JP2006261857A - Decoder and decoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a decoder for outputting a result of hard decision associated with data S on the basis of redundant data P<SB>1</SB>(P<SB>2</SB>) attached to the data that adopts a highly accurate likelihood for combinations of codes of the S, P<SB>1</SB>more than that of prior arts. <P>SOLUTION: The decoder includes: data conversion means 401, 402 for inverting part bits of received data and providing an output of the resulting data; a redundant data conversion means 401, 402 for inverting part bits of received redundant data and providing an output of the resulting data; discrimination means 404, 406 for discriminating whether or not the received data are zero and whether or not the received redundant data are zero; and a likelihood calculation means for summating either of an output of the data conversion means or a first fixed value X and an output of the redundant data conversion means or a second fixed value Y to provide an output of a likelihood. The likelihood calculation means adds the first fixed value when the received data are zero and adds the second fixed value when the received redundant data are zero. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to the technical field of wireless communication, and more particularly to a decoder and decoding method used in a wireless receiver.

  Fading and other communication environments in the radio propagation path greatly affect the quality of signals to be communicated. One technique for improving communication quality is to perform encoding (decoding) and decoding (decoding). The transmitter encodes and transmits the signal, and the receiver determines whether or not there is an error in the signal and / or corrects the error, thereby improving the transmission quality. Among these, turbo coding and turbo decoding are promising technologies in the present and future in terms of excellent data processing efficiency and large error correction capability.

  In turbo coding and decoding, an information sequence (data) S (systematic bits) and a redundant information sequence (redundant data) P (parity bits) derived from the information sequence are created and transmitted. Redundant data is formed from encoded data obtained by performing convolutional coding on data, and encoded data obtained by performing convolutional coding after performing interleaving processing on the data. The receiver receives the data S and the redundant data P, calculates the likelihood related to the combination of the most significant bits (MSB), executes a predetermined operation based on the likelihood, and performs a hard decision result of the data S Is output and decryption is performed. In general, the MSB of data represents the sign of the data, and the bit string after the MSB represents the likelihood.

In the turbo decoder, the likelihood L _ij (i, j = 0 or 1) regarding the combination of the MSB of the data S and the redundant data P is calculated as follows:
The likelihood L ₀₀ that the MSB of the data S is “0” (positive) and the MSB of the redundant data P is also “0” (positive) is obtained by inverting the MSB of S and the MSB of P Expressed as the sum of the inverted version;
The likelihood L ₀₁ that the MSB of the data S is “0” and the MSB of the redundant data P is “1” (negative) is obtained by inverting the MSB of S and all the bits other than the MSB of P Expressed as the sum of the inverted version;
The likelihood L ₀₁ in which the MSB of the data S is “1” and the MSB of the redundant data P is “0” is obtained by inverting all the bits other than the MSB of S and by inverting the MSB of P. And the likelihood L ₁₁ that the MSB of the data S is “1” and the MSB of the redundant data P is also “1” is obtained by inverting all the bits other than the MSB of S and P This is expressed as the sum of all bits other than the MSB.

For example, assuming that S = 0111 and P = 0111 (binary number), the likelihood L _ij is as follows.

L ₀₀ = 0 111 + 0 111 = 1111 + 1111 = 11110
L ₀₁ = 0 111 + 0 111 = 1111 + 0000 = 01111
_{L 10 = 0 111 + 0 111} = 0000 + 1111 = 01111
L ₁₁ = 0 111 +0 111 = 0000 + 0000 = 00000.
However, the underlined bits represent being inverted. Comparing the likelihood values, it can be seen that L ₀₀ is the maximum. This indicates that it is most probable that both the signs of data S and redundant data P are positive (both MSBs are 0). Actually, S = P = 0111 is a positive value and corresponds to +7 in the decimal system. Therefore, it is determined that it is most probable that both of the possible combinations of the codes of the data S and the redundant data P are positive, and the basis for the subsequent decoding calculation can be given accurately. .

A conventional turbo decoder that performs decoding by calculating likelihood in this way is described in Patent Document 1, for example.
JP 2001-35652 A

By the way, in the above calculation, 2's complement expression is employed to increase the range of numerical values that can be expressed. For this reason, inconvenience occurs when the value of data and / or redundant data is zero. For example, assuming that data S = 0111 and P = 0000, the likelihood L _ij is calculated as follows according to the above calculation rule.

