RU2789785C1 - Method for compression interference-resistant data encoding for transmitting and storing information - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: invention relates to telecommunication systems and computing technology. Claimed method for compression interference-resistant data encoding for transmitting and storing information consists in additionally, on the transmitting side, combining successive ternary code pulses with durations T₀ and T₁=1.5 T₀, as well as T₀ and T₂=2T₀, occurring in the PWM₃ signal sequence, into longer pulses with total time intervals 2.5 T₀ and 3T₀, respectively, while the other PWM₃ signals remain in the same succession order; using the sequence of signals with additional durations 2.5 T₀ and 3T₀, formed as a result of the combinations, for phase modulation of the carrier frequency signal with phase values 0° and 180° when inverting the polarities of the PWM₃ pulse signals; on the receiving side, determining the durations of the PWM₃ restored during demodulation and matching with the following code constructions (CC_i, i=0, 1, 2, 3, 4): T₀ ↔ {<00>₂,<11>₂}, T₁ ↔ {<10>₂,<001>₂} and T₂ ↔ <101>₂, T₃ ↔ {<110>₂,<0001>₂} and T₄ ↔ <1101>₂.

EFFECT: higher interference resistance of transmitted and archived information and error detection.

1 cl, 9 dwg

Description

The invention relates to telecommunications systems and computer technology. It can be used in data transmission systems over communication channels to resolve the contradiction between the required high rates of information transfer and the limited bandwidth of communication channels, which depends, among other things, on the intensity of interference of various origins on transmitted messages. Also, the invention can be used to store large amounts of transmitted and received information with their protection against errors that appear, for example, as a result of exposure to cosmic radiation on radio-electronic equipment.

Its use allows:

1) improve the noise immunity of transmitted and archived information without introducing structural redundancy into transmitted messages and data compression;

2) detect errors arising during transmission and storage of data, both single and multiple;

3) increase, if necessary, the speed of information transfer and provide indicators of its energy and structural secrecy.

The invention defines a new direction in the theory of transmission and archiving of information, which is called: "noise-immune data compression".

It is based on the developed innovative technologies in the form of economical noise-immune coding, the origins of which are determined by patents RU No. [7], 2658795 [8], 2672392 [9].

Currently, “noise-immune data compression” is focused on the following three main components that determine such stages of formation, transmission and reception of information as:

economical representation, low-redundancy and non-redundant coding of digital signals, data and messages using residual images and equivalent structural-algorithmic transformations (SAT) (patents RU No. 2586605 [4], 2586833 [5], 2609747 [6], 2649291 [ 7]);

additional coding of information with the replacement of data streams represented by a binary code, including after the primary economical noise-immune SAP, with their equivalent analogue in the form of a logical noise-correct code using two-dimensional ternary symbols S _i (T _i ), where i=0,1,2 , repeating each other in terms of transmitted information with decodings in the form: S ₀ (T ₀ ) ↔ {<00> ₂ ,<11> ₂ }, S ₁ (T ₁ ) ↔ {<00> ₂ ,<11> ₂ } and S ₂ (T ₂ ) ↔ <101> ₂ (here S _i - determine during primary modulation three allowed states i=0,1,2 pulse amplitudes, where T _i are three values of T ₀ , T ₁ =1, 5T ₀ and T ₂ =2T ₀ their duration in time, and T ₀ is the duration of the binary symbols "1" and "0" of the original stream (patents RU No. 475861 [1], 2480840 [2], 2581774 [3]) ;

processing of recovered information with detection and correction of message transmission errors (patent RU No. 2658795 [7]) and adaptive nonlinear filtering of the received data (patent RU No. 2672392 [8]).

Known are “Methods for transmitting information and systems for their implementation” (patent RU No. 2475861 with priority dated April 27, 2013 [1] and patent RU No. 2581774 with priority dated September 30, 2014 [3]).

The patent [1] shows the sequence of operations by which the encoding algorithm is implemented, and also presents one of the possible options for implementing the patent based on logic circuits. The main objective of the patent [1] was to show the possibility of a fairly simple implementation of the possibility of switching from a binary code to the proposed ternary coding using duplicate ternary symbols S ₂ (T ₂ ), S ₁ (T ₁ ) and S ₀ (T ₀ ) based on binary logic, which is used in the existing element base.

The disadvantage of the invention [1] lies in the fact that the potential of the proposed ternary coding has not been fully disclosed. In particular, the new properties of the proposed ternary code have not been demonstrated in terms of providing noise-resistant data compression.

