message is transmitted, if each of the above hypotheses is true, is known. Given this circumstance, in the prototype, according to the signal observed at each rth cycle of the IVPS, with respect to each of the set of hypotheses of the EP sequences of the received message (taking into account the array of decision statistics (PC) of the element-wise reception calculated on the r-1st cycle), the corresponding these hypotheses for the implementation of residual and accompanying components of PI.

Explanation of terminology. The residual component of the PI (or the residual signal of the MSI) is an interference occurring due to the temporary overlap with the analyzed EP of those EPs that were transmitted before it; the accompanying component of IP is an interference occurring due to temporary overlap with the analyzed EP of those EPs that were transmitted after it [3, p. 193].

The entire IVPS is divided into measures, the duration of each of which is equal to the sum of the durations of the ES and the KFM, and the time interval between the moments of the beginning of the time-adjacent measures is equal to the duration of the ES. The process (set of operations) of signal processing on the time interval of each rth beat is hereinafter referred to as the rth beat of the communication system or simply the rth beat.

Decisive statistics (PC) is the result of a set of operations in the general case of spatio-temporal signal processing. An array of PCs is formed in each digital communication system, each of which is designed (tuned) to receive a specific alternative to the received electronic signature. In this case, the PC array is used directly for decision making by comparing each formed PC with a threshold and / or with other PCs. In the description of the prototype PC - the calculation result located in the ratio (1) in its curly brackets (see also PC z _i calculated in accordance with the expression (1) below).

Bitmap PC - is a PC designed to provide bitmap decision making.

At each rth beat, the IHPS compensate for the realization of the signal observed at this beat of each of the combinations of the results of the formation of realizations of all the above-mentioned IP components.

Then, according to the signal observed at the IVPS and each combination of the sequence of EP components of the IP components determined for the corresponding hypothesis, the prototype sequentially computes the PC of the element-by-element reception for each r-th EP message (r = 1 ... R is the serial number of the ES in the message). Further, when describing the prototype and the proposed method, we believe that the operation of forming the PC of the bitwise reception includes, among other things, the operation of forming implementations of the residual and accompanying components of the IP, as well as the operation of compensation in the implementation of the signal observed at the current beat of each of the combinations of the results of forming implementations the specified components of IP.

For each of the element-wise PCs, element-wise decision-making is implemented.

Next, the prototype calculates the array of PC reception as a whole, each of which corresponds to one of the hypotheses of the sequence of electronic messages received.

A number of the above operations of the prototype are mathematically described as follows. Let the realization of the residual component of the PI calculated on the assumption that all ES accepted before the start of the analysis (decoding) of the rth ES be decoded correctly, denoted as q _{r ost} (t) (here, the index number of the hypothesis of the sequence of ES in the message i is omitted due to the fact that, as has just been noted, all ESs received at the time of analysis of the i-th ES are considered decoded correctly, and the element-wise decision made on them in the prototype is the only one), and the accompanying interference component as q _{ir sopp} (t). In the prototype, each i-th (with index i varying in the range of values from 1 to M, where M is the possible number of hypotheses of EP sequences in the message) PC reception as a whole Z _{i is} calculated as

where r is the serial number of the analysis subinterval (or the time subinterval on which the rth ES is located) corresponding to the rth tact of the communication system, and the value of r varies in the range of values 1 ... R, where R is the number of ESs in the message; T _e - the duration of a single EP that has passed the multipath channel (equal to the sum of the durations of a single transmitted EP and KFM); T - the period of following the ES in the message (equal, as a rule, to the duration of a single transmitted ES T _e ).

Each integration result in (1) is a piecewise reception PC, and each summation result (i.e., Z _i ) is a reception PC as a whole.

номера истинной гипотезы последовательности ЭП в принятом сообщении по формулеThe operation of deciding on the fidelity of one of the hypotheses of EP sequences using an array of PC reception as a whole Z _i provides for the calculation of the estimate

numbers of the true hypothesis of the EP sequence in the received message by the formula

-той гипотезы, которой соответствует PC приема в целом Z_i наименьшего уровня.those. numbers (index)

the hypothesis, which corresponds to the PC reception as a whole Z _{i the} lowest level.

The disadvantage of the prototype is the relatively low noise immunity of reception (decoding) due to the following. The principle of its operation is based on the fact that the estimate of the energy of a difference signal in calculating the PC (see formula (1)), which corresponds to the really accepted alternative to the EP sequence, is smaller than all other PCs only in the sense of a trend, i.e. not guaranteed at all. This phenomenon is characteristic of all possible variants of PC formation. However, in this case, this effect is manifested to a special extent for the following reason. In the case of correctly guessing the accepted alternative to the EP sequence, there is a coincidence of the formed temporary implementation of the PI and PI, which is a component of the received signal implementation; when calculating the difference signal, the PI, which is a component of the received signal implementation, is compensated, and due to this, the energy of the difference signal decreases. But with a relatively small signal-to-noise ratio, this tendency is often (and, possibly, as a rule) can be violated. This negative effect is due to the fact that the reception of messages is carried out with a disturbing effect (along with IP) and background noise. In the case of radio communications, this may be the electrical noise of the input circuits of the receiving stations, and in the case of sound underwater (hydroacoustic) communications, the noise of the sea. In this case (i.e., with a small ratio of the total level of the received signal and SP to the background noise level), the compensation of the SP taking place in the calculation of the difference signal does not lead to a significant decrease in the level of this difference signal, since under such conditions the level of the difference signal is determined not so much IP how much background noise. The background noise during the formation of the difference signal is not compensated.