L ₀₀ = 0 111 + 0 000 = 1111 + 1000 = 10111
L ₀₁ = 0 111 + 0 000 = 1111 + 0111 = 10110
_{L 10 = 0 111 + 0 000} = 0000 + 1000 = 01000
L ₁₁ = 0 111 +0 000 = 0000 + 0111 = 00111.
According to this calculation result, since L ₀₀ = 10111 is the maximum, it is most likely that both of the possible combinations of the codes of the data S and the redundant data P are positive. However, in practice, since the data S is positive and the redundant data P is zero, such likelihood is not obtained. Therefore, in this case, an incorrect basis is given to the subsequent decoding calculation. In this example, since the data S is positive, it may be said that the calculated likelihood is incorrect for the redundant data P but not appropriate for the data S. However, for example, if the data S = 0000 and P = 0000, the likelihood L _ij is calculated as follows, which leads to a more unreasonable result.

L ₀₀ = 0 000+ 0 000 = 1000 + 1000 = 10000
L ₀₁ = 0 000 + 0 000 = 1000 + 0111 = 01111
L ₁₀ = 0 000 + 0 000 = 0111 + 1000 = 01111
L ₁₁ = 0 000 +0 000 = 01111 + 0111 = 011110.
According to this calculation result, since L ₀₀ = 10000 is the maximum, it is most likely that both of the possible combinations of the codes of the data S and the redundant data P are positive. However, since the data S and the redundant data P are actually zero, the likelihood should be an equal value between L _{00 and} L _11. Thus, the likelihood that L ₀₀ is relatively high and The end result is undesirable.

Note that the data S and the redundant data P become zero when, for example, zero is set when the transmission side supplements the portion where the puncture process is performed and the bits are deleted.
The present invention has been made in view of the above problems, and the problem is that in a decoder or decoding method that outputs a hard decision result related to data based on the data and redundant data accompanying the data, This is to obtain a likelihood for a combination of redundant data codes with higher accuracy than in the past.

  In the present invention, a decoder that outputs a hard decision result of data based on the likelihood of the combination of data and the code of redundant data accompanying the data is used. The decoder includes: data conversion means for inverting and outputting some bits of input data; redundant data conversion means for inverting and outputting some bits of input redundant data; and input data Determining means for determining whether or not the input redundant data is zero, and at least one of the output of the data conversion means or the first fixed value And likelihood calculating means for adding the output of the redundant data converting means and at least one of the second fixed values to output the likelihood. The likelihood calculation means selects a fixed value in the addition when the determination means determines that the value is zero.

  According to the present invention, in a decoder or decoding method that outputs a hard decision result related to data based on the data and redundant data accompanying the data, the likelihood of the combination of the data and the code of the redundant data is higher than in the past. The accuracy can be obtained.

  According to one aspect of the present invention, whether or not the received data is zero is output to the decoder that outputs a hard decision result of the data based on the likelihood for the combination of the data and the redundant data code accompanying the data. And determining means for determining whether or not the received redundant data is zero, the first fixed value is added if the received data is zero, and the second if the received redundant data is zero. The likelihood is calculated by adding the fixed values of. Accordingly, it is possible to effectively prevent the zero value from being inappropriately signal-converted and deriving a biased likelihood.

  The data conversion means includes first conversion means for inverting and outputting the most significant bit of the received data, and data conversion means having second conversion means for inverting and outputting the bits other than the most significant bit of the received data. May be provided. The redundant data converting means includes first converting means for inverting and outputting the most significant bit of the received redundant data, and second converting means for inverting and outputting the bits other than the most significant bit of the received redundant data. May be.

  According to one aspect of the invention, the encoder may constitute a turbo decoder. In the D calculation unit or the γ calculation unit in the turbo encoder, the likelihood related to the combination of the codes of data and redundant data is calculated. Thus, the present invention is particularly advantageous for turbo decoders.

  According to one aspect of the invention, the first fixed value is equal to the second fixed value.

  According to the aspect of the present invention, the determination unit that determines whether data or redundant data is zero may perform the determination based on the puncturing flag. When data or the like is punctured, it is known where the zero is inserted in the received bit string, so the timing for selecting a fixed value can be easily obtained by using the puncturing flag. it can.