In the patent [3], which was awarded a diploma of Rospatent in the nomination "100 best inventions of Russia - 2016", the proposed ternary code with the symbols S ₂ (T ₂ ), S ₁ (T ₁ ) and S ₀ (T ₀ ) was used for increasing the information load of the carrier frequency emitted by the transmitter, for which complex modulation in amplitude, frequency and phase was used. In addition, “comparability of the results of demodulation of relative phase modulation (RPM) with base two, applied over the main communication channel, and the proposed ternary phase modulation with base two (PM ₂ ³ ), which is based on logical error-correcting codes with base three, presented by symbols S ₂ (T ₂ ), S ₁ (T ₁ ) and S ₀ (T ₀ )” (clause 4 of the claims [3]). However, to implement this possibility, an additional (duplicating) information transmission channel is required, which cannot always be provided in a variety of practical applications.

The essence of the proposed logical error-correcting coding of the generated messages with a replacement ternary code, as in patents-analogues [1-3], is as follows.

The method is based on the transformation formulas F associated with the replacement of the sequence of characters a _i0 , a _i1 , ..., a _in the binary alphabet A={0,1} with sequences of characters d _i0 , d _i1 , ..., d _{im of} the alphabet D={00, 11 , 10, 001, 101} based on the following logical coding schemes:

The proposed encoding establishes a logical correspondence between the binary symbols presented in brackets and their ternary equivalents S ₀ , S ₁ and S ₂ .

In addition, when receiving signals S ₀ , S ₁ and S ₂ provide their duplication of signals T ₀ , T ₁ and T ₂ (Fig. 1 (B)). As a result, two modulating sequences are formed based on the signals S ₀ , Si, S ₂ and the signals T ₀ , T ₁ , T ₂ . Moreover, if the signals S ₀ , S ₁ , S ₂ are presented in the form of pulse amplitude modulation into three states (AIM ₃ ), then the corresponding signals T ₀ , T ₁ , T ₂ are displayed in the form of pulse width modulation (PWM ₃ ), which also has three allowed positions of the pulse duration T ₀ , T ₁ =1.5T ₀ , T ₂ =2T ₀ , where T ₀ is the time duration of the symbols "1" and "0" of the source binary code. As a result, a stream of digital data generated for transmission, displayed initially by the sequences of symbols "1" and "0" of the binary code, after recoding into a logical error-correcting ternary code with the symbols S ₂ (T ₂ ), S ₁ (T ₁ ) and S ₀ ( T ₀ ) are represented by two redundant streams at the level of primary (pulse) modulation: AIM ₃ and PWM ₃ . The essential characteristics of the method [3] also consist in providing duplication of logical error-correcting coding with ternary symbols S ₂ (T ₂ ), S ₁ (T ₁ ) and S ₀ (T ₀ ), associated with the primary pulse modulation of the transmitted signals, which forms the basis of the synthesis various problem-oriented structural-code (StKK) and signal-code constructions (SiKK) [1-3].

The proposed algorithm for recoding the original binary code with the symbols "1" and "0" is as follows. Given a sequence of binary characters <10101000111100101> ₂ (N ₁ =17). It is necessary to translate it into a new alphabetic coding using upper and lower partitions (K1 and K2):

K1, K2: S ₂ S ₂ S ₁ S ₀ S ₁ S ₀ S ₀ S ₀ S ₁ S ₁ S ₂ is a new sequence of 11 characters S _i (i=0,1,2) used to transmit information.

As a result of encoding, the last binary symbol of the previous encoded message S _i repeats the first binary symbol of the subsequent ternary message S _i+1 (repetitions of binary symbols are highlighted in bold and circled). If these conditions are not met when receiving, then this indicates an error. Error correction is provided with reliable reception of symbols S ₂ , which are unambiguously decoded S ₂ ⇔ (101). In FIG. 1 shows an example of the proposed transition from a 16-bit binary code word X _j =<1011100010100101> ₂ , corresponding to the decimal representation of X _j =<788> ₁₀ , to the proposed ternary code with duplicate symbols S _i (T _i ), i=0.1 ,2. It follows from the above illustration that the number of symbols S _i (T _i ), i=0.1.2 is reduced by 1.6 times compared with the original binary symbols "1" and "0", having a duration of T ₀ . This corresponds to the theoretical value determined by the following analytical representation: k=log ₂ 3=1.6. Another example of translating into the proposed ternary code the value X _j+1 =<1010101100001101> ₂ corresponding to the decimal representation of X _j+1 =<353> ₁₀ (Fig. 2 ).