The aim of the proposed method is to increase the noise immunity of the transmission.

The goal is achieved in that in the method of transmitting digital information, the implementation of which

on the transmitting side

- convert the sequence of information symbols to be transmitted into a stream of information ES,

- convert the resulting flux of ES into fluctuations in the hydroacoustic or electromagnetic field,

on the receiving side

- convert the received vibrations of the hydroacoustic or electromagnetic field into electrical signals,

- according to the signal observed at the IVPS as applied to each r-th informational ES of the message (r = 1 ... R is the serial number of the informational ES in the message), the PC of the element-wise reception of informational ES is calculated,

- for each of the PCs of the element-wise reception of informational electronic devices, element-wise decision-making is implemented,

- perform the operation of deciding on the fidelity of one of the hypotheses of sequences of informational ES,

moreover

on the transmitting side

- convert the sequence of information symbols to be transmitted into redundant EP,

- combine the flow of information and excess ES in the final stream of ES,

on the receiving side

- according to the signal observed at the IVPS in relation to the excess EF, the element-wise reception PC is calculated,

- on the PC of the element-wise reception of excess ES, implement the element-wise decision-making,

- based on the totality of the results of the formation of the PC of the element-wise reception of information and excess EP, find the number of that set of the results of the formation of the PC of the element-wise reception, which does not contain errors,

- the decision-making operation on the fidelity of one of the hypotheses of the sequences of informational EPs is performed on the basis of the number of the totality of the results of the formation of the PC of the bitwise reception that does not contain errors.

The meaning of the inventive concept underlying the proposed method is to use the standard (arbitrary) method of receiving a message as a whole with the addition of the operation of generating excess EF on the transmitting side of the system, which makes it possible to detect detection errors, and on the receiving side, deciding on the fidelity of one of hypotheses of sequences of informational EPs based on the result of the operation of detecting detection errors (in the prototype, as in all other analogues, this is this is taken on the basis of the analysis of the PC reception array as a whole; in the claimed method, the operation of forming the PC reception array as a whole is absent).

A flowchart illustrating the inventive method is shown in FIG. 1, where are indicated:

- 1 - conversion of the sequence of information symbols to be transmitted into the stream of information ES;

- 2 - conversion of the sequence of information symbols to be transmitted into redundant electronic signature;

- 3 - combination of information and excess ES flows into the final ES flow;

- 4 - conversion of the final flux of electromagnetism into vibrations of the hydroacoustic or electromagnetic field;

- 5 - conversion of the received vibrations of the hydroacoustic or electromagnetic field into electrical signals;

- 6 - calculation of the PC of the element-by-element reception by the signal observed at the IVPS as applied to each r-th informational ES message (r = 1 ... R is the serial number of the informational ES in the message),

- 7 - decision-making on each of the PC element-wise reception of a sequence of information ES messages,

- 8 - calculation of the PC of the element-wise reception according to the signal observed at the IVPS as applied to the excessive electronic message,

- 9 - making a decision on the PC of the element-wise reception of the excess ES messages,

- 10 - finding the totality of the results of the formation of the PC of the step-by-step reception of the information of the electronic information message and the excess EF of the number of the set of the results of the formation of the PC of the step-by-step reception of information ES messages that does not contain errors,

- 11 - making a decision on the fidelity of one of the hypotheses of sequences of informational ES.

Operation 1 (the conversion of the sequence of information symbols to be transmitted into the informational ET stream) is performed, for example (see [6, 7]), by generating a pseudo-random sequence (SRP) with cyclic time shift (CVS) for each of these symbols, the value of which ( in units equal to F ^-1 s, where F is the bandwidth of the frequency bandwidth in Hertz) is equal to the digit corresponding to the digital information symbol to be transmitted. So, when transmitting a character, for example, Ω = the 5-bit binary code is “00010” (note that the binary code “00010” is equivalent to the decimal code “002”), the specified DAC entered into the memory bandwidth is 2 units (unit size introduced in the SRP TsVS is specified above). Next, each PSP formed in this way is multiplied by a high (working) frequency sinusoidal signal. The result of this multiplication with the participation of each of these PSPs is the informational EP corresponding to this PSP. There is a one-to-one correspondence between the number of the transmitted alternatives to ES m and the binary code of the corresponding information symbol (the same applies to the redundant symbol).

Operation 2 (converting the sequence of information symbols to be transmitted to the redundant EP) consists of the following two particular operations. When performing the first of them, the sequence of information symbols is converted into a redundant symbol, and the second is the conversion of the redundant symbol to a redundant EP. The second sub-operation completely coincides with the operation 1 described above, and in this case the same correspondence table can be used as when performing the operation 1.

A particular operation of converting a sequence of information symbols into a redundant symbol (and further into a redundant EP) provides the possibility of detecting detection errors (by detection errors are the decision-making errors of PC element-wise reception of a sequence of informational ES messages). As such an operation, an arbitrary error-correcting coding operation with error detection can be used. Further, as a variant of such a particular operation, the CRC (Cyclic redundancy code — cyclic redundancy code) algorithm (procedure) is considered [8 ... 11]. When implementing this procedure, the redundant character is the so-called checksum. The procedures are described in a wide range of CRC checksum bit depth variations, for example, in [11]. The particular operation under consideration “... is based on the properties of division with the remainder of the polynomial into a polynomial. In fact, the result of CRC checksumming is the remainder of dividing the polynomial corresponding to the source data by the generating polynomial of fixed length ”[8]. The same generating polynomial of fixed length is also present (i.e., known) at the receiving end of the transmission system.

