JPH1056483A - Data transmission system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize flexible data communication by changing the number of bits mapped to a signal point. SOLUTION: An incidence probability predict section 12 receiving a symbol stream from a data source 11 of a transmitter estimates incidence probability based on the incidence frequency of symbols and reports it to an area division section 13. The area division section 13 conducts area division bit share based on the incidence probability of the symbol and reports it to a mapping section 14. The mapping section 14 selects an area in response to the symbol stream from the data source from the area divided by the area division section and sends its representative point to a modulation section 15 as a signal point, from which the signal is sent. A receiver uses a demodulation section 21 to demodulate waveform data and converts the data into a signal point, a Demapping section 22 inputs a symbol stream corresponding to the area including this signal point to an incidence probability predict section 23, which predicts the incidence probability. As the area division section 24 uses the probability to conduct area division and the result is used by the Demapping section 22.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to a data transmission system.

[0002]

2. Description of the Related Art In conventional data communication, symbol mapping basically uses a predetermined signal point arrangement. In this method, the same data communication speed and the same error rate are realized for any data. If the data structure has redundancy, separate compression is used to achieve faster data communication, and a separate error correction method is implemented when low-error, high-precision communication is required. It has been realized by doing.

[0003]

SUMMARY OF THE INVENTION The present invention realizes flexible data communication by adaptively changing the signal point arrangement or changing the number of bits mapped to signal points according to the purpose. Since the data transmission method itself includes control of compression and error accuracy, data communication according to the purpose can be performed without using other devices.

[0004]

According to a first aspect of the present invention, there is provided a data transmission system for modulating one symbol (signal point) to which a plurality of bits of digital data are allocated and transmitting the digital data, the symbol appearance frequency. Characterized by adaptively changing the mapping of signal points according to
Thus, flexible data transmission suitable for the data source (purpose) can be realized.

According to a second aspect of the present invention, in the first aspect of the present invention, the amplitude modulation method is used for symbol modulation, and only the code section is changed while using the conventional amplitude modulation data transmission apparatus. By doing so, flexible data transmission suitable for the data source (purpose) can be realized.

A third aspect of the present invention is characterized in that, in the first aspect of the present invention, frequency modulation is used for symbol modulation, and only the code section is changed while using a conventional frequency modulation data transmission apparatus. Thus, flexible data transmission suitable for the data source (purpose) can be realized.

According to a fourth aspect of the present invention, in the first aspect of the present invention, phase modulation is used for symbol modulation, and only the code section is changed while using a conventional phase modulation data transmission apparatus. Thus, flexible data transmission suitable for the data source (purpose) can be realized.

A fifth aspect of the present invention is characterized in that, in the first aspect of the present invention, quadrature amplitude modulation is used for symbol modulation, so that only a code part is used while using a conventional quadrature amplitude modulation data transmission apparatus. By making the change, flexible data transmission suitable for the data source (purpose) can be realized.

According to a sixth aspect of the present invention, in the first aspect of the present invention, the method of arithmetic coding is used for adaptive mapping of signal points. The error rate of a symbol having a high appearance frequency is reduced.

According to a seventh aspect of the present invention, in the first aspect of the present invention, a conditional probability in a case where there is a symbol sequence earlier than the symbol occurrence probability is used instead of the symbol appearance probability. Is used to reduce the error rate for more likely symbols.

According to an eighth aspect of the present invention, in the first aspect of the present invention, the probability of appearance of a symbol is approximated by using the frequency of the symbol transmitted and received up to the time of transmission, whereby the probability of appearance is known in advance. It is possible to adaptively follow without the need.

According to a ninth aspect of the present invention, there is provided a data transmission system for transmitting digital data by modulating one symbol (signal point) to which a plurality of bits of digital data are allocated, wherein 1 is set in accordance with the appearance frequency of 0 and 1. Characterized by adaptively changing the number of bits assigned to one symbol,
Thus, by adaptively changing the number of allocated bits according to the frequency of appearance of symbols, the number of bits per transmitted symbol can be increased and the transmission rate can be increased as compared with the conventional mapping. Things.

According to a tenth aspect of the present invention, in the ninth aspect of the present invention, the number of bits mapped to one symbol (signal point) is adaptively changed within a range in which a predetermined minimum signal point distance can be maintained. By defining the minimum distance between signal points in advance, the worst value of the error rate can be guaranteed while reducing the process of calculating the position of the signal point.

According to an eleventh aspect of the present invention, in the tenth aspect of the present invention, the state of the transmission path is checked in advance, and the minimum distance between signal points is determined according to the state. By determining the distance according to the state of the transmission line, the transmission rate can be adaptively determined according to the state of the transmission line.