Known analogous methods use the following algorithms for recovering transmitted information. Let us assume that as a result of demodulation the following fragment of the sequence of ternary signals S _i is restored:

1 The first operation of restoring data in the original binary code is to extract signals S ₂ that can be unambiguously decoded: S ₂ ↔ <101”. As a result, in the received sequence (2) a sign of receiving the symbol S ₂ is formed, which is allocated to control the reliability of the reception of the symbols S ₁ in the time interval between adjacent symbols S ₂ :

2 The conversion to binary code of the entire sequence of received ternary symbols begins with the fact that the result of unambiguous decoding of symbols S ₂ is written in binary code:

To control the reliability of the reception of symbols S ₁ in the time interval between adjacent symbols S ₂ use the following rule: the number of symbols S ₁ must be even.

3 Then, using the first of the received signs of receiving the symbol S ₂ , they begin to decrypt the following ternary symbol S ₁ , which is assigned not one, but two code combinations made up of binary symbols: S ₁ ↔ "10" and "001 ". Since, according to the condition for the formation of the proposed ternary code, the last binary symbol of the previous decryption 101 must be the first symbol of the subsequent decryption of the ternary symbol (in the given reception of signal S ₁ ), code combination 10 should be chosen as a candidate for replacement. Selection of the appropriate next code combination , corresponding to the received signal S ₁ , is determined by the last binary character obtained as a result of decoding the previous signal of the ternary code S ₂ (matching characters are in bold).

Further, the next restored ternary symbol corresponds to the signal S ₀ , which also has two possible decoding options: S ₀ ↔ "00" and "11". Since the previous decoding 10 ended with the binary character "0", then the next signal S ₀ must be replaced with a binary code combination consisting of two characters "0" - 00. Similarly, the next restored ternary symbol S ₁ ↔ "10" and "001" must be replaced by binary codeword 001. Then two characters S ₀ ↔ "00" and "11" follow in a row, and since the decryption of the previous binary codeword ended with the character "1", then the next two ternary characters So must be replaced to a sequence consisting of two characters "1" - 11. The fact that the decoding was done correctly is evidenced by the fact that the next ternary character will again be S ₂ ↔ "101", which begins with the same binary character "1", which received at the end of the previous decoding of the signal S ₀ . The subsequent process of restoring transmitted messages in traditional binary is similar. The correcting ability of the proposed code lies in the fact that the entire chain of ternary symbols restored upon reception is determined by the reference signals S ₂ ↔ "101" (Fig. 1 and Fig. 2).

4 The next operation assumes that the received sequence of ternary symbols S ₁ and S ₀ , allowing ambiguous interpretations

replaced with a binary code so that the last binary character of the decoding of the preceding character S _i , where i=0,1,2, matches the first character of the decoding of the subsequent character S ₁ or S ₀ .

As a result, for the sequence (6), the following restored sequence of the original binary code is obtained:

where "dots" separate the results of decoding the symbols S ₂ , S ₁ and S ₀ of the ternary code.

Then the matching binary symbols on the boundaries of the decodings of the ternary signals S _i , where i=0,1,2, are combined and replaced with one corresponding binary symbol:

where matched characters are highlighted and underlined, which combine and replace with a single binary character.

As a result, the following fragment of the original binary code sequence will be restored: 10100011101.

Currently, the problems of increasing the efficiency of data transmission over high-speed radio links, when the required volumes of transmitted information exceed the bandwidth of the communication channel, are most acutely identified. Also, one of the objectives of this invention is to substantiate technical solutions, which are based on non-traditional methods of noise-immune compression and their extended application to resolve existing contradictions in the development of info-telecommunication systems (ITCS). Of particular importance is the use of the proposed invention, both for the transmission of information and for its storage. The inability to transmit large amounts of information from spacecraft (SC) in a short time of their communication session leads to the need to store a significant part of it in on-board storage devices, where they must wait for their turn in a long transmission process. However, in contrast to the conditions of information storage in terrestrial computers, hard cosmic radiation is accompanied by the appearance of a large number of errors in information storage. This is observed in many cases, the characteristic features of which are manifested in the fact that each of the transmitted pieces of information stored on board and waiting for its turn, when it is received, becomes progressively worse in quality and reliability indicators compared to previously received data.

This invention is aimed at eliminating the noted shortcomings associated with insufficient information transfer rate, which is limited by the bandwidth of the radio channel, and the need to eliminate internal errors that appear in the radio-electronic equipment (REA) itself, for example, in data storage units, under the influence of hard cosmic radiation, this invention is directed.