Thus, the redundant character in this case is the indicated remainder of division. This remainder from the division can be formed on the transmission interval of the entire message both once or several times (in this case, in the first case, the mentioned source data, which corresponds to the polynomial used as the dividend, is the data array of the entire message, and in the second, data subarrays, formed by splitting the data array of the entire message into these disjoint modarrays). In the present description (as well as in the claims), an example of the implementation of the proposed method is considered in the embodiment of the presence of one excess ES located in the final ES stream after the last (i.e., Rth) information ES.

With the length (length) of the checksum equal to p, the number of alternatives to the excess EP m _ν is V = 2 ^p . Each m _ν -th excessive VC corresponds to the binary code generated in step 2 (this code is the remainder of said division in the preceding paragraph).

Due to the fact that, on the one hand, the detailed detail (description) of the CRC checksumming procedure is very cumbersome, and, on the other hand, this procedure is exhaustively described in accessible sources and is known to specialists (and is widely used in data exchange networks and in storage devices for checking information for authenticity, as well as protection against unauthorized changes), this detailed specification is omitted in the present description, we only indicate that if there is a stream of information binary symbols at its input s at its output a binary code is generated redundancy symbols equal to said reference amount.

In order to ensure a negligible probability of decision-making errors on the PC by the element-wise reception of the redundant electronic signaling, carried out on the receiving side, when performing a particular sub-operation of converting a sequence of information symbols into an excess symbol, in addition to the indicated actions, an encoding operation of this symbol with error correction can be performed. Moreover, the indicated redundant symbol is supplemented with binary symbols formed in accordance, for example, with the Reed-Solomon algorithm.

Operation 3 (combining information and excess EF flows into the final EF stream) is implemented by storing the totality of information and excess EFs and their sequential output (reading) to the input of the device performing operation 4 (converting the resulting EF stream into hydroacoustic or electromagnetic field oscillations). Moreover, the sequence of the specified issuance may be arbitrary; it is only important that it be known at the receiving end of the transmission system (data exchange system). As indicated above, for specificity, an example of the implementation of the proposed method is described in the embodiment of the arrangement of excess ES in the final ES stream after the last information ES.

Operation 4 (converting the resulting ES stream to vibrations of a hydroacoustic or electromagnetic field, or, more simply, transmitting the resulting ES stream), as a rule, is performed on analog signals. As a result of the implementation of the above operation 3, digital signals are generated. In this regard, it is understood that the operation of analog-to-digital conversion (ADC) is implemented either as the last part of operation 3 (in this case, the result of the operation 3 becomes the ES stream in analog form), or as part of operation 4 (in this case, operation 4 begins with the ADC). ADC as an independent operation of the proposed method is not considered, since it is not associated with any functional transformations of information.

Operation 4, for example, in the case of a sonar or sonar communication system is implemented by a sonar emitter. The transmission and reception time interval of one ES unit (minus the propagation time of signals between the transmission and reception points) is the transmission and reception cycle of information (i.e., the communication system operation cycle).

Operation 5 (conversion of the received vibrations of the hydroacoustic or electromagnetic field into electrical signals) in the considered example of a sound underwater communication system provides for the conversion of acoustic vibrations of the aqueous medium into electrical signals. In this case, it is realized by a hydrophone or, in a more complicated case, an antenna array containing a set of hydrophones, a set of delay lines and an adder (see [12], Fig. 1.5b, 1.6 and 1.7).

Operation 6 (calculation of the PC of the element-wise reception by the signal observed at the IVPS as applied to each r-th informational ES message (r = 1 ... R is the sequence number of the informational ES in the message) consists of three phases (sub-operations) and is performed, for example, as follows.

альтернатива ЭП; здесь m - номер альтернативы ЭП, причем m=1…M, где M - количество возможных альтернатив отдельной ЭП, k_r-1-индекс, равный порядковому номеру той из K_r-1 PC поэлементного приема, которая отобрана в итоге выполнения операции 7 на r-1-м такте ИВПС (описание операции 7 приведено ниже).The first of these suboperations generates an array of realizations of the residual component of SP. It is based, as well as the basis of the corresponding operation of the prototype, is the feedback on solutions (OCP). The meaning of the use of OCP in this case is as follows. In order to form the residual component of IP at each r-th beat of the IVPS in the situation under consideration, information is needed on which particular alternative to the ES was transmitted at the r-1-th beat of the IVPS. Namely, the ES transmitted on the r-1st beat of the IVPS partially overlaps with the ES received on the rth beat of the IVPS, i.e. it determines the form of the residual component of the PI to be compensated. This information is generated by making a decision based on the PC bit-wise reception, developed on the specified r-1-st tact of IVPS. In the inventive method, at each, including, for example, the r-1-st cycle of the IHPS based on the PC bit-wise reception, a set of K _r-1 > 1 decisions is taken. At the same time, each k _r-1- th of these solutions has a presumptive decision that at the indicated r-1-st clock

alternative to ES; here m is the number of alternatives to the electromotive force, and m = 1 ... M, where M is the number of possible alternatives to an individual electromagnet, k _{r-1 is the} index equal to the serial number of that of K _r-1 PC bitmap, which was selected as a result of operation 7 on the r-1st cycle of the IVPS (operation 7 is described below).