A twelfth aspect of the present invention is characterized in that, in the ninth aspect of the present invention, an arithmetic code technique is used for adaptive symbol mapping, and the arithmetic code is theoretically used for mapping. Using the held means, it is possible to easily change the number of bits adaptively.

According to a thirteenth aspect, in the twelfth aspect, when the number of bits to be mapped to one symbol is P (S), the appearance probability of a certain symbol S is P (S) ^-1.
Is adaptively changed in a range smaller than a predetermined value, and by adaptively changing the mapping in a range where ΣP (S) ^-1 is smaller than a predetermined value, the error rate can be reduced. The worst value can be held.

According to a fourteenth aspect of the present invention, in the ninth aspect of the present invention, a conditional probability in a case where there is a symbol series earlier than the symbol occurrence probability is used instead of the symbol appearance probability. Instead of using a low probability of occurrence, a conditional probability based on a previously transmitted (received) symbol sequence is used, thereby increasing the prediction probability and realizing a higher transmission rate. is there.

According to a fifteenth aspect of the present invention, in the ninth aspect of the present invention, the symbol appearance probability is approximated by using the frequency of the symbols transmitted and received up to the time of transmission.
Therefore, by adaptively changing the appearance probability of a symbol without knowing the appearance probability in advance by using the frequency of the symbols transmitted and received by the time of transmission, it is possible to adaptively change the appearance probability of the symbol without knowing the appearance probability in advance. It is made to be able to follow.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Here, for the sake of simplicity, a case will be described in which a phase modulation method (PSK) is used as a modulation method and mapping is performed by applying an arithmetic code method. First, compression using general arithmetic codes will be briefly described. In FIG. 1, it is assumed that the data is binary (here, represented by A or B in order to avoid confusion with other numerical values), the occurrence frequency is biased, and P _A and P _B are assumed respectively. At this time, 0
Consider the first area from the region of which 1 to P _A represents B, P _A to 1 can be viewed as a region representing the A. Therefore, if B is sent from the data source, the area P
With _A to 1 may represent B, it followed when A is sent, the further P _A region P _A to 1 in Figure 1: P _B
, The area of P _{A to} P _A + (1−P _A ) P _A can represent the bit string BA. As described above, a bit string can be represented by a certain area. At the time of encoding, by encoding one representative point included in this area, the original bit string can be restored. Since a bit string with a high appearance frequency occupies a wide area, a point representing an area corresponding to a bit string with a high appearance frequency can be represented by a small number of bits, and compression can be performed. For example, P _A = 3/4, P _B
= 1/4, the bit string "AABA" becomes 2
This corresponds to an area of 7/64 to 135/256. However, 0.5 in this section, that is, (0.1) ₂ in a binary system is set as a representative point, and only the decimal part is encoded. In this case, only one bit is required. What the restoring side needs to know is only the number of bits included in one representative point and the respective appearance probabilities.

Next, a general PSK modulation method will be briefly described with reference to FIG. In the PSK modulation method, data is carried on phase information of a carrier wave, and data transmission is performed. FIG. 2 shows an eight-level PSK modulation method. In this case, one symbol carries 3-bit information. If this is regarded as dividing the region between 0 and 2π, it can be seen that the region is divided by each symbol, as shown in FIG.
8). Thus, there is a deep similarity between mapping by the PSK modulation scheme and division of arithmetic coding. The present invention makes use of this similarity and applies division of arithmetic codes to mapping.

FIG. 4 is a diagram showing an example of the mapping according to the present invention. This mapping is the same as the example used in the arithmetic code described above. If the areas 0 to 1 in the arithmetic code are regarded as the phase angles 0 to 2π, the area for each bit string can be set in exactly the same way as the area allocation of the arithmetic code. Ie, A, the probability of occurrence of B, respectively, and P _A, 1-P _A, region representing the A is 0~2π
The area representing P _A and B is 2πP _{A to} 2π. Further AA
Is a range of 0 to 2πP _A ² , and a region of AB is 2πP _A ²
The area representing ２2πP _A , BA is 2πP _A ２2π (P _A +
_{(1-P A) P A} ), the area representing the BB is 2π (P _A + (1
−P _A ) P _A ) 〜2π. As a signal point representing each bit string, a point representing each region, that is, an intermediate point of an arc is used. If this mapping method is used, it is difficult for an error to occur in a symbol (bit string) having a high appearance probability, and conversely, a symbol (bit string) having a low appearance probability.
Has a higher error probability.

