TWI387274B - Mediation variable method and device using soft decision for orthogonal amplitude modulation - Google Patents

Description

Mediation variable method and device using soft decision for orthogonal amplitude modulation

The present invention relates to soft decision demodulation of quadrature amplitude modulation (hereinafter referred to as QAM) signals, and more particularly to a soft decision demodulation method for improving soft decision processing speed by using some functions and modes when demodulating a received signal.

The QAM mode carries more than two bits on a specified waveform symbol. When such a waveform is represented by a mathematical formula, it can be represented by a real number and an imaginary number (ie, at a complex number α + βi). Since the orthogonal signal component corresponds to α and the in-phase signal component corresponds to β, the change in the α value does not affect the β value. The general quadrature phase component is called the Q-channel, and the in-phase component signal is called the I-channel.

The amplitudes of the two waveforms are interconnected to form a plurality of combinations. Such combinations are located on the complex plane to have equal conditional probabilities, and such a position is called a QAM constellation diagram. Figure 2 is an example of such a combined distribution map having a size of 16 combinations. Further, each point shown in FIG. 2 is referred to as a constellation point. Moreover, the binary combination below each point is a symbol set at each point, that is, a combination of bits.

In general, the QAM demodulator converts the signals entering the I channel and the Q channel, that is, the received signals specified by α+βi, into the in-situ combination according to the combined profile at the previously agreed position. However, at this time, the received signal is affected by the noise interference, and most of the parallel portions are located at a predetermined position, that is, the parallel portion is located on the combined distribution map, so that the demodulator is restored to the original signal by the noise-changing signal. However, there are some problems in the reliability of the communication, so that the demodulator bears the effect of eliminating noise. Therefore, the channel decoder in the next stage bears these effects, and the more effective and reliable communication system is embodied. However, in order to perform these hard decision processes, that is, bit quantization based on the binary bit detector, the demodulated signal of the continuous value is corresponding to the two levels of discrete signals, and the loss information is lost. Thus, instead of using a bit detector, the similarity measure of the distance between the received signal and the agreed distribution point is changed from hamming to Euclidean distance to obtain additional gain (Gain).

As shown in Figure 1, the modulation is transmitted after the signal encoded by the channel encoder. These are decoded by the channel decoder of the receiver through the soft decision coding process. The demodulator should use the same phase signal component and orthogonality. The received signal formed by the phase signal generates a soft decision value manner corresponding to each channel encoder output bit. These methods are roughly divided into two types, namely, the simple metric procedure proposed by Nokia and the dual minimum metric procedure proposed by Motorola. Both methods calculate the LLR (Log Likelihood Radio) for each output bit. These are used as input soft decision values for the channel decoder.

The simple metric procedure transforms the complex LLR calculation formula into the event equation of the simple form approximation formula. Although the LLR calculation is simple, the approximation formula is used, which causes the LLR to be distorted and degraded according to these properties. In contrast, the dual minimum metric procedure uses a more correct approximation formula to input the calculated LLR to the channel decoder's event algorithm, greatly improving the performance degradation that occurs when using the simple metric procedure, but requires more than the simple metric procedure. The amount of calculations, when embodying hardware, may increase its complexity a lot.

The object of the present invention is to solve the above-mentioned conventional technical problem, a soft decision method for demodulating quadrature amplitude modulation (QAM) received signals composed of in-phase signal components and quadrature phase signal components, from the positive of the received signal The phase component value and the phase component value are calculated by using a function including conditional judgment calculation to calculate respective soft decision values corresponding to hard decision bit positions, that is, distance probability vector values, thereby improving processing speed and saving actual hardware. production cost. In order to perform these processes, the morphology of the existing QAM combination profile and the specific demodulation method according to it are first explained as follows. The combined distribution map of QAM is roughly classified into three types according to the combination of bits set at its distribution point. The first is the form of distribution as shown in Figures 2 to 4, the second is the distribution pattern of Figures 5 to 7, and the third is not included in the scope of this patent.