. В обеспечение этого на каждом r-1-м такте формируется массив из K_r-1 PC поэлементного приема, обозначаемых как PC_kr-1 (операция их формирования совпадает с таковой, реализуемой на каждом r-м такте, описанной ниже (см. описание второй подоперации, входящей в состав операции 6)). Каждой из этих PC_kr-1 соответствует

альтернатива ЭП. Т.е. наибольшая из указанных PC_kr-1 поэлементного приема, имеющая по определению ранг K_r-1, характеризуется индексом k_r-1=1, и ей соответствует

альтернатива ЭП, следующая PC поэлементного приема, имеющая ранг K_r-1-1, характеризуется индексом k_r-1=2, и ей соответствует

альтернатива ЭП и т.д.In other words, in the claimed method, an array of K _r-1 realizations of the residual component of the IP is formed at each rth beat

. To ensure this, an array of K _r-1 PC of element-wise reception, designated as PC _kr-1 (the operation of their formation coincides with that implemented at each r-th beat described below (see description the second suboperation, which is part of operation 6)). Each of these PCs _kr-1 corresponds to

alternative to EP. Those. the largest of the indicated PC _{kr-1 of the} bitwise reception, which by definition has the rank K _r-1 , is characterized by the index k _r-1 = 1, and it corresponds to

an alternative EP, the next PC of the bitwise reception, having the rank K _r-1 -1, is characterized by the index k _r-1 = 2, and it corresponds to

alternative to ES, etc.

может быть, например, следующим. Пусть Sm(t) - временная реализация m-й альтернативы передаваемой ЭП, а

- временная реализация той альтернативы передаваемой ЭП, которая, в соответствии с результатом ОСР, принята на r-1-м такте работы системы. Тогда реализация подлежащей компенсации остаточной компоненты ИП, соответствующей этой

альтернативе ЭП, вычисляется (с учетом того, что ЭП, принятая на r-1-м такте, опережает ЭП, принятую на r-м такте, на время T_э) какThe rule of formation of each implementation

may be, for example, the following. Let Sm (t) be the temporary implementation of the mth alternative of the transmitted electron beam, and

- a temporary implementation of that alternative of the transmitted ES, which, in accordance with the result of the OCR, was adopted at the r-1-st clock cycle of the system. Then the implementation of the compensable residual component of the IP corresponding to this

alternative to the ES, it is calculated (taking into account the fact that the ES received on the r-1st step is ahead of the ES received on the r-th time by the time T _e ) as

при t<0 или при t>Т_э.moreover

at t <0 or at t> T _e .

и их задержек относительно луча, пришедшего в точку приема первым,

(

- номер луча). Вычисления реализаций остаточных компонент по формуле (3) производятся K_r-1 раз, т.е. раздельно применительно к каждому из значений k_r-1=1…K_r-1.Acceptance of the prominence of the KFM (it is measured in advance) means the popularity of the parameters that characterize it (3) as arrays of L (this is the number of rays arriving at the receiving point) of the ray amplitudes

and their delays with respect to the beam arriving at the receiving point first,

(

- beam number). The calculations of the implementations of the residual components by the formula (3) are performed K _r-1 times, i.e. separately for each of the values of k _r-1 = 1 ... K _r-1 .

.As a result of the implementation of the described suboperation, at each rth step, K _r-1 results of compensation of the residual component of the SP are formed

.

The second sub-operation, which is part of operation 6, provides compensation for each of the results of the formation of realizations of the residual component of the IP. It is realized by K _r-1- fold (i.e., for each k _r-1 separately) calculating a difference of the form

where S _r (t) is the implementation of the signal observed at the rth cycle of the system.

.As a result of this second sub-operation, at each r-th beat, K _r-1 results of compensation of the residual component of IP are formed

.

The third suboperation, which is part of operation 6, provides the calculation of the actual PC array of element-wise reception. It is implemented, for example, as follows. With the known (measured in advance) KFM, each mth alternative of the transmitted ET Sm (t), as in the prototype, is recalculated to the receiving point (i.e., predicted) by calculation using the following formula (here the index k _{r is} omitted as not carrying at described prediction no information)

и опорной функцией, совпадающей с соответствующей (т.е. m-й при переборе всех значений m=1…M) альтернативой, принимаемой на этом такте ЭП y_m(t), определяется по формулеThe correlation value between each result of compensation of the residual component of IP formed on the rth cycle of the IVPS

and a support function that coincides with the corresponding (i.e., the mth alternative when enumerating all the values m = 1 ... M) taken at this ES step y _m (t), is determined by the formula

, и номером этой PC k_r - произвольное). Далее индекс, характеризующий номер m альтернативы ЭП, применительно к которой на r-м такте вычислена PC, уточняется как m_k,r.The index k _r at the value calculated on the rth step PC (6) is the number of this PC. In total, such PCs on the rth step of the PC compute K _r-1 M pieces, i.e. these PCs are calculated for each of the K _r-1 M combination of indices k _r-1 and m (the ratio between the indices k _r-1 and m, for which the PC value is calculated

, and the number of this PC k _r is arbitrary). Next, the index characterizing the number m of the ES alternative, in relation to which the PC is calculated on the rth beat, is specified as m _{k, r} .

) может, как и в прототипе (см. соотношение (1)), вычисляться с учетом также и сопровождающей компоненты ИП q_{ir сопр}(t). Данное уточнение не принципиально и поэтому для упрощения описания опускается. С другой стороны, при практическом отсутствии МСИ в процедуре (6) вместо реализации

может использоваться реализация наблюдаемого сигнала S_r(t), в которой ИП не компенсирована.The temporary implementation of the IP compensation result (indicated as

) can, as in the prototype (see relation (1)), be calculated taking into account the accompanying component of the IP q _{ir sopr} (t) as well. This refinement is not essential and therefore is omitted to simplify the description. On the other hand, with the practical absence of ISI in procedure (6) instead of implementation

the implementation of the observed signal S _r (t) can be used, in which the PI is not compensated.