By the way, when performing the mapping using the arithmetic code as it is, the appearance probability of the symbol "AAA" is represented by P
When _AAA, probability E _AAA symbol AAA is received in error is inversely proportional to the square of the distance between the signal points representing the next symbol can be expressed as follows. _{E AAA α (P AAA π)} -2 Therefore, the average error rate E is a ^{_{EαΣP S (P S π) -2}} = π -2 ΣP S -1 ... (1). Here, Σ represents the sum of all symbols S. On the other hand, when mapping is performed at regular intervals, E∝ΣP _S (π / n) ⁻² = n ² π ⁻² (∵ΣP _S = 1), where n is the number of symbols. 'If we calculate the, E-E' E-E becomes a ^{_{= π -2 ΣP S -1 -n 2}} π -2 = π -2 (ΣP S -1 -n 2), ΣP S -1 Since −n ² ≧ 0, the probability of error increases.

Assuming that P _S = 1 / n for all S, ΔP _S ⁻¹ = n ² . If two of them are S
ΣP _S ^-1 = (n−2) / n
+ 1 / (n + k) + 1 / (nk). Here, k is n> k ≧ 0. When the last two terms are calculated, 1 / (n
+ K) + 1 / (n -k) = 2n / (n 2 -k 2) ≧ 2n / n 2
= 2 / n, and it can be seen that ΔP _S ⁻¹ −n ² ≧ 0.

Therefore, instead of adaptively changing the size of the area, the invention of claim 9 considers changing the number of bits allocated to each area. For example, when division of a region is repeated so as to maintain a predetermined minimum signal point distance, the number of bits mapped for each region differs. In this case, the probability that a symbol is erroneously received is almost the same as in the normal mapping, but the expected value of the number of bits mapped to one symbol is different from the conventional mapping. FIG. 5 shows an example in which the distance between signal points is determined in advance (that is, using a conventional signal point arrangement diagram), and mapping is performed on the distance instead of dynamically dividing the area.

FIG. 5B shows an example in which the area division in the conventional signal point arrangement is applied to FIG. 4, and FIG. 5A shows an approximate bit assignment in accordance with this area. The area of the symbol "AAA" in FIG.
In (A), the symbol is further divided into symbols “AAAAA” +
“AAAAB” + “AAAAB”. On the other hand, the symbols “BBB”, “BBA”, “BA” in FIG.
5A is integrated into a symbol "BB" area in this manner, and although the number of bits mapped for each signal point is different in this mapping method, the area division in the arithmetic code is performed. When this method is used, the error probability of the symbol is the same as that of the normal mapping, and the number of bits per symbol is different. In this example, the number of bits to be mapped is expected value, when the probability of occurrence of a was _{P a, 2 (1-P} a) 2 +2 (1-P a) P a + 3P a (1-P a) 2 + 3P a (1-P _A ) P _A + 3P _A (1-
P _A ) + 4P _A ³ (1-P _A ) +5 P _A ⁴ (1-P _A ) +5 P _A ⁵ = P _A ⁴ + P _A ³ + P _A +2 FIG. 6 is a plot of this value. As you can see, as the P _A approaches 1, so that the number of bits mapped is increased. In general mapping, in the case of an 8-point signal point arrangement, 3 per symbol is used.
Therefore, if P _Aよ り 大 き い P _A is greater than 0.6,
More bits can be put in one symbol.

The error probability is proportional to ΔP _S ^{-1 as} can be seen from equation (1). So check this value,
If the number of bits to be mapped is adaptively changed so as to be smaller than a predetermined value, mapping can be performed while suppressing the error probability. Here, a description has been given as a simple occurrence probability, but a conditional probability based on a bit sequence before a coded bit may be used. Even if the probability is not known in advance, the probability can be adaptively approximated by recording the frequency of the transmitted and received symbols.

In the method of adaptively arranging signal points, in the case of amplitude modulation instead of phase modulation, 0 to 2
Region allocation is performed for 0 to the maximum amplitude instead of the phase angle of π. In the case of frequency modulation, a region is allocated to the frequency shift width. In the case of quadrature amplitude modulation, symbols are divided into groups each having a similar appearance probability, and a region is assigned to each group in the amplitude direction. That is, the area of each group has a ring shape. Even if the area division in the ring is performed evenly, the area division as an approximation is realized.