At the same time, structural-algorithmic transformations (SAT) of data, messages and digital signals are used as a preliminary first stage of compressed noise-correcting coding (RF patents No. 2586605 [4], 2586833 [5], 2609747 [6]). They can be considered as an additional economical coding carried out in relation to the original binary code. An example of such additional coding at the level of binary codes is shown in FIG. 3. Its essence lies in the fact that in order to provide an additional possibility of detecting and correcting errors in transmitted information with internal redundancy, the simplest SAP-1(p) operation was used. In terms of the possibility of increasing the noise immunity of transmitted data with internal redundancy (correlation dependence of neighboring values), the structural-algorithmic transformation SAP-1(p) is a substitute for the cognitive (knowledge-generating) model SAP-1(o) [4]. The latter involves dividing the values of the initial data or messages X _i , represented by N=2n - bit binary code, into the selected optimal comparison modules m ₁ =2 ⁿ -1 and m ₂ = 2 ⁿ +1 with obtaining residual images b _1i (mod m ₁ ) and b _2i (mod m ₂ ). In this case, each of the obtained residual images b _1i (mod m ₁ ) and b _2i (mod m ₂ ) requires n-bit binary codes to represent its values, which is half the original bit depth N. Therefore, the encoding result C _i =<b _1i (mod m ₁ ); b _2i (mod m ₂ )> ₂ assumes the use of the same bit grid (N) for binary representation.

(фиг. 4) при «мягком» декодировании слов и ли сообщений приведут к возможности обнаружения и исправления ошибок (фиг. 4 (средний график)) на основе выделяемого из переданных данных инварианта в виде свойства «равноостаточности» [4-8]. Для сравнения на фиг. 4 (нижний график) представлен результат восстановления переданных данных телеизмерений (ТИ) при использовании «жесткого» алгоритма декодирования без исправления ошибок. Но его использование, несмотря на значительные искажения ΔC_i восстановленного телеметрируемого параметра (ТМП), позволяет установить временные участки, на которых выполняется условие корреляционной зависимости, что необходимо для управления работой «мягкого» декодера. При использовании алгоритма «мягкого» декодера было исправлено около 94% ошибок. Оставшиеся 6% ошибок (фиг. 4 (средний график)) - это ошибки второго рода. Их появление связано с тем, что результаты ТИ, искаженные помехой, случайно оказались на месте разрешенных позиций помехоустойчивого кода. Они также могут быть исправлены при предварительной обработке. При использовании патента [8] этот дополнительный технический результат получают за счет повторного применения на приемной стороне (уже при обработке) алгоритма дополнительного кодирования САП-1(п), но с минимальным кодовым расстоянием d_min ⁽²⁾, которое отличается от ранее использовавшегося при передаче информации - d_min ⁽¹⁾:d_min ⁽²⁾ ≠ d_min ⁽¹⁾.The use of the SAP-1(p) replacement model differs in that it is the simplest in terms of its technical implementation in high-speed data transmission systems (HTS). It involves the division of the original code word or message X _i into the highest ( a _sti ) and the youngest ( a _ml ) half-words, followed by rearrangement of their places (Fig. 3) [6]. Then, in order to implement the mode of "hard" decoding of words or messages when receiving or restoring, it is necessary to return the half-words to their places (Fig. 4). In this case, to detect errors, it is necessary to use the “soft” decoding mode, the operability of which is ensured if there is a correlation between adjacent transmitted or stored values of words or messages. When using it, the results of noise-proof coding C _i , including those distorted by noise ΔC _i :

(Fig. 4) with "soft" decoding of words or messages will lead to the possibility of detecting and correcting errors (Fig. 4 (middle graph)) based on the invariant extracted from the transmitted data in the form of the "equal residual" property [4-8]. For comparison, in Fig. Figure 4 (lower graph) shows the result of reconstructing the transmitted telemetry data (TI) using a "hard" decoding algorithm without error correction. But its use, despite significant distortions ΔC _i of the recovered telemetered parameter (TMP), allows you to set the time intervals on which the condition of the correlation dependence is met, which is necessary to control the operation of the "soft" decoder. When using the "soft" decoder algorithm, about 94% of errors were corrected. The remaining 6% of the errors (FIG. 4 (middle plot)) are Type II errors. Their appearance is due to the fact that the results of the TI, distorted by the noise, accidentally ended up in the place of the allowed positions of the noise-correcting code. They can also be corrected by pre-processing. When using the patent [8], this additional technical result is obtained due to the repeated application on the receiving side (already during processing) of the additional coding algorithm SAP-1(p), but with a minimum code distance d _min ⁽²⁾ , which differs from that previously used when information transfer - d _min ⁽¹⁾ :d _min ⁽²⁾ ≠ d _min ⁽¹⁾ .

The "hard" decoding algorithm for words or messages is universal: its performance does not depend on the correlation dependence of neighboring values. In this case, in the general case, a “soft” decoder is connected to work only when, during the “hard” decoding of words or messages, the presence of correlation dependence properties of neighboring words and messages is established, even if it is local, performed for a limited number (i) of recovered data: i≥4.