, сформированных на r-м такте. Путем сравнения всех K_rM PC указанного массива между собой определяется и запоминается отдельно от этого массива максимальная из них. При этом она из указанного массива удаляется. Далее процедура повторяется по отношению к массиву из K_r-1M-1 оставшихся PC и т.д. В итоге K_r-кратного выполнения указанной процедуры запомнены K_r PC, являющиеся в исходном массиве наибольшими по уровню (т.е. старшими членами вариационного ряда указанных PC). Имеются и варианты выполнения операции 7, являющиеся более экономичными по вычислительным затратам, чем описанный выше.Operation 7 (decision-making on each of the PCs by the element-wise reception of a sequence of informational electronic messages) is performed on each r-th step by selecting K _{r the} highest level-wise PC reception of elements and storing the corresponding indices m _{k, r} and m _{k, r-1} . The specified selection is carried out, for example, as follows. There is an (source) array of K _r-1 M RS element-wise reception

formed on the rth beat. By comparing all K _r M PC of the indicated array with each other, the maximum of them is determined and stored separately from this array. At the same time, it is deleted from the specified array. Next, the procedure is repeated with respect to the array of K _r-1 M-1 remaining PCs, etc. As a result of K _r -fold execution of the indicated procedure, K _r PCs are memorized, which are the highest in the initial array (i.e., the senior members of the variational series of these PCs). There are also options for performing operation 7, which are more economical in computational cost than described above.

The index k (or k _r ) for the numbers (or indices) of the alternatives to EP m is omitted in some cases; its use is mainly due to the need for numbering the PCs selected in step 7 at each clock cycle of the system.

, в расчете которой по формуле (4) использована та реализация

, которая в свою очередь вычисленная по формуле (3) применительно к индексу

при сигнале

, совпадающему с индексом m_k,r-1 (другими словами, индексы m_k,r и m_k,r-1 считаются относящимися к их сочетанию, если индекс m_k,r выработан или сформирован в итоге совокупности операций, в которой использована реализация остаточной компоненты ИП, вычисленная применительно к ЭП с индексом m_k,r-1).Regarding the memorization of combinations of indices m _{k, r} and m _{k, r-1} (here, the pair of information carriers are the indices m and r). Each combination of indices m _{k, r} and m _{k, r-1} fixed on the r-th bar corresponds to a combination of (element-wise) decisions that, at the r-1-th bar, it is possible (presumably) m _r-1 the nth alternative to EP, and on the rth beat - also, possibly, the m _rth alternative to EP (there are K _{r for} such possible or valid alternatives to EP) In this case, the following condition holds for the possibility of assigning indices m _{k, r} and m _{k, r-1} to the indicated combination. The indices m _{k, r} and m _{k, r-1} are considered to be related to the indicated combination, if the index m _{k, r is} generated on the basis of the PC that is generated on the rth beat using the implementation included in procedure (6)

, in the calculation of which, according to formula (4), that implementation

, which in turn is calculated by formula (3) as applied to the index

at signal

coinciding with the index m _{k, r-1} (in other words, the indices m _{k, r} and m _{k, r-1} are considered to be related to their combination if the index m _{k, r is} generated or generated as a result of a set of operations in which the implementation is used the residual component of PI calculated as applied to EP with index m _{k, r-1} ).

Each combination of EP indices m _{k, r-1} and m _{k, r} can characterize a possible fragment of the EP sequence in the message (in the theory of Viterbi algorithm, every possible EP sequence of the entire message is called a trajectory [13]; in this connection, we can also say that the indicated a pair of indices can characterize a fragment of a trajectory).

As a result of the operation 7, K _r combinations (or pairs) of indices m _{k, r-1} and m _{k, r are} fixed (remembered); fixing each such pair of indices means that a possible alternative to the fragment is the sequence of EPs of the entire message (this fragment is a subsequence consisting of two EPs) contains alternatives to EPs characterized by these two indices.

Operation 7 (as well as the operations preceding it) is implemented at each step of the method. In this regard, as a result of its implementation, for example, on the r-1st step, K _r-1 pairs of indices m _{k, r-2} and m _{k, r-1 are} recorded (stored).

Operation 8 (calculation of the PC of the element-by-element reception from the signal observed at the IVPS as applied to excess EF messages) is performed similarly to operation 6, but with refined support functions and integration limits, namely, as

where Y _ν (t) is the form of the νth alternative of the excess EP, recalculated to the receiving point similarly to (3), i.e. according to the formula

where S _ν (t) is the form of the νth alternative to the excess EF at the transmission point; ν = 1 ... V. The number (index) of this alternative of excessive EP is denoted as m _ν .

To increase the noise immunity of detecting excess EF between the moments of the location of the Rth (i.e., the last) information EF and excess EF during step 3, a protective time interval of width Δ approximately equal to the duration of the KFM can be introduced, which ensures the absence of IP in the reception interval of excess EP. In this case, instead of the temporal realization of the signal δ _{R + 1 stop} (t) in (7), the temporal realization of the signal S _ν (t) can (should) appear. In the presence of the specified protective time interval, the integration limits in (7) are shifted “to the right” by the value of this interval (that is, Δ is added to each limit), while the time boundaries of the R + 1-st cycle of the system operation become equal to RT + Δ ... RT + T _e + Δ.