Hereinafter, a turbo decoder and a decoding method according to an embodiment of the present invention will be described. First, the transmitter encodes an information sequence (data), creates a redundant information sequence (redundant data), and wirelessly transmits the data and redundant data. Any suitable number of coding rates may be adopted, but for the sake of simplicity, in this embodiment, coding is performed when the coding rate is 1/3 as shown in FIG. In the illustrated example, the information sequence (data) S, the first redundant data P ₁ obtained by encoding the data S with the encoder 11, and the data S after being interleaved with the interleaver (π) 13 encoder 12 and the second redundant data P ₂ obtained by encoding are prepared in, signals of three systems are sent wirelessly. The encoder may consist of a recursive systematic convolutional encoder as shown for example in FIG. Interleaving and encoding may be performed for each frame of the information sequence.

  FIG. 3 shows an exemplary block diagram of a turbo decoder provided in the receiver. The turbo decoder includes an input buffer 302, a D operation unit 304, an A operation unit 306, a B operation unit 308, a B operation memory 310, an L operation unit 312, a likelihood output unit 314, and an external likelihood. A degree calculation unit 316, an interleaving memory 318, and a likelihood memory 320 are included.

The input buffer 302 stores a signal received by the receiver. Three systems of signals are stored according to the transmitted three systems of signals (data S, first redundant data P ₁ , second redundant data P ₂ ). The number of signal systems can be changed as appropriate according to the coding rate used for turbo coding. In the figure, “FP” indicates a pulse representing the head of the frame. “DTEN” is a signal indicating an enabled or disabled state, and various signals are stored in the enabled state.

D calculation section 304, the signal _S of the three systems _received, based on P 1, _{P 2} and extrinsic likelihood information _{L e,} a likelihood relates to the combination of the code data S and the redundant data _{P 1,} the data S and the redundant It calculates a likelihood concerning the combination of the code data P _2. This likelihood has significance as a branch metric. Details of the D calculation unit 304 will be described later.

  Based on the likelihood calculated by the D calculation unit 304, the A calculation unit 306 cumulatively calculates path metrics in the order in which signals are input.

  The B operation unit 308 also calculates path metrics cumulatively based on the likelihood calculated by the D operation unit 304, but the B operation unit 308 calculates the path metrics in the reverse order of the signal input order. It differs from the A calculation unit 306 in that it is calculated.

  The B calculation memory 310 stores the B calculation result.

  The L calculation unit 312 determines a path with high likelihood based on the path metrics of the A calculation unit 306 and the B calculation unit 308.

  The likelihood output unit 314 performs a hard decision on the data based on what is a path with a high likelihood, and outputs the result as a decoding result.

External likelihood computation section 316, the data S, based on the output L and previous external likelihood L _e of the likelihood output unit 314 calculates an extrinsic likelihood L _e to be used for the next operation.

  The interleave memory 318 is a memory for rearranging bit strings.

  The likelihood memory 320 stores past external likelihoods.

In general, the received signal is stored in the input buffer 302 and given to the D operation unit 304, the likelihood regarding the combination of the codes of the data S and the redundant data P ₁ and P ₂ is calculated, and the branch metrics are A and B operations. Are given to the units 306 and 308, a path metric is calculated, and a path with a high likelihood is determined. Based on the calculation result, the external likelihood is calculated and fed back to the D calculation unit 304, and thereafter the same calculation is repeated to improve the decoding accuracy.

FIG. 4 is a functional block diagram of the D operation unit 304 according to an embodiment of the present invention. In Figure 4, for simplicity, the data S, the first redundant data P ₁ and although only depicted state of the signal processing relating to the external likelihood L _e, the D calculation section 304, the data S, the second redundant similar functions of the signal processing relating to the data P ₂ and external likelihood L _e are also provided. 4 shows a data S, each signal path of the first redundant data P ₁ and the external likelihood L _e, the first signal converter 401 and a second signal conversion unit 402 is depicted. Further, each of the data S and the first redundant data P ₁ of the signal path, the first selector 411 and second selector 412 is depicted. Further, FIG. 4 illustrates a first determination unit 404, a second determination unit 406, addition units 421, 422, 423, and 424, a normalization unit 426, and addition units 431, 432, 433, and 434. .

  The first signal conversion unit 401 inverts the most significant bit (MSB) of a group of n-bit strings input thereto, and outputs the other n−1 bits as they are. For example, if 1011 is input to the first signal conversion unit, 0011 is output.