The morphological features of Fig. 2 are schematically illustrated as follows. When the size of the QAM is 2 ^{2 n} , the number of bits set at each point is 2n, wherein the first half, that is, the distance probability vector value from the 1st bit to the nth bit is according to any of the received signals α or β. A demodulation, the distance probability vector value of the second half to the last 2nth bit of the second half is demodulated according to the remaining received signal, and the same applies to the first half and the second half of the two demodulated equations. That is to say, the first half of the demodulation method is substituted into the second half of the received signal value to obtain the latter half result. (refer to these forms as "first type")

The morphological features of Fig. 5 are schematically illustrated as follows. When the size of the QAM is 2 ^{2 n} , the number of bits set at each point is 2n, and the demodulation method of the distance probability vector of the odd bit is the same as the calculation method of the distance probability vector of the even-numbered bit. However, the received signal value of the distance probability vector for calculating the odd bit here uses either one of α or β according to the combined profile, and the received signal value of the even bit uses the remaining one. That is to say, the demodulation method is the same when calculating the first and second distance probability vectors, except that the received signal values are different. (refer to these forms as "second type")

The present invention replaces the log-likelihood ratio of the soft decision-making demodulation method of the square QAM signal mainly used in industry, and applies the distance probability vector equation to significantly improve the processing speed.

There are two demodulation methods for newly developed square QAM signals, which are respectively classified into a first type and a second type. Embodiments of this are transmitted through the first embodiment and the third embodiment and the second type of the first type. The second embodiment and the fourth embodiment are explained. Moreover, the output range of the final distance probability vector value is between any real number "a" and "-a".

First, let's explain a few basic premise. The size of QAM is determined by Mathematical Formula 1. The number of bits at each point of the distribution map is determined by Mathematical Formula 2.

[Mathematical Formula 1] 2 ^{2 n} -QAM, n=2,3,4...[Mathematical Formula 2] The number of bits set at each point = 2n The number of distance vector values based on these final output values also becomes 2n.

The first method is described in the demodulation method of the square QAM signal of the present invention.

First, a method of softly deciding the first type of square QAM received signal will be described. As illustrated by the features of the first type described above, the first type uses a quadrature phase component (real part or α) or an in-phase signal component (imaginary part or β) in the received signal in order to calculate the distance probability vector of the first half bit combination. ) any of the values. In order to facilitate the understanding of the first half using the beta value, the latter half is demodulated after using the alpha value, and is limited to a value between 1 and -1 according to these output ranges. Further, k is used as a variable of each element order.

The calculation method of the distance probability vector in the first bit of the first type, i.e., k = 1, is represented by Mathematical Formula 3, and these are shown in Figure 5.

[Mathematical Formula 3] □ The first distance probability vector (k=1) Unconditional output value . Here, the value of n is determined according to the QAM size and the above mathematical formula 1.

The calculation method of the distance probability vector of the first type second bit (k=2) is represented by Mathematical Formula 4, and these are shown in Fig. 6.

[Mathematical Formula 4] □ Second Distance Probability Vector (k=2) Unconditional Output Value .

Here, n is the QAM size variable of Mathematical Formula 1, and c is a constant.

The distance probability vector calculation method of the third to nth bits (k=3, 4, ..., n-1, n) of the first type is expressed by Mathematical Formula 5. Here, the distance probability vector of the third bit or more as shown in the ninth figure shows the form of the repeated (V word), and a formula is repeatedly used by these properties.

[Mathematical Formula 5] □ First, the distance probability vector corresponding to each element is distinguished by the basic V-shape form as the (2 ^{k - 3} +1) field.

□ According to the basic formula, the basic formula is .

□ Use the specified β to find the domain to which it belongs, minus the center value m of each field (for example, when k=4, the repeated field is one, and the field is 2 ^{n - 2} ≦|β|<3.2 ^{n - 2} , (|β|-m) whose center value is m=2 ^{n - 1} ) is taken as the new β, which is substituted into the basic formula and determines the output value.

□ Finally, in the field of distinction between the left and the outermost areas, that is, in the field of (2 ^{k - 2} -1) 2 ^{n - k + 2} <|β|, the center value is m = 2 ^{2 n} , (|β |-m) As a new β, substitute into the basic formula and determine the output value.

Here, d is a constant that varies according to the value of k.

The second half of the first type, that is, the distance probability vector calculation method from the n+1th to the 2nth bit, is obtained by α-replacement β in the method of calculating the distance probability vector according to the feature of the first type described above. That is to say, the condition that all β of the mathematical formula 3 is replaced by α becomes the first distance probability vector of the second half, that is, the distance probability vector calculation formula of the n+1th bit.