The calculation of the PC of the element-wise reception of the excess EP z _ν (for example) according to formula (8) is carried out at the R + 1-st cycle of the system for each value of the index ν in the range 1 ... V; at the same time, by the element-by-element PC of excess EP calculated on this R + 1-st cycle, we mean the entire array of V values z _{ν, R + 1} for ν = 1 ... V.

Operation 9 (decision making on each of the PCs of the element-wise reception of the excess ES message) is performed, for example, as follows. At the R + 1-st cycle of the system, the index ν = ν ₀ is selected, which corresponds to the value z _ν0 , which is the highest in level among all values of z _ν (here, the index R + 1, which characterizes the number of the cycle at which the PC of ES, omitted, because with the considered variant of the proposed method, the message contains the only redundant ES, the specified PC is calculated only on the R + 1-st cycle of the system). Index ν ₀ is that number (binary code) according to which, in conjunction with the binary codes of each of the possible sequences of information ES, then (as a result of operation 10), a decision is made on the presence / absence of detection errors for these sequences of information ESs.

The value of the index ν ₀ determined in this way is the result of a decision on the PC by the element-by-element reception of an excess ES message.

Operation 9 can also be performed similarly to operation 7, i.e. with the formation as a result of its execution of a set of indices ν _k0 corresponding to K _{R + 1 the} highest in terms of PC level of the element-wise reception of excess EP.

Operation 10 (finding the number of the totality of the results of the formation of the PC of the element-wise reception of information-wise information messages that does not contain errors from the aggregates of the results of the PC formation of the element-wise reception of information and excess EP) includes several phases and is performed, for example, as follows.

In the first phase, according to the combinations of indices (alternative EP numbers) m _{k, r-1} and m _{k, r} for r = 1 ... R that are formed as a result of step 7 of the entire set of R clock cycles of the IVPS, the combination of possible sequences of ETs in the message is determined and this is possible EP sequences (as in the prototype) are assigned indices of their numbers i. Each of the possible EP sequences in the message is determined by comparing the sets of indicated pairs (combinations) of indices (i.e., m and r indices) developed on R clocks and selecting those EP sequences, each of which starting at r = 1-st ends at r = R-m, never interrupting.

The rule for determining the totality of possible sequences of EP is as follows. Suppose, for example, that on the r-1 st measure combinations of indices m _{k, r-2} and m _{k, r-1 are formed} , and on the r-st measure, combinations of indices m _kr-1 and m _{kr are formed} . If, at the same time, the index m _{k, r-1} generated by the r-1 st measure does not coincide with any of the indices m _{k, r-1} formed on the r-st step, then the sequence of electromagnets, a fragment of which included an alternative to EM, whose index on the r-1-st cycle, it was defined as m _{k, r-1} , it is excluded from further consideration (ie, by analogy with the terminology of the theory of Viterbi algorithm [13], the corresponding trajectory is not surviving). Hereinafter, the coincidence of the indices m _{k, r-1} and m _{k, r} means their coincidence only in part of the index (number) of the alternative to ES m.

If the index m _{k, r-1} generated by the r-1 st measure coincides with one of the indices m _{k, r-1} formed on the r-st measure, then a possible sequence of electromagnets, a fragment of which included an alternative to EMs whose index is on r- The 1st measure was defined as m _{k, r-1} , and is considered continued on the rth measure. At the same time, it (that is, this sequence) retains that number or index i, under which it was previously continued to the r-1st cycle due to the coincidence of the generated r-2-second cycle of index m _{k, r-2} with index m _{k, r-2} , formed by r-1-step. Moreover, the element-wise reception PCs calculated on the r-1st and r-th clocks corresponding to the indicated coincident indexes of EP alternatives are considered to correspond to the indicated possible sequence of EPs continued on the r-th clock, i.e. entering into it. The aforesaid applies to the corresponding indexes m _{k, r} .

) из индексов m_k,r-1, сформированных на r-м такте, то реализуется увеличение имевшего место до констатации указанного совпадения индексов количества возможных последовательностей ЭП на

, т.е. формирование

новых таких последовательностей с соответствующей их нумерацией. В остальном при

выполняются те же процедуры (только

), что и в ситуации, рассмотренной в предыдущем абзаце.If one of the r-1-th beat formed, the index m _r-1 coincides with more than one (i.e., with

) from indices m _{k, r-1} , formed on the rth measure, then an increase in the number of possible sequences of EPs occurring prior to stating the indicated coincidence is realized by

, i.e. the formation of

new such sequences with their corresponding numbering. Otherwise, when

the same procedures are performed (only

), as in the situation discussed in the previous paragraph.

Each index m _{k, r} that is included in the ith sequence of EPs (trajectory) is specified as m _{k, r, i} . In accordance with the text of the previous paragraph, the same index m _{k, r} (the corresponding alternative to EP) can be included in several sequences of EP. The set of possible sequences of EP in the message is the set of surviving trajectories, each i-th of which is characterized by an array of indices {m _{k, r, i} }.