  The second signal conversion unit 402 inverts all n−1 bits other than the most significant bit of a group of n bit strings input thereto, and outputs the MSB as it is. For example, if 1011 is input to the second signal converter, 1100 is output.

  The first and second selectors 411 and 412 select either the output from the first or second signal conversion unit 401 or 402 or a predetermined fixed value according to the control signal from the determination unit 404 or 406. And output. As will be described later, the first selector 411 selects the output of the first signal conversion unit 401 if the signal input to the first signal conversion unit 401 is not zero. Select a fixed value. The second selector 412 selects the output of the second signal converter 402 if the signal input to the second signal converter 402 is not zero, and selects a predetermined fixed value if it is zero. . The fixed value can be any value, but it must be a constant value that is invariant during at least one likelihood calculation performed iteratively. The fixed value X used for the calculation related to the data S and the fixed value Y used for the calculation related to the redundant data P may be the same value or different.

  The first and second determination units 404 and 406 determine whether or not the signal input to the first and second signal conversion units 401 and 402 is zero (0000 if n = 4).

The adders 421, 422, 423, and 424 add the outputs from the selectors 411 and 412 to calculate the likelihood L _ij (i, j = 0 or 1), respectively.

The normalizing unit 426 normalizes the likelihood L _ij calculated by each adding unit, adjusts the magnitude of the likelihood with an appropriate scale, and outputs the result.

Adders 431, 432, 433, and 434 add the normalized likelihood and the external likelihood after signal conversion, respectively, and use the likelihoods γ ₀₀ , γ ₀₁ , γ ₁₀ as outputs of D calculation unit 304. , Γ ₁₁ is output. Thereby, the likelihood γ _ij in which the past likelihood L _e is cumulatively added can be output.

FIG. 5 shows an example of a procedure in the D operation unit of the decoder according to the embodiment of the present invention. In step 502, a signal is received at a receiver. The signal data S, includes first redundant data P ₁ and the second redundant data P _2, they are input to the input buffer 302 of FIG.

  In step 504, the signals input to the first and second signal converters 401 and 402 in the D operation unit are converted and applied to one input of the selectors 411 and 412. A predetermined fixed value is given to the other input of the selector.

  In step 506, the determination units 404 and 406 determine whether or not the signals input to the first and second signal conversion units 401 and 402 are zero. Any appropriate method may be adopted as the determination method. For example, an n-bit signal may be directly compared with 00 ... 0.

In step 508, the outputs or fixed values of the first and second signal conversion units 401 and 402 are selected by the selectors 411 and 412 and given to each of the addition units 421, 422, 423, and 424. Each adding unit adds the input values and outputs likelihood L _ij . By adding the external likelihood to this likelihood, the likelihood γ _ij is output from the D operation unit 304. As described above, the D calculation section 304, the data S, also performed a similar process using the second redundant data P ₂ and the external likelihood L _e. For convenience of explanation, FIG. 5 depicts that the determination of step 506 is made after the signal conversion is performed in step 504, but such order is not essential to the present invention. All or a part of these may be performed simultaneously, or in the reverse order.

Next, an example of likelihood calculation in the D operation unit will be described using specific numerical examples.
(1) It is assumed that the data S and the first redundant data P ₁ are given by S = 0111 and P ₁ = 0111. The first signal conversion unit 401 relating to the data S outputs 0 111 = 1111 (the underline means that the bit is inverted), and gives it to one input of the selector 411. The fixed value X input to the other selector is X = 1000. The determination unit 404 determines whether the data S is zero and provides a control signal to the selector 411. In the current example, since S = 0111 ≠ 0000, the control signal is generated so that the selector 411 selects the output from the first signal converter 401. As a result, 1111 is given to the adders 421 and 422, respectively. The second signal conversion unit 402 related to the data S outputs 0 111 = 0000 and gives it to one input of the selector 412. A fixed value X = 1000 is input to the other selector. The selector 412 selects 0000 according to the control signal from the determination unit 404 and supplies it to the adding units 423 and 424.