The proposed ternary code with the symbols S _i (T _i ), i=0,1,2, the transition to which is the next structural-algorithmic transformation (SAP-2), is one of the universal ones: it provides an opportunity to increase the efficiency of the SPT and, in general, when there is no correlation between neighboring values.

The essential characteristics of the proposed "Method of noise-immune data compression for transmission and storage of information" are as follows.

It involves the following sequence of operations: code words and messages to be transmitted or stored are represented by a positional binary code with the symbols "1" and "0" having a duration T ₀ , from which a digital group signal (DTS) is formed in the form of a sequence of bits, which is then subjected to structural-algorithmic transformations of the first stage (SAP-1), the simplest technical implementation of which is associated with the division of the original code words or messages X _i into high ( a _sti ) and low ( a _{m i} ) half-words with subsequent rearrangement of their places, as a result after which a new DTS stream is formed, which has the properties of increased noise immunity in the presence of a correlation between neighboring transmitted or stored values of code words or messages, after which they proceed to the second stage of the SAP (SAP-2) based on the transformation of the new bit sequence into a replacement ternary error-correcting code with symbols S _i (T _i ), i=0,1,2, which forms based on the accepted logical coding rule that establishes the following correspondences "↔", for example, S ₀ (T ₀ ) ↔ {<00> ₂ ,<11> ₂ }, S ₁ (T ₁ ) ↔ {<10> ₂ ,<001 > ₂ } and S ₂ (T ₂ ) ↔ <101> ₂ , each of which has three encoding symbols conventionally defined as "S ₀ " and "T ₀ ", "S ₁ "and "T ₁ ""S ₂ ” and “T ₂ ”, while the symbols “S ₀ ”, “S ₁ ” and “S ₂ ”, are represented in the form of amplitude-pulse modulation (AIM ₃ ) with three allowed code positions, and the second modulating component of the video signal with the symbols of the tribasic code "T ₀ ", "T ₁ " and "T ₂ " with durations T ₀ , T ₁ \u003d 1.5T ₀ and T ₂ \u003d 2T ₀ , where T ₀ is the duration of one symbol of the source binary code, is used for pulse-width modulation (PWM ₃ ) of the second video signal subject to further structural-algorithmic transformation (SAP-3) of the third stage.

The invention is focused on the next third conditional stage of the distributed structural-algorithmic transformation (SAP-3). Its task is to combine into a single pulse signal the previously generated pulse signals PWM ₃ , following in such a sequence as: 1) T ₀ and T ₁ \u003d 1.5T ₀ with the formation of a new duration T ₄ \u003d 2.5T ₀ and 2) T ₀ and T ₂ =2T ₀ , resulting in a combined pulse signal duration of 3T ₀ . A block diagram of a device designed for such a combination is shown in Fig. 5. Illustrations of the types of new PWM signals ₅ , consisting of the previously used PWM symbols ₃ of the ternary code T ₀ , Ti=1.5T ₀ and T ₂ =2T ₀ and supplemented by two new pulses with durations T ₄ =2.5T ₀ and T ₅ =3T ₀ are shown in FIG. 6, fig. 7, fig. 8 and FIG. 9. As a result of such an addition, the new code will no longer be 3-position (M=5), but 5-position (M=5). But when using phase (PM) and relative phase modulation (RPM), during transmission, it requires a change in the form of only two states of the phase of the carrier frequency: 0 ° and 180 °, which compares favorably with its counterparts.

Completed preparatory description allows you to proceed to the formulation of the claims.