The second distance probability vector of the second half, that is, the distance probability vector of the n+2th bit is also the mathematical formula 4 for calculating the first half of the second distance probability vector, and the α substitution β can be discriminated, and the distance from the n+3th to the 2ndth bit thereafter The probability vector can be discriminated as described above by modifying the mathematical formula 5.

Thereafter, a method of soft decision-making second type square QAM reception signal will be described. In order to facilitate understanding and to use the alpha value for discriminating the distance probability vector of the odd bit, in order to discriminate the even bit using the beta value and then demodulate, the output range is limited to a value between 1 and -1 according to the first type. Further, k is used as a variable of each element order.

The distance probability vector calculation method for the first bit of the second type (k = 1) is expressed by Mathematical Formula 6, and these are shown in Fig. 9.

[Mathematical Formula 6] The first bit (k=1) unconditional output value . Here, the value of n is determined according to the QAM size and the above mathematical formula 1.

The distance probability vector calculation method of the second type second bit (k=2) is obtained by substituting α for β in Mathematical Formula 6 for calculating the distance probability vector according to the feature of the second type described above.

The distance probability vector calculation method of the second type third bit (k=3) is expressed by Mathematical Formula 7.

[Mathematical Formula 7] If α. When β≧0, the unconditional output value of the third bit (k=3) .

If α. When β<0, the formula is α. All α of the calculation formula at the time of β ≧ 0 is replaced by the formula of β.

Here, n is the QAM size variable of Mathematical Formula 1, and c is a constant.

Divided into α like this. ≧0 and α. The method of calculating the distance probability vector after β<0 is another feature of the second type QAM. These characteristics are applicable when calculating the distance probability vector of the third or more bits of the second type, and the mutual permutation characteristics such as the above-described β permutation to α are also included in this feature.

The distance probability vector calculation method of the fourth bit (k=4) of the second type is to replace α with β and α with α according to the mathematical formula 7 for calculating the third distance probability vector according to the feature of the second type described above. You can get it.

The distance probability vector calculation method for the fifth bit (k=5) of the second type is applied to Mathematical Formula 8. Here, the distance probability vector of the fifth bit (k=5) or more as shown in the tenth figure shows the form of the repeated (V word), and a formula is repeatedly used by these properties. Here, when calculating the distance probability vector of the fifth bit or more, the formula used when calculating the odd-numbered decision value is used according to the characteristic even-numbered value of the second type, and these can only be applied when the size of the QAM is 64 or less. When 256 or more, if the remaining part of the first type is divided into two, the first half and the second half can be calculated.

[Mathematical Formula 8] If α. When β≧0, □ first distinguishes the output map by the basic V-shape, the distance probability vector corresponding to each element is divided into (2 ^{k - 5} +1) fields.

□ According to the basic formula, the basic formula is .

□ Use the specified α to find the domain to which it belongs, minus the center value m of each field (for example, when k=6, the repeated field is one, and the field is 2 ^{n - 2} ≦|α|<3.2 ^{n - 2} , (|α|-m) whose center value is m=2 ^{n - 1} ) is taken as the new α, substituted into the basic formula and determines the output value.

□ Finally, in the field of distinction between the left and the outermost areas, that is, in the field of (2 ^{k - 2} -1) 2 ^{n - k + 2} <|α|, the center value is m = 2 ⁿ , (|α| -m) As a new α, substitute into the basic formula and determine the output value.

If α. When β < 0, the α of the above formula is replaced by β according to the characteristics of the second type described above.

The distance probability vector of the sixth bit of the second type is calculated when the QAM size is 64-QAM, and α is replaced with β in the mathematical formula 8 for calculating the fifth distance probability vector according to the feature of the second type described above, and β is It can be obtained by substituting α. However, when the QAM size is 256-QAM or more, the first half is calculated by dividing the total number of remaining vectors into half, and the second half is the formula for replacing the received value (α or β) in the first half. At this time, the change value in the front office type is only the received value, the bit number (k) is unchanged and the first half of the number is used.

As a result, when the QAM size is larger than 256, the calculation of the distance probability vector of the second type fifth to n+2th bits is determined according to the above mathematical formula 8.

The calculation of the distance probability vector of the n+3th to the 2nth bit of the second type is determined by replacing the variable α of the mathematical formula 8 with β as described above.