The described procedure for determining the set of possible sequences of EP essentially coincides with the similar procedure performed to ensure the calculation of the formation of the PC reception array as a whole in the prototype (the description of this operation implemented in the prototype is absent in the literature, but it is in the prototype, as in any other a variant of the method of receiving the whole, key, without this operation its implementation is impossible). The only feature of this procedure in the claimed method is that as a result of performing operation 7 on each rth measure (this procedure is performed precisely on the results of performing operation 7 on a set of R measures), a number of EP sequences are excluded from consideration (deleted), i.e. . (in the mentioned terminology of the theory of Viterbi algorithm [13]) the survivors on each rth step are no more than K _r trajectories (hypotheses of EP sequences in the message). As a result of the specified procedure (the first phase of operation 10), arrays of possible sequences or surviving trajectories are formed, each i-th of which contains an array of indices {m _{k, r, i} }. The total number of possible sequences of EPs (it is designated below as I) or arrays of indices {m _{k, r, i} } is random.

In the second phase of operation 10, each i-th possible sequence of EPs or the corresponding array of indices {m _{k, r, i} } is converted to the corresponding i (this array) i-th stream of binary symbols. Moreover, in the case of Ω-bit information symbols, with respect to each number (index) i, the Ω-bit code of the symbol received at r = 1 clock is stored (this symbol corresponds to the index m _{r = 1, i} ; hereinafter, indices not bearing information k omitted); further to the right of it (that is, from the side of its highest order), the Ω-bit code of the symbol received at r = 2-nd step is “appended” (this symbol corresponds to the index m _{r = 2, i} ), etc ... and finally, the last one is similarly “added” to the Ω-bit code of a symbol received at r = 2-nd clock (this symbol corresponds to the index m _{r = R, i} ). As a result, I formed information sequences containing RΩ binary symbols each.

Along with the indicated set of information sequences, there is a binary checksum code corresponding to the index ν ₀ generated as a result of operation 9 (an implementation option of the proposed method is considered in which one solution is formed as a result of operation 9).

In the third phase of operation 10, each i-th of the I indicated possible informational sequences is checked for the presence of errors of element-wise reception of informational electronic signals (i.e., detection errors). To do this, in accordance with, for example, the CRC checksumming procedure, each i-th binary symbol stream is divided by the same generating polynomial that was used to perform the same procedure in step 2. As a result, each i-th binary stream is divided characters the corresponding i-th checksum is formed. The procedure for generating each checksum completely coincides with the similar procedure (which is part of operation 2), the description of which is omitted.

, соответствующего тому из индексов i контрольных сумм, при котором имеет совпадение этой (т.е. i-й суммы) с опорной контрольной суммой.In the fourth phase of operation 10, at first, the results of checking each i-th of the I indicated generated possible information sequences are formed for the presence of errors of element-wise reception of informational electronic signals. For this purpose, each of the indicated I checksums is compared with the binary code of the checksum generated as a result of the operation 9 (conditionally referred to as its reference). Finding the number of the totality of the results of the formation of the PC of the element-wise reception of informational electronic messages (i.e., one of the possible sequences (informational) electronic messages) that does not contain errors is carried out by determining the index

corresponding to that of the i checksum indices at which this (i.e., i-th sum) coincides with the reference checksum.

определяется путем сравнения каждой из указанных выше I контрольных сумм с каждой из K_R+1 опорных контрольных сумм. Вероятностью более чем однократного совпадения сравниваемых пар контрольных сумм можно пренебречь.In the embodiment of operation 9, it is similar to operation 7 (i.e., with the formation of the totality of the indices ν _k0 corresponding to K _{R + 1} > 1 with the highest level of element-by-element reception of excess ES) K _{R + 1} reference checksums are formed, and index

determined by comparing each of the above I checksums with each of the K _{R + 1} reference checksums. The probability of more than one coincidence of the compared pairs of checksums can be neglected.

возможной последовательности (информационных) ЭП сообщения.Operation 11 (deciding on the fidelity of one of the hypotheses of sequences of informational EPs) involves the selection and broadcasting of the output method

possible sequence of (informational) electronic messages.

All operations of the proposed method are implemented by universal programmable microprocessors. It is possible to combine the implementation of several operations of the method in one microprocessor.

The inventive method, like the prototype, is designed for use in a synchronous communication system. In such a system, at the receiving end, the moments of the beginning of arrival of each ES are known. In principle, it is possible, for example, the operation of the transmitting and receiving parts of the transmission system in a single time system. Moreover, with regard to the synchronization of the operation of devices that implement signal processing operations at the receiving end, the propagation time of the signal from the transmitter to the receiver is known, and the equipment that implements the reception operations includes a timer that generates (generates) clock signals that control the execution of all receiving operations.

The set of synchronization operations is not included in the composition of the claimed object, since the vast majority of digital (discrete) communication systems are synchronous, and the features of the claimed object are not associated with any specifics of the specified set of operations.

The above description of the totality of the features of the proposed method actually contains a description of its work in dynamics.

The technical effect in the claimed method (increasing the noise immunity of a transmission) by making a choice (making an appropriate decision) about the fidelity of one of the hypotheses of sequences of information ESs based on the identification of one of the totality of possible sequences of ESs that (in accordance with the result of the application of the noiseless coding procedure, for example, CRC procedures) does not contain detection errors (i.e. errors made by element-wise decisions). Data transmission errors in this case occur for one of the following three reasons:

- among the set of possible sequences of EPs, that set that actually does not contain detection errors is absent (i.e., it was mistakenly lost in the process of selecting K _{r the} highest bit-wise reception by level PC during the implementation of operation 7 at least on one IVPS cycle);

- there is an erroneous determination of the index z _ν0 (and, as a result, an erroneous formation of the reference checksum) during operation 9;

- at the same time, errors are specified in the two hyphens above, and as a result, one of the I checksums coincides with the reference checksum.