A method for compressed error-correcting data coding for transmitting and storing information, which consists in performing the following sequence of operations: code words and messages to be transmitted or stored are represented by a positional binary code with symbols "1" and "0" having a duration of T ₀ , from which digital group signal (CTS) in the form of a sequence of bits, which is then subjected to structural-algorithmic transformations of the first stage (SAP-1), the simplest technical implementation of which is associated with the separation of the original code words or messages into the highest ( a _st ) and the youngest ( a _ml ) half-words with their subsequent rearrangement, as a result of which a new DTS stream is formed, which has the properties of increased noise immunity in the presence of a correlation between adjacent transmitted or stored values of code words or messages, after which they proceed to the second stage of SAP (SAP-2) based on the transformation of the new bit sequences in replacement sparing ternary error-correcting code with symbols S _i (T _i ), i=0,1,2, which is formed on the basis of the accepted logical coding rule that establishes the following correspondences "↔", for example, S ₀ (T ₀ ) {<00> ₂ ,<11> ₂ }, S ₁ (T ₁ ) ↔ {<10> ₂ ,<001> ₂ } and S ₂ (T ₂ ) ↔ <101> ₂ , each of which has three encoding symbols, conventionally defined as "S ₀ " and "T ₀ ", "S ₁ " and "T ₁ ""S ₂ " and "T ₂ ", while the characters "S ₀ ", "S ₁ " and "S ₂ " are represented as amplitude-pulse modulation (AIM ₃ ) with three allowed code positions, and the second modulating component of the video signal with the symbols of the tribasic code "T ₀ ", "T ₁ " and "T ₂ " with durations T ₀ , T ₁ =1.5T ₀ and T ₂ =2T ₀ , where T ₀ is the duration of one symbol of the original binary code, used for pulse width modulation (PWM ₃ ). It differs from analogues in that on the transmitting side, the following successive pulses of the ternary code with durations T ₀ and T ₁ \u003d 1.5T ₀ , as well as T ₀ and T ₂ - 2T ₀ , occurring in the sequence of _PWM signals 3, are combined into longer pulses with total time intervals 2.5T ₀ and 3T ₀ , respectively, leaving other PWM signals ₃ in the same sequence, formed as a result of the performed combinations, a sequence of signals with additional durations 2.5T ₀ and 3T ₀ is used for phase modulation of the carrier frequency signal with the values phases 0 ° and 180 ° when changing the polarity of the PWM ₃ pulse signals, as a result of which the intensity of the phase change of the carrier frequency signal is further reduced, providing the possibility of increasing the speeds and volumes of transmitted information, on the receiving side, the durations of the restored during demodulation are determined and the following code codes are assigned to them constructions (CC _i , i=0,1,2,3,4): T ₀ ↔ {<00> ₂ ,<11> ₂ }, T ₁ ↔ {<10> ₂ ,<001> ₂ } and T ₂ ↔ <101> ₂ used previously, as well as new decodings: T ₃ ↔ {<110> ₂ ,<0001> ₂ } and T ₄ ↔ <1101> ₂ , retaining the same properties recurrent code, according to which the last binary character "1" or "0" of the previous decryption is the first in the decryption of the next code structure (CC _i , i=0,1,2,3,4).

In FIG. 5 shows a device that implements the proposed method on the transmitting side. In this case, the following designations are used: 1 - converter of PWM durations ₃ into digital form; 2 - clock signal generator; 3 - counting pulse generator (GSI); 4 - PWM duration selector ₃ ; 5, 10, 11, 12 and 13 - delay lines for time T ₀ ; 6 - pulse expander T ₀ PWM ₃ 1.5 times; 7 - pulse expander T ₀ PWM ₃ 2 times; 8, 9 - logical elements "AND"; 15 - logical element "Ban" with two prohibiting inputs; 14, 16 - logical elements "Prohibition" with one inhibiting input; 17 - logical element "OR".

The operation of the device is as follows. The generated sequence of PWM pulses ₃ is fed to the input of the Converter 1 duration PWM ₃ in digital form to the input of the shaper 2 clock signals. The clock pulses generated in 2 start the generator of 3 counting pulses (GSI). In the converter 1 of the PWM durations ₃ they are counted, as a result of which the durations are converted into a digital amplitude-pulse form. In the selector 4 durations of PWM ₃ , the obtained values are compared with the set thresholds and, as a result of comparison with them, the PWM pulses ₃ are divided into signals T ₀ (output 24), T ₁ \u003d 1.5T ₀ (output 22) and T ₂ \u003d 2T ₀ (output 23). PWM pulses ₃ with a duration T ₀ (output 22) are delayed for a time T ₀ in the delay line 5 and are fed to the corresponding expanders 6 and 7 of the pulse duration T ₀ PWM ₃ 1.5 and 2 times, respectively. When they coincide in the logic elements "AND" with the original PWM pulses ₃ with the appropriate durations T ₁ =1.5T ₀ and T ₂ =2T ₀ form PWM pulses with additional durations T ₃ =2.5T ₀ and T ₄ =3T ₀ . Then they are delayed in the delay lines for a time T ₀ , and the generated signals are prohibited by means of the logic element "Prohibition" 15 with two prohibiting inputs, the passage of signals of PWM pulses ₃ with durations T ₀ to the input of the logic element "OR" 17. Also additionally generated PWM pulses with durations T ₃ =2.5T ₀ and T ₄ =3T ₀ are prohibited using the logic elements "Prohibition" 14 and 16 passage to the input of the logic element "OR" 17 PWM pulses ₃ with the corresponding durations T ₁ =1.5 T ₀ and T ₂ =2T ₀ the appearance of which was preceded by PWM pulses ₃ with durations T ₀ . As a result, the PWM ₃ pulse sequence of the upgraded ternary code is restored in the form of PWM ₅ .