Similarly, the first signal conversion unit 401 related to the first redundant data P ₁ outputs 0 111 = 1111 and supplies it to one input of the selector 411. A fixed value Y = 1000 is input to the other selector. The determination unit 406 determines whether or not the data P ₁ is zero and provides a control signal to the selector 411. In the current example, since P ₁ = 0111 ≠ 0000, the control signal is generated so that the selector 411 selects the output from the first signal converter 401. As a result, 1111 is given to the adders 421 and 423, respectively. The second signal conversion unit 402 for the data P ₁ outputs 0 111 = 0000 and gives it to one input of the selector 412. A fixed value Y = 1000 is input to the other selector. The selector 412 selects 0000 according to the control signal from the determination unit 406 and supplies it to the addition units 422 and 424.

The adder adds the values given from the selectors and calculates the likelihood. In the current example, the following likelihood L _ij is obtained.

L ₀₀ = 0 111 + 0 111 = 1111 + 1111 = 11110
L ₀₁ = 0 111 + 0 111 = 1111 + 0000 = 01111
_{L 10 = 0 111 + 0 111} = 0000 + 1111 = 01111
L ₁₁ = 0 111 +0 111 = 0000 + 0000 = 00000.
Comparing the likelihood values, it can be seen that L ₀₀ is the maximum. This indicates that it is most probable that both of the combinations of codes of the data S and the redundant data P are positive. Actually, S = P = 0111 is a positive value and corresponds to +7 in the decimal system. Therefore, it is determined that it is most probable that both of the possible combinations of the codes of the data S and the redundant data P are positive, and the basis for the subsequent decoding calculation can be given accurately. .

(2) Assume that the data S and the first redundant data P ₁ are given by S = 0111, P ₁ = 0000. The first signal conversion unit 401 related to the data S outputs 0 111 = 1111 and gives it to one input of the selector 411. The fixed value X input to the other selector is X = 1000. The determination unit 404 determines whether the data S is zero and provides a control signal to the selector 411. In the current example, since S = 0111 ≠ 0000, the control signal is generated so that the selector 411 selects the output from the first signal converter 401. As a result, 1111 is given to the adders 421 and 422, respectively. The second signal conversion unit 402 related to the data S outputs 0 111 = 0000 and gives it to one input of the selector 412. A fixed value X = 1000 is input to the other selector. The selector 412 selects 0000 according to the control signal from the determination unit 404 and supplies it to the adding units 423 and 424.

The first signal conversion unit 401 related to the first redundant data P ₁ outputs 0 000 = 1000, and gives it to one input of the selector 411. A fixed value Y = 1000 is input to the other selector. The determination unit 406 determines whether or not the data P ₁ is zero and provides a control signal to the selector 411. In the current example, since P ₁ = 0000, the control signal is generated so that the selector 411 selects the fixed value Y. As a result, Y = 1000 is given to the adders 421 and 423, respectively. The second signal conversion unit 402 for the data P ₁ outputs 0 000 = 0111 and gives it to one input of the selector 412. A fixed value Y = 1000 is input to the other selector. The selector 412 selects a fixed value Y = 1000 according to the control signal from the determination unit 406 and supplies it to the adding units 422 and 424.

The adder adds the values given from the selectors and calculates the likelihood. In the current example, the following likelihood L _ij is obtained.

L ₀₀ = 0 111 + Y = 1111 + 1000 = 10111
L ₀₁ = 0 111 + Y = 1111 + 1000 = 10111
L ₁₀ = 0 111 + Y = 0000 + 1000 = 01000
L ₁₁ = 0 111 + Y = 0000 + 1000 = 01000.
Comparing the likelihood values, it can be seen that L ₀₀ and L ₀₁ are the maximum. This is because it is more probable that S is a positive combination of the signs of data S and redundant data P, but P ₁ is equally likely to be positive and negative. Show. Actually, S = 0111 is a positive value and corresponds to +7 in the decimal system. Since P ₁ is zero, its sign is neither positive nor negative. Therefore, it is reasonable that L ₀₀ and L ₀₁ are the same value and have values larger than L ₁₀ and L ₁₁ , and the basis for the subsequent decoding calculation can be given accurately.

(3) It is assumed that the data S and the first redundant data P ₁ are given by S = 0000, P ₁ = 0000. The first signal converter 401 relating to the data S outputs 0 000 = 1000, give it to one input of the selector 411. The fixed value X input to the other selector is X = 1000. The determination unit 404 determines whether the data S is zero and provides a control signal to the selector 411. In the current example, since S = 0000, the control signal is generated so that the selector 411 selects the fixed value X = 1000. As a result, X = 1000 is given to the adders 421 and 422, respectively. The second signal conversion unit 402 relating to the data S outputs 0 000 = 0111 and gives it to one input of the selector 412. A fixed value X = 1000 is input to the other selector. The selector 412 selects X = 1000 according to the control signal from the determination unit 404 and supplies it to the adding units 423 and 424.