Through the received signals of these processes, soft decision demodulation of square QAM is performed using the α+βi value. However, in the method described above, in the method of selecting the received signal and substituting the discriminant formula, in order to facilitate understanding of the arbitrary order, it is more widely applicable in practical application, and the α or β character represented by the formula is based on the combined distribution form of QAM. Inverting any one, whereby the range of output values is also limited not only between a and -a, but also within values of an asymmetrical type such as a and b. These broaden the scope of this invention and increase its significance. Moreover, the above calculation formulas may appear to be complicated, and these are widely applicable, and the popular calculation formulas are known to be very simple by practically applicable embodiments.

- First embodiment -

The first embodiment of the present invention belongs to the first type described above, and is applicable to the first type. In the first embodiment, an example of 1024-QAM having a size of 1024 for QAM is exemplified. The order of selection of the received signals is that β is applied in the first half and α is applied in the second half.

Basically, the QAM according to the two embodiments of the present invention is determined as follows. Mathematical formula 1 determines the size of QAM, and Mathematical Formula 2 shows the number of bits set at each point of the combined profile according to the size of QAM.

[Mathematical Formula 1] 2 ^{2 n} -QAM, n=2, 3, 4... [Mathematical Formula 2] The number of bits set at each point = 2n Basically, the QAM size in the first embodiment of the present invention It is determined according to the following formula. According to these final output values, the number of distance probability vector values also becomes 2n.

Using such mathematical formulas 1 and 2, when n is 5, that is, according to the mathematical formula 1 2 ^{2 * 5} -QAM=1024-QAM, the number of bits set at each distribution point is 2x5= according to the mathematical formula 2 bits. 10 cases. First, when it is determined again that the calculation formula of the first half of the first ten bits is known among the entire ten bits of the feature of the first type before the calculation formula is applied, the calculation formula of the remaining five second half bits can be known.

First, the first distance probability vector is calculated. When k=1, the unconditional output value is Thereafter, the second (ie, k=2) distance probability vector unconditional output value .

Here, c is a constant.

Thereafter, the third (ie, k=3) distance probability vector is calculated as follows, and the basic formula according to the basic form is At this time, it is divided into two fields. If |β|<2 ⁴ , the output value is Other output value .

Then, the fourth (ie, k=4) distance probability vector is calculated as follows, and the basic formula according to the basic form is At this time, it is divided into three fields. If |β|<2 ³ , the output value is ^{, 2 3 ≦ | β | <} 3.2 3 , the output value Other times, the output value is Next, the fifth (ie, k=5) distance probability vector is calculated as follows, and the basic formula according to the basic form is At this time, it is divided into five fields. If |β|<2 ² , the output value is , 2 ² ≦|β|<3.2 ² , the output value is , 3.2 ² ≦|β|<5.2 ² , the output value is , 5.2 ² ≦|β|<7.2 ² , the output value is Other times, the output value is .

Then, the calculation formula of the sixth to tenth distance probability vectors uses a formula in which α is substituted for β in the calculation formula of the first to fifth distance probability vectors according to the feature of the first type.

- Second embodiment -

The second embodiment of the present invention belongs to the second type described above, and the second type is applied. In the second embodiment, a 1024-QAM example in which the size of the QAM is 1024 is exemplified. The order in which the received signals are selected is first applied after selecting α.

The mathematical formula 1 of the first embodiment described above determines the size of the QAM, and the mathematical formula 2 shows the number of bits set at each point of the combined profile according to the size of the QAM.

[Mathematical Formula 1] 2 ^{2 n} -QAM, n=2, 3, 4... [Mathematical Formula 2] The number of bits set at each point = 2n Basically, the QAM size in the second embodiment of the present invention It is determined according to the following formula. According to these final output values, the number of distance probability vector values also becomes 2n.

Using such mathematical formulas 1 and 2, when n is 5, that is, according to the mathematical formula 1 2 ^{2 * 5} -QAM=1024-QAM, the number of bits set at each distribution point is 2x5 according to the mathematical formula 2 =10 case.

First, the calculation formula of the first distance probability vector, when k=1, the unconditional output value .

Thereafter, the calculation formula of the second (ie, k=2) distance probability vector is a form in which the first calculation formula is replaced, and the unconditional output value is .

Thereafter, when the third (ie, k=3) distance probability vector is calculated (1) αβ≧0, the unconditional output value is Here, c is a constant.

