KR100433028B1 - A demodulation