The first signal conversion unit 401 related to the first redundant data P ₁ outputs 0 000 = 1000, and gives it to one input of the selector 411. A fixed value Y = 1000 is input to the other selector. The determination unit 406 determines whether or not the data P ₁ is zero and provides a control signal to the selector 411. In the current example, since P ₁ = 0000, the control signal is generated so that the selector 411 selects the fixed value Y. As a result, Y = 1000 is given to the adders 421 and 423, respectively. The second signal conversion unit 402 for the data P ₁ outputs 0 000 = 0111 and gives it to one input of the selector 412. A fixed value Y = 1000 is input to the other selector. The selector 412 selects a fixed value Y = 1000 according to the control signal from the determination unit 406 and supplies it to the adding units 422 and 424.

The adder adds the values given from the selectors and calculates the likelihood. In the current example, the following likelihood L _ij is obtained.

L ₀₀ = X + Y = 1000 + 1000 = 10000
L ₀₁ = X + Y = 1000 + 1000 = 10000
L ₁₀ = X + Y = 1000 + 1000 = 10000
L ₁₁ = X + Y = 1000 + 1000 = 10000.
Comparing the respective likelihood values, it can be seen that the values are the same and the likelihood is the same. This is because the combination of the signs of the data S and the redundant data P is equally as probable that S is positive and negative, and that P ₁ is also positive and negative. It shows what is certain. In fact, the sign of S = P = 0000 is neither positive nor negative. Therefore, it is appropriate that all the likelihoods L _ij have the same value, and the basis for the subsequent decoding calculation can be given accurately.

For simplicity, in the above example, the fixed value of the selector related to the data S and the fixed value of the selector related to the redundant data P ₁ are both X = Y = 1000, but they may be different fixed values. For example, the fixed value of the selector related to the data S may be X = 1000, and the fixed value of the selector related to the redundant data P ₁ may be Y = 0100. Furthermore, the fixed value can be any value, even zero (0000) is allowed. Whatever the fixed value is adopted, when the data S is zero, some same fixed value X is given to all the adding units, and when the redundant data S is zero, some same fixed value Y is given. This is because it may be given to all the adders.

  When the transmitted / received information sequence is a punctured code (when punctured), a signal (puncturing flag) indicating which part of the information sequence is zero is also notified to the receiver. Therefore, in such a case, as indicated by the broken-line arrows in FIG. 4, the determination units 404 and 406 use the puncturing flag to input the first and second signal conversion units 401 and 402. It is possible to easily determine whether or not the signal is zero.

  Hereinafter, the means taught by the present invention will be listed as an example.

(Appendix 1)
A decoder that outputs a hard decision result of data based on likelihood for a combination of data and a code of redundant data accompanying the data,
Data conversion means for inverting and outputting some bits of input data;
Redundant data conversion means for inverting and outputting some bits of the input redundant data;
Determining means for determining whether the input data is zero, or determining whether the input redundant data is zero; and
The likelihood is output by adding at least one of the output of the data conversion unit or the first fixed value and at least one of the output of the redundant data conversion unit or the second fixed value. Likelihood calculating means to perform,
The likelihood calculation means selects a fixed value in the addition when the determination means determines that the determination means is zero.

(Appendix 2)
Data conversion means, wherein the data conversion means includes first conversion means for inverting and outputting the most significant bit of the received data, and second conversion means for inverting and outputting bits other than the most significant bit of the received data And
The redundant data converting means includes first converting means for inverting and outputting the most significant bit of the received redundant data, and second converting means for inverting and outputting the bits other than the most significant bit of the received redundant data. The decoder according to appendix 1, wherein the decoder is provided.

(Appendix 3)
The decoder according to appendix 1, wherein the encoder is a turbo decoder.

(Appendix 4)
The decoder according to claim 1, wherein the first fixed value is equal to the second fixed value.

(Appendix 5)
The decoder according to appendix 1, wherein the determination unit performs determination based on a puncturing flag.