(2) When αβ<0, at this time, the method of determining the third distance rate vector output method (αβ≧0) described above is determined by β-substitution α.

Then, when the fourth (ie, k=4) distance probability vector is calculated (1) αβ≧0, the unconditional output value is (2) When αβ<0, at this time, the method of determining the fourth distance probability vector output described above (αβ≧0) is calculated by β-substitution α.

Then, when the fifth (ie, k=5) distance probability vector is calculated by (1) αβ≧0, the basic formula according to the basic form is At this time, it is divided into two fields. If |α|<2 ⁴ , the output value is Other times, the output value is .

(2) When αβ<0, at this time, the method of determining the fifth distance probability vector output described above (αβ≧0) is calculated by β after α.

Then, when the sixth (ie, k=6) distance probability vector is calculated by (1) αβ≧0, the basic formula according to the basic form is At this time, it is divided into three fields. If |α|<2 ³ , the output value is , 2 ³ |α|<3.2 ³ , the output value is , other times, the output value is .

(2) When αβ<0, the equation (αβ≧0) of the sixth distance probability vector output described above is determined in this case, and β is replaced by α.

Next, when the seventh (ie, k=7) distance probability vector is calculated by (1) αβ≧0, the basic formula according to the basic form is At this time, it is divided into five fields. If |α|<2 ² , the output value is , 2 ² |α|<3.2 ² , the output value is , 3.2 ² <|α|<5.2 ² , the output value is , 5.2 ² <|α|<7.2 ² , the output value is , other times, the output value is .

(2) When αβ<0, the equation of the method (αβ≧0) for determining the output of the seventh distance probability vector described above is determined by β after α.

The method of calculating the eighth to tenth distance probability vectors is obtained by substituting α for β and β for α by calculating the fifth to seventh distance probability vectors.

The connection embodiment of the present invention is described in detail. However, this is only for the purpose of carrying out the embodiments, and is not intended to limit the invention, and it is obvious to those skilled in the art that the modifications within the scope of the invention are readily understood.

The invention replaces the log-likelihood ratio of the soft QD signal of the square QAM signal mainly used in industry, and applies the distance probability vector equation to significantly improve the processing speed, and saves the manufacturing cost when the hardware is actually embodied.

Figure 1 is a block diagram of a general digital communication system.

2 is a combined distribution diagram of a soft decision demodulation method according to a first embodiment of the present invention.

3 and 4 are diagrams showing the bit distribution in the combined profile of Fig. 2.

Figure 5 is a combination of a soft decision-making demodulation method of a second embodiment of the present invention.

6 and 7 are diagrams showing the bit distribution in the combined profile of Fig. 5.

FIG. 8 is a functional configuration diagram illustrating a distance probability vector decision process according to the present invention.

Figure 9 is an output diagram of each distance probability vector of the first type 1024-QAM.

Figure 10 is an output diagram of each distance probability vector of the second type 1024-QAM.

Claims (14)