The problem with errors caused by the event indicated in the first hyphen is largely overcome by increasing the values of K _r . The problem with errors caused by the event indicated in the second hyphen is largely overcome by applying error-correcting coding when generating excessive ESs (such a “protection” of the totality of all informational ESs, which are relatively numerous, will lead to a significant loss in transmission speed; one for the whole message). Probability of an event specified in the third hyphen equal to 2 ^-p, is negligible (for example, at p = 20 or 32, respectively, said probability is ≈10 ^-6 or ≈10 ^-10). Similar resources listed to reduce the likelihood of errors in the prototype are missing, which is due to the achievement of the technical effect.

About the criterion of "inventive step". The inventive object is fundamentally based on methods for receiving in general and noise-resistant coding with error detection, separately known. At the same time, the method of error-correcting coding with error detection in known objects is implemented with respect to a single sequence of information ES and as a result gives a decision only on whether there are errors in this sequence or not. In the inventive method, the set of error-correcting coding operations with error detection is implemented on the receiving side of the transmission system with respect to the set of I> 1 sequences of information ES, which ultimately provides a significant reduction in the probability of errors. Those. in the inventive method, the set of error-correcting coding operations with error detection actually performs a function equivalent to error-correcting coding functions with error correction. In this case, the error-correcting coding procedure with error correction is characterized by a deliberately large required number of redundant EFs (i.e., it reduces the transmission rate by a greater extent) than the error-correcting coding procedure with error detection. Thus, the combination of two well-known technical solutions used in the inventive object is associated, firstly, with a new feature (implementation of checking for errors in each of the set of I> 1 sequences of informational ES), and secondly, with a new effect.

Possible implementations of the proposed method, providing, in particular, the implementation of an arbitrary set of receiving operations as a whole. All of them should be considered as equivalents of the above described variant of this proposed method.

Used sources

1. Fink L.M. Theory of discrete message transmission. - M .: Owls. radio. 1970 .-- 728 pp., Ill.

2. Klovsky D. D., Shirokov S. M. A method for optimal reception of discrete messages in general. A.S. USSR No. 930696.

3. Zyuko A.G., Klovsky D.D., Korzhik V.I., Nazarov M.V. Theory of electrical communication. - M .: Radio and communication. 1999 .-- 432 p.: Ill.

4. Alyshev Yu.V., Borisenkov A.V .. Klovsky D. D., Nikolaev B. I. Digital processing of signals during soft decoding in channels with multipath and code symbol interleaving // 4th International Conference DSPA-2002 (http://www.autex.spb.ru/download/dsp/dspa/dspa2002/tom1_03.pdf).

5. Klovsky D. D. Transmission of discrete messages over the air. - M .: Radio and communication. 1982. - 304 p.: Ill.

6. Nikolaev R.P., Popov A.R. A method of transmitting information in a communication system with noise-like signals. RF patent No. 2286017.

7. Ozerov I.A., Ozerov S.I. A method of transmitting information in communication systems with noise-like signals and a software product. RF patent No. 2277760.

8. The CRC checksum algorithm (http://all-ht.ru/inf/systems/p_0_13.html).

9. Ross N. Williams. A basic guide to CRC error detection algorithms. (http://www.info-system.ru/library/algo/crc1.pdf).

10. Peterson W., Weldon E. Codes corrective errors. Per. from English - M .: World. 1976 .-- 595 p.: Ill.

11. Best CRC Polynomials. Koopman P, Carnegie Mellon University. (https: //users.ece.cmu. edu / ~ koopman / crc /).

12. Smaryshev M.D., Dobrovolsky Yu.Yu. Hydroacoustic antennas. Directory. - L .: Shipbuilding. 1984.

13. Sergienko A.B. Digital communication. Tutorial. - SPb .: SPbGETU "LETI", 2012, 164 pp., Ill.

Claims (18)

A method for transmitting and receiving digital information in general, in accordance with which on the transmitting side - convert the sequence of information symbols to be transmitted into a stream of information elementary premises (EP), - convert the resulting flux of ES into fluctuations in the hydroacoustic or electromagnetic field, on the receiving side - convert the received vibrations of the hydroacoustic or electromagnetic field into electrical signals, - according to the signal observed at the time interval of the reception of the message (IVPS) in relation to each r-th information message ES (r = 1 ... R is the serial number of the information message in the message), the decisive statistics (PC) of the element-wise reception of information message information are calculated, - for each of the PCs of the element-wise reception of informational electronic devices, element-wise decision-making is implemented, - perform the operation of deciding on the fidelity of one of the hypotheses of sequences of informational ES, characterized in that on the transmitting side - convert the sequence of information symbols to be transmitted into a stream (set) of excess EPs, - combine the flows of information and excess ES in the final stream of ES, on the receiving side - according to the signal observed at the IVPS as applied to each k-th redundant ES of the message (k = 1 ... K is the serial number of the redundant EL in the message), the element-by-element PC of receiving the redundant EL is calculated, - for each of the PCs of the element-wise reception of excess EPs, element-wise decision-making is implemented, - from the totals of the decision-making results for each of the PCs of the element-wise reception of a sequence of informational electronic components and the decisions for each of the PCs of the element-wise reception of a sequence of redundant electronic devices, find the number of the set of results of the formation of the PC of the element-wise reception of informational electronic components that does not contain errors, - the operation of deciding on the fidelity of one of the hypotheses of sequences of informational EPs is performed on the basis of the number of the totality of the results of the formation of the PC of the bitwise reception that does not contain errors.

Images