When receiving new signals, it is determined on the basis of operations similar to the mathematical operation of differentiating the pulse sequence, the boundaries of the PWM ₅ pulses, as well as their duration in time, which should take on the values T ₀ , T ₁ =1.5T ₀ , T ₂ =2T ₀ , T ₄ \u003d 2.5T ₀ and T ₅ \u003d 3T ₀ . The obtained values, corresponding to the duration of the PWM pulses ₅ , put in line with their following decoding: T ₀ ↔ <11.00>₂; T ₁ \u003d 1.5T ₀ ↔ <11.001>₂; T ₂ =2T ₀ ↔ <101> ₂ that was before. And for the new durations of PWM ₅ , the correspondence system will be as follows: T ₃ \u003d 2.5T ₀ ↔ <110, 0001> ₂ and T ₄ \u003d 3T ₀ ↔ <1101> ₂ . As a result of this, the original property of the recurrent code will be preserved, in which the last binary symbol of the previous decryption repeats the first binary symbol of the subsequent decryption.

As an example, consider the option of receiving the previously used ternary code (PWM ₃ ), the symbol sequence of which is shown in Fig. 6:

As a result, we get the following restored message:

X _i =<1011100010100101> ₂ =<788> ₁₀ .

As a result, we also get the following restored message:

X _i =<1011100010100101> ₂ =<788> ₁₀ .

The gain is as follows. When encoding in binary code, to transmit the value X _i =<788> ₁₀ , it was necessary to transmit 16 binary characters. When replacing a binary code with a ternary one, the required number of ternary symbols is 10 (Table 1). The syntactic compression ratio at the level of code symbols is equal to: k ₃ =log ₂ 3=1.6.

When switching to the proposed extended PWM code ₅ , the number of symbols, which is equal to the number of carrier frequency phase changes, takes the value 8 (Table 2). This means that the limit k ₅ =2, determined by the lossless data compression theory [10,11,12,14], has been reached. In another case, shown in FIG. 7, to transmit information of the same amount (16 binary symbols), only 7 PWM symbols ₅ will be required and, therefore, the limit set by the theory of lossless data compression will be exceeded: k ₅ \u003d 2.285, which is more than 2.

In addition, it becomes possible to detect and correct information transmission errors if the algorithms of the previous structural-algorithmic transformation (SAP-1) were used on the transmitting side, which preceded the conversion of the binary code into ternary.

As a result, the resulting sequence of symbols of the extended ternary code with symbols (T _i ), i=0,1,2,3,4, which is PWM ₅ , is converted as a result of an additional decoding operation into a binary code with symbols "1" and "0" (Table 2). The binary code recovered during decoding is subjected to the next recovery operation based on the reverse SAP-1 (OSAP-1) algorithms using the universal "hard" decoding mode, which provides unconditional restoration of information in its original form without correcting transmission or storage errors. Under its control, a “soft” decoding algorithm operates, which allows detecting and correcting errors in the presence of a correlation dependence that manifests itself in relation to successive code words or messages when their number is more than three.

As a conclusion, allowing to sum up the novelty and essential characteristics of the proposed invention, the following should be noted.

The use of the proposed invention greatly simplifies the procedures for phasing (synchronization [13]) of the transmitted information streams during its diversity reception, since a number of new enlarged information features appear that can be used to identify the alignment of information streams received, for example, by different measuring points having different geographical coordinates accommodation. There are only 2 such signs in binary coding - these are the recoverable characters "1" and "0". With ternary coding, there are 3 of them, represented as three PWM pulse widths ₃ . When using the proposed compressed error-correcting coding, there are already 5 of them. This significantly simplifies the operation of combining streams of the same information that is received with different time delays.

The simulation of operations that form the basis of the proposed method in relation to data compression, which was carried out to ensure comparability with existing methods of their archiving used in computer technology, showed the possibility of achieving higher rates of information redundancy reduction at the syntactic level compared to known technical solutions. At the same time, the possibility of further improving a number of other indicators of the effectiveness of the SPT and increasing the information security of stored and transmitted information was shown: noise immunity and reliability of the initial data recovery.

The above description of the essential characteristics of the invention also significantly enriches the theory of data compression and noise-resistant economical coding of information [15] based on the use of a new applied mathematical theory. Such a new interpretation significantly expands the areas of possible use of the invention, contributes to the emergence of new information technologies that have an unusual, from the point of view of existing practice, name "noise-immune compression".

List of information sources used

1. "Method of information transmission and device for its implementation", patent RU No. 2475861.

2. "Method of information transmission and system for its implementation", patent RU No. 2480840.

3. "Method of information transmission and device for its implementation", patent RU No. 2581774.

4. "Method of information transmission and system for its implementation", patent RU No. 2586605.

5. "Method of information transmission and system for its implementation", patent RU No. 2586833.

6. "Method of information transmission and system for its implementation", patent RU No. 2609747.

7. "Method of economical encoding of information and a system for its implementation", patent RU No. 2649291.

8. "Method of primary processing of information with the detection and correction of transmission errors", patent RU No. 2658795.

9. "Method of primary information processing with adaptive nonlinear data filtering", patent RU No. 2672392.

10. Kukushkin S.S. Finite field theory and informatics: In 2 vols. - vol. 1: Methods and algorithms, classical and non-traditional, based on the use of the constructive remainder theorem. - M.: MO RF, 2003. -284 p.

11. Lidl R., Niederreiter G. Finite fields. In 2 volumes. Per from English. M.: Mir, 1988. 882 p.

12. Emelyanov G.A., Shvartsman V.O. Transfer of discrete information. Moscow: Radio and communication, 1982. 240 p.

13. "Method for synchronizing transmitted messages and a device for its implementation", patent RU No. 2 591 565 C1, publication date 10.07.16, bul. No. 20).

14. Bylinski P., Ingram D. Digital transmission systems // Translated from English, ed. A.A. Wenzelya /, M .: "Communication", 1980. - 360 p.

15. Semenyuk V.V. Economic coding of discrete information. - St. Petersburg: SPbGITMO (TU), 2001. - 115 p.

Claims (1)

A method for compressed error-correcting data coding for transmitting and storing information, which consists in performing the following sequence of operations: code words and messages to be transmitted or stored are represented by a positional binary code with symbols "1" and "0" having a duration of T ₀ , from which a digital group signal (CGS) in the form of a sequence of bits, which is then subjected to structural-algorithmic transformations of the first stage (SAP-1), the simplest technical implementation of which is associated with the division of the original code words or messages X _i into the highest ( a _sti ) and the lowest ( a _mli ) half-words with their subsequent rearrangement, as a result of which a new CGS stream is formed, which has the properties of increased noise immunity in the presence of a correlation between adjacent transmitted or stored values of code words or messages, after which they proceed to the second stage of SAP (SAP-2) based on converting the new bit sequence into a replacement ternary error-correcting code with symbols S _i (T _i ), i=0, 1, 2, which is formed on the basis of the accepted logical coding rule that establishes the following correspondences "↔", for example S ₀ (T ₀ ) ↔ {<00> ₂ ,<11> ₂ }, S ₁ (T ₁ ) ↔ {<10> ₂ ,<001> ₂ } and S ₂ (T ₂ ) ↔ <101> ₂ , each of which has three encoding symbols, conditionally defined as "S ₀ " and "T ₀ ", "S ₁ " and "T ₁ ""S ₂ " and "T ₂ ", while the characters "S ₀ ", "S ₁ " and "S ₂ " are represented as pulse amplitude modulation (AIM ₃ ) with three allowed code positions, and the second modulating component of the video signal with the symbols of the tribasic code "T ₀ ", "T ₁ " and "T ₂ " with durations T ₀ , T ₁ \u003d 1.5T ₀ and T ₂ \u003d 2T ₀ , where T ₀ is the duration of one symbol of the original binary code, is used for pulse-width modulation (PWM ₃ ), characterized in that on the transmitting side occurring in the sequence of PWM ₃ signals, consecutive pulses of the ternary code with durations T ₀ and T ₁ \u003d 1.5T ₀ , as well as T ₀ and T ₂ =2T ₀ are combined into longer pulses with total time intervals 2.5T ₀ and 3T ₀ , respectively, leaving other PWM signals ₃ in the same order, formed as a result of the combined combinations, a sequence of signals with additional durations 2.5T ₀ and 3T ₀ are used for phase modulation of the carrier frequency signal with phase values of 0° and 180° when changing the polarity of the PWM ₃ pulse signals, as a result of which the intensity of the phase change of the carrier frequency signal is further reduced, providing the possibility of increasing the speeds and volumes of transmitted information , on the receiving side, the durations of PWM ₃ restored during demodulation are determined and the following code structures are assigned to them (CC _i , i=0, 1, 2, 3, 4): T ₀ ↔ {<00> ₂ ,<11> ₂ } , T ₁ ↔ {<10> ₂ ,<001> ₂ } and T ₂ ↔ <101> ₂ previously used, as well as new transcripts: T ₃ ↔ {<110> ₂ ,<0001> ₂ } and T ₄ ↔ <1101> ₂ retaining the same properties of the recurrent code , according to which the last binary symbol "1" or "0" of the previous decryption is the first in the decryption of the next code structure (CC _i , i=0, 1, 2, 3, 4).

Images