A soft decision demodulation method for quadrature amplitude modulation, comprising: a soft decision demodulation method for square quadrature amplitude modulation (QAM) received signals formed by in-phase signal components and quadrature phase signal components, orthogonal to received signals The phase component value and the in-phase component value are used to calculate the soft decision value corresponding to the hard decision bit position, that is, the distance probability vector value, and the distance probability vector and the decision of the first half among the entire bit. The method of the distance probability vector of the remaining half-bit is the same, and the orthogonal phase component value and the in-phase component value are respectively determined, and the demodulation method of the odd-bit distance probability vector and the distance probability vector of the even-numbered bit are respectively determined. The calculation method is the same, and the received signal value of the odd bit distance probability vector is calculated using any one of α or β according to the specified combination profile, and the remaining signal value of the even bit is used. A soft decision demodulation method for quadrature amplitude modulation as described in claim 1 wherein the first distance probability vector selects any one of α or β according to the form of the combined profile and is determined according to the following formula: Unconditional output value Where Ω is the selected received value, that is, any one of α or β , n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, and α is any real number determined according to the output range. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the second distance probability vector is calculated in the first distance of the second type In the rate vector method, the received value is not selected and the selected received value is replaced and calculated. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the third distance probability vector selects any one of the received values α or β according to the form of the combined profile, αβ 0, using the following formula, when αβ <0, the received value selected by the following formula is replaced with the unselected received value and determined: unconditional output value Where Ω is the selected received value, n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, α is an arbitrary real number determined according to the output range, and c is an arbitrary constant. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the fourth distance probability vector is in the method of calculating the third distance probability vector of the second type, αβ When 0 and αβ < 0, the selected received values are respectively replaced with the received values that are not selected and calculated. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the fifth distance probability vector selects any one of the received values α or β according to the form of the combined profile, the above αβ When 0, the following formula is used. When αβ < 0, the selected received value is replaced by the received value that is not selected by the following formula: First, the distance probability vector corresponding to each element is distinguished when the output map is distinguished by the basic V-shape. For two areas; the basic formula according to the basic form is Find the domain to which you belong with the specified Ω, subtract (|Ω|-m) of the center value m of each field as the new Ω, substitute it into the basic formula and determine the output value; distinguish between the left and right outermost areas of the field. That is, in the field of 7.2 ^n-3 <|Ω|, the center value is m= ²ⁿ , and (|Ω|-m) is taken as the new Ω, which is substituted into the basic formula and determines the output value; Is the selected received value, n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, d is a constant, and α is a constant that determines the output range. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the sixth distance probability vector is in the method of calculating the fifth distance probability vector of the second type, wherein the size of the QAM is 64-QAM , αβ When 0 and αβ < 0, the selected received values are respectively replaced with the received values that are not selected and calculated. The soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein when the size of the QAM is 256-QAM or more, the distance probability vector from the fifth to the n+2 is according to the form of the combined distribution map. Select any one of the received values α or β , αβ 0, using the following formula, when αβ < 0, the selected received value is replaced with the received value that is not selected by the following formula: First, the distance probability vector corresponding to each element when the output map is distinguished by the basic V-shaped form is first determined. Divided into (2 ^k-3 +1) fields; the basic formula according to the basic form is Use the specified Ω to find the domain to which it belongs, minus the center value m of each field (for example, when k=6, the field to be repeated is one, this field is 2 ^n-2 ≦|Ω|<3.2 ^n-2 , center The value Ω|-m with m=2 ^n-1 ) is taken as the new Ω, which is substituted into the basic formula and determines the output value; the field is located in the outermost left and right areas, that is, at (2 ^k-2 -1) 2 ^n-k+2 <|Ω| field, the center value is m=2 ⁿ , |Ω|-m is taken as a new Ω, substituted into the basic formula and determines the output value; where k is the distance probability vector The number (k=5,6,...,n), Ω is the selected receiving value, n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, α is any real number set according to the output range, and d is changed according to the value of k. Constant. A soft decision demodulation method for quadrature amplitude modulation according to claim 1, wherein the distance probability vector αβ from the n+ 3th to the 2nd is when the size of the QAM is 256-QAM or more. When 0 is used, the received value which is not selected when the fifth to n+2 distance probability vectors of the second type are determined is used. When αβ < 0, the selected received value is obtained by replacing the selected received value. A soft decision demodulation method for quadrature amplitude modulation, comprising: a soft decision demodulation method for square quadrature amplitude modulation (QAM) received signals formed by in-phase signal components and quadrature phase signal components, orthogonal to received signals The phase component value and the in-phase component value are used to calculate the soft decision value corresponding to the hard decision bit position, that is, the distance probability vector value, and the distance probability vector and the decision of the first half among the entire bit. The method of the distance probability vector of the remaining lower half is the same, and the orthogonal phase component value and the in-phase component value are respectively determined, and the distance probability vector value from the 1st bit to the nth bit is determined according to the received signal or Any one of the