FIG. 7 is a block diagram showing an implementation example of the present invention. An appearance probability predicting unit 12 which has received a symbol sequence from a data source 11 of a transmitter 10 generates an appearance probability based on an appearance frequency of a symbol. The probability is estimated and reported to the area dividing unit 13 (see FIG. 8). The region dividing unit 13 performs region division and bit allocation based on the symbol appearance probability,
Notify g section 14. FIG. 10 is a flowchart showing a method of using an arithmetic code method in area division, and FIG. 11 is a flowchart showing a method of adaptively changing the number of bits to determine and assign a minimum area in area division. It is represented by. The mapping unit 14 selects a region corresponding to the symbol string from the data source 11 from the regions divided by the region dividing unit 13 and sends the representative point to the modulation unit 15 as a signal point (see FIG. 9). . The modulation unit 15
The signal point is modulated and transmitted. In the receiver 20, first, the demodulation unit 21 demodulates the waveform data and converts it into signal points.
The mapping unit 22 outputs a symbol sequence corresponding to the area including the signal point. The output symbol sequence enters the appearance probability prediction unit 23 at the same time, and calculates the appearance frequency to thereby predict the appearance probability. Region dividing unit 24
In this case, the region is divided using the probability in the same manner as that of the transmitter 10, and is used by the demapping unit 22. Here, the description has been given based on the above implementation example. However, if the appearance probability is known in advance, it is not necessary to use the appearance rate prediction unit, and a predetermined area division is used. Instead of using the appearance probabilities, conditional appearance probabilities may be used based on the preceding appearance symbol sequence.

[0029]

【The invention's effect】

According to the first aspect of the present invention, in a data transmission method for transmitting digital data by modulating one symbol (signal point) to which a plurality of bits of digital data are allocated, mapping of signal points according to the appearance frequency of the symbol Is adaptively changed, so that flexible data transmission according to the data source (purpose) can be realized.

Operation and Effect Corresponding to Claim 2 In the invention of claim 1, since the amplitude modulation method is used for symbol modulation, only the code section is changed while using the conventional amplitude modulation data transmission apparatus. By doing so, flexible data transmission suitable for the data source (purpose) can be realized.

Operation and effect corresponding to claim 3: In the invention of claim 1, since frequency modulation is used for symbol modulation, only the code section is changed while using a conventional frequency modulation data transmission apparatus. Thereby, flexible data transmission suitable for the data source (purpose) can be realized.

Operation and Effect Corresponding to Claim 4: In the invention of claim 1, since phase modulation is used for symbol modulation, only the code section is changed while using a conventional phase modulation data transmission device. Thereby, flexible data transmission suitable for the data source (purpose) can be realized.

Operation and Effect Corresponding to Claim 5: In the invention of claim 1, since the quadrature amplitude modulation is used for symbol modulation, only the encoding section is used while using the conventional quadrature amplitude modulation data transmission apparatus. By making the change, flexible data transmission suitable for the data source (purpose) can be realized.

Operation and Effect Corresponding to Claim 6: In the invention of claim 1, since the arithmetic coding method is used for the adaptive mapping of signal points, the error rate for symbols with a high appearance frequency can be reduced. .

Operation and Effect Corresponding to Claim 7: In the invention of claim 1, instead of the appearance probability of a symbol, a conditional probability in the case where there is a symbol sequence earlier than that is used, so that it is more likely. The error rate for symbols can be reduced.

Operation and Effect Corresponding to Claim 8: In the invention of claim 1, since the appearance probability of a symbol is approximated by using the frequency of symbols transmitted and received up to the time of transmission, the appearance probability is known in advance. Can be adaptively followed without the need.

According to a ninth aspect of the present invention, in a data transmission system for transmitting digital data by modulating one symbol (signal point) to which a plurality of bits of digital data are assigned, according to the appearance frequency of 0 and 1 The number of bits to be allocated to one symbol is adaptively changed by changing the number of bits to be allocated adaptively according to the symbol appearance frequency. The number of bits per hit increases, and the transmission rate can be increased.

Operation and effect corresponding to claim 10: claim 9
Since the number of bits to be mapped to one symbol (signal point) is adaptively changed within a range capable of holding a predetermined minimum signal point distance, the minimum signal point distance is determined in advance. The worst value of the error rate can be guaranteed while reducing the process of calculating the position of the point.

Operation and effect corresponding to claim 11: claim 1
In the invention of No. 0, the state of the transmission path is checked in advance, the minimum distance between signal points is determined according to the state, and the minimum distance between signal points is determined according to the state of the transmission path. The transmission rate can be determined adaptively according to the state.

Operation and effect corresponding to claim 12: claim 9
According to the invention, the arithmetic coding method is used for the adaptive mapping of symbols, so that it is possible to easily change the number of bits adaptively using the theoretically held means.

Operation and Effect Corresponding to Claim 13: Claim 1
In the invention of the second aspect, when the number of bits to be mapped to one symbol is P (S), where the appearance probability of a certain symbol S is P (S),
Since の P (S) ^-1 is adaptively changed in a range smaller than a predetermined value, and ΣP (S) ^-1 is adaptively changed in a range smaller than a predetermined value, an error is generated. The worst value of the rate can be maintained.

Action and effect corresponding to claim 14: claim 9
In the invention of the above, instead of the probability of appearance of a symbol, the conditional probability in the case where there is an earlier symbol sequence is used, and instead of using the simple probability of appearance of a symbol, the symbol is transmitted (received) earlier. Since the conditional probability based on the symbol sequence is used, the prediction probability can be increased, and a higher transmission rate can be realized.

Operation and effect corresponding to claim 15: claim 9
In the invention, the symbol appearance probability is adaptively changed without knowing the appearance probability in advance by approximating the symbol appearance probability using the frequency of the symbols transmitted and received by the time of transmission. Since the approximation is made using the frequency of the symbols transmitted and received until the time of transmission, it is possible to adaptively follow without knowing the appearance probability in advance.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram for explaining compression by an arithmetic code.

FIG. 2 is a conceptual diagram illustrating a PSK modulation scheme.

FIG. 3 is a conceptual diagram showing how a region is divided by a symbol.

FIG. 4 is a diagram illustrating an example of mapping according to the present invention.

FIG. 5 is a diagram for explaining an example in which a distance between signal points is determined in advance, and mapping is performed thereon.

FIG. 6 is a diagram illustrating a relationship between an appearance rate and the number of bits.

FIG. 7 is a block diagram for explaining an embodiment of the present invention.

FIG. 8 is a flowchart for explaining the operation of an appearance probability unit.

FIG. 9 is a flowchart illustrating the operation of a mapping unit.

FIG. 10 is a flowchart for explaining the operation of a region dividing unit.

FIG. 11 is another flowchart for explaining the operation of the area dividing unit.

[Explanation of symbols]

10: transmitter, 11: data source, 12, 23: appearance probability prediction unit, 13, 24: region dividing unit, 14: mapping unit,
15: Modulator, 20: Receiver, 21: Demodulator, 22: De
mapping unit, 25 ... data output unit.

Claims (15)

[Claims] In a data transmission system for transmitting digital data by modulating one symbol (signal point) to which digital data of a plurality of bits are assigned, mapping of a signal point is adaptively performed according to the appearance frequency of the symbol. A data transmission method characterized by changing. 2. The data transmission method according to claim 1, wherein an amplitude modulation method is used. 3. The data transmission method according to claim 1, wherein frequency modulation is used. 4. The data transmission method according to claim 1, wherein phase modulation is used. 5. The data transmission method according to claim 1, wherein quadrature amplitude modulation is used. 6. The data transmission method according to claim 1, wherein an arithmetic code method is used for adaptive mapping of signal points. 7. The data transmission method according to claim 1, wherein a conditional probability in a case where there is an earlier symbol sequence is used instead of a symbol appearance probability. 8. The data transmission method according to claim 1, wherein the appearance probability of the symbol is approximated using the frequency of the symbol transmitted and received up to the time of transmission. 9. In a data transmission system for transmitting digital data by modulating one symbol (signal point) to which a plurality of bits of digital data are allocated, bits allocated to one symbol in accordance with the appearance frequency of 0 and 1 A data transmission method characterized by adaptively changing the number. 10. The data transmission method according to claim 9, wherein the number of bits to be mapped to one symbol (signal point) is adaptively changed within a range that can maintain a predetermined minimum distance between signal points. 11. The data transmission method according to claim 10, wherein the state of the transmission path is checked in advance, and the minimum distance between signal points is determined according to the state. 12. The data transmission method according to claim 9, wherein an arithmetic code method is used for adaptive symbol mapping. 13. The method according to claim 12, wherein the number of bits to be mapped to one symbol is within a range where ΣP (S) ⁻¹ is smaller than a predetermined value, when the appearance probability of a certain symbol S is P (S). A data transmission method characterized by being adaptively changed. 14. The data transmission method according to claim 9, wherein a conditional probability in a case where there is a symbol sequence earlier than the symbol occurrence probability is used instead of the symbol appearance probability. 15. The method according to claim 9, wherein the symbol appearance probability is adaptively changed without previously knowing the appearance probability by approximating the symbol appearance probability using the frequency of the symbols transmitted and received up to the time of transmission. Data transmission method.