(Appendix 6)
A decoding method for outputting a hard decision result of data based on likelihood for a combination of data and a code of redundant data accompanying the data,
Selecting one of the converted data or the first fixed value in which a part of the received data is inverted, selecting one of the converted redundant data or the second fixed value in which a part of the received redundant data is inverted, Calculate the likelihood by adding the selected values,
If the received data is zero, the first fixed value is selected, and if the received data is not zero, the conversion data is selected,
A decoding method, wherein the first fixed value is selected when the received redundant data is zero, and the converted data is selected when the received data is not zero.

(Appendix 7)
A turbo decoder that outputs a hard decision result of data based on likelihood for a combination of data and redundant data code accompanying the data,
A data conversion means for outputting an inverted version of a specific part of the input data and an inverted version of the other part of the bit;
Redundant data conversion means for outputting an inverted version of a specific part of the input redundant data and an inverted version of the other part of the bits;
When the input data or the input redundant data does not correspond to the punctured portion, the addition result for each possible combination of the two outputs of the data conversion means and the two outputs of the redundant data conversion means When the input data corresponds to the punctured portion, the addition result is output for each possible combination of two fixed values of the same value and the two outputs of the redundant data conversion means. Likelihood calculation means;
A turbo decoder comprising:

It is a figure which shows an example of a turbo encoder. It is a figure which shows an example of an encoder. FIG. 3 shows a turbo decoder according to an embodiment of the present invention. The functional block diagram of the D calculating part by one Example of this invention is shown. It is a flowchart which shows an example of the procedure in the D calculating part by one Example of this invention.

Explanation of symbols

11, 12 Encoder 13 Interleaver 302 Input buffer 304 D operation unit 306 A operation unit 308 B operation unit 310 B operation memory 312 L operation unit 314 Likelihood output unit 316 External likelihood operation unit 318, 320 Memory 401, 402 Signal conversion unit 404, 406 determination unit 411, 412 selector 421, 422, 423, 424 addition unit 426 normalization unit 431, 432, 433, 434 addition unit

Claims (5)

A decoder that outputs a hard decision result of data based on likelihood for a combination of data and a code of redundant data accompanying the data,
Data conversion means for inverting and outputting some bits of input data;
Redundant data conversion means for inverting and outputting some bits of the input redundant data;
Determining means for determining whether the input data is zero, or determining whether the input redundant data is zero; and
The likelihood is output by adding at least one of the output of the data conversion unit or the first fixed value and at least one of the output of the redundant data conversion unit or the second fixed value. Likelihood calculating means to perform,
The likelihood calculation means selects a fixed value in the addition when the determination means determines that the determination means is zero. Data conversion means, wherein the data conversion means includes first conversion means for inverting and outputting the most significant bit of the received data, and second conversion means for inverting and outputting bits other than the most significant bit of the received data And
The redundant data converting means includes first converting means for inverting and outputting the most significant bit of the received redundant data, and second converting means for inverting and outputting the bits other than the most significant bit of the received redundant data. The decoder according to claim 1, comprising: a decoder. The decoder according to claim 1, wherein the encoder is a turbo decoder. A decoding method for outputting a hard decision result of data based on likelihood for a combination of data and a code of redundant data accompanying the data,
Selecting one of the converted data or the first fixed value in which a part of the received data is inverted, selecting one of the converted redundant data or the second fixed value in which a part of the received redundant data is inverted, Calculate the likelihood by adding the selected values,
If the received data is zero, the first fixed value is selected, and if the received data is not zero, the conversion data is selected,
A decoding method, wherein the first fixed value is selected when the received redundant data is zero, and the converted data is selected when the received data is not zero. A turbo decoder that outputs a hard decision result of data based on likelihood for a combination of data and redundant data code accompanying the data,
A data conversion means for outputting an inverted version of a specific part of the input data and an inverted version of the other part of the bit;
Redundant data conversion means for outputting an inverted version of a specific part of the input redundant data and an inverted version of the other part of the bits;
When the input data or the input redundant data does not correspond to the punctured portion, the addition result for each possible combination of the two outputs of the data conversion means and the two outputs of the redundant data conversion means When the input data corresponds to the punctured portion, the addition result is output for each possible combination of two fixed values of the same value and the two outputs of the redundant data conversion means. Likelihood calculation means;
A turbo decoder comprising:

Images