demodulation, the distance probability vector of the second half of the last nth bit to the last 2nth bit is demodulated according to the remaining received signal, and the first half and the second half of the applicable equations are the same. A distance probability vector is selected according to the form of the combined profile to select any one of the received values α or β according to the following formula: the unconditional output value is Where Ω is any one of the selected received values α or β , α is any real number set according to the output range. A soft decision demodulation method for quadrature amplitude modulation, comprising: a soft decision demodulation method for square quadrature amplitude modulation (QAM) received signals formed by in-phase signal components and quadrature phase signal components, orthogonal to received signals The phase component value and the in-phase component value are used to calculate the soft decision value corresponding to the hard decision bit position, that is, the distance probability vector value, and the distance probability vector and the decision of the first half among the entire bit. The method of the distance probability vector of the remaining lower half is the same, and the orthogonal phase component value and the in-phase component value are respectively determined, and the distance probability vector value from the 1st bit to the nth bit is determined according to the received signal or Any one of the demodulation, the distance probability vector of the second half of the last nth bit to the last 2nth bit is demodulated according to the remaining received signal, and the first half and the second half of the applicable equations are the same. The two-path probability vector is determined according to the received value selected when determining the first distance probability vector and the following formula: the unconditional output value is Where Ω is the selected received value, n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, α is an arbitrary real number determined according to the output range, and c is an arbitrary constant. A soft decision demodulation method for quadrature amplitude modulation, comprising: a soft decision demodulation method for square quadrature amplitude modulation (QAM) received signals formed by in-phase signal components and quadrature phase signal components, orthogonal to received signals The phase component value and the in-phase component value are used to calculate the soft decision value corresponding to the hard decision bit position, that is, the distance probability vector value, and the distance probability vector and the decision of the first half among the entire bit. The method of the distance probability vector of the remaining lower half is the same, and the orthogonal phase component value and the in-phase component value are respectively determined, and the distance probability vector value from the 1st bit to the nth bit is determined according to the received signal or Any one of the demodulation, the distance probability vector of the second half of the last nth bit to the last 2nth bit is demodulated according to the remaining one of the received signals, and the two methods of the first half and the second half of the applicable equations are the same. The third to nth distance probability vectors are determined according to the received value selected when determining the first distance probability vector and the following formula: first, the basic V-shaped form region is used. Members corresponding to FIG output from the probability vector elements divided into (2 ^k-3 +1) when the art; The basic form of the basic formula Find the domain to which you belong with the specified Ω, subtract (|Ω|-m) of the center value m of each field as the new Ω, substitute it into the basic formula and determine the output value; distinguish between the left and right outermost areas of the field. That is, in the field of (2 ^k-2 -1) 2 ^n-k+2 <|Ω|, the center value is m=2 ⁿ , (|Ω|-m) is used as the new Ω, and is substituted into the basic formula. And determine the output value; where Ω is the selected received value, n is the size of QAM, that is, the variable of 2 ²ⁿ is determined, k is the number of the distance probability vector (k=3, 4,..., n), and d is according to k The constant of the value change, α is a constant that determines the output range. A soft decision demodulation device for quadrature amplitude modulation (QAM) received signal, comprising: a soft decision demodulation device for receiving signals by square quadrature amplitude modulation (QAM) comprising an in-phase signal component and a quadrature phase signal component, including Calculating the distance probability vector decision unit of the distance softness decision vector value from the orthogonal phase component value and the in-phase component value of the received signal by using a function including the conditional judgment calculus, corresponding to each soft decision value of the hard decision bit position The probability vector decision unit determines the distance probability vector of the first half and the distance probability vector for determining the remaining second half of the entire bit, and replaces the orthogonal phase component value and the in-phase component value, respectively, and determines the distance probability. The distance probability vector value of the vector calculation unit from the 1st bit to the nth bit is demodulated according to any one of the received signals, and the distance probability vector of the second half of the last nth bit to the last 2nth bit is left according to the remaining A received signal demodulation, the first half and the second half of the equations applicable to the two demodulations are the same, wherein the first distance probability vector is based on the combination FIG form after selecting any one from among the received value α or β determined according to the formula: output value is unconditionally Where Ω is any one of the selected received values α or β , α is any real number set according to the output range. A soft decision demodulation device for quadrature amplitude modulation (QAM) received signal, comprising: a soft decision demodulation device for receiving signals by square quadrature amplitude modulation (QAM) comprising an in-phase signal component and a quadrature phase signal component, including Calculating a distance probability vector decision unit for each soft decision value corresponding to a hard decision bit position, that is, a distance probability vector value from a quadrature phase component value and an in-phase component value of the received signal, using a function including a conditional judgment calculus, The demodulation calculation of the distance probability vector of the odd-numbered bit of the distance vector calculation unit is the same as the calculation of the distance probability vector of the even-numbered bit, and the received signal value of the odd-bit distance probability vector is calculated according to the specified combination map using α or Any one of β , the received signal value of the even bit uses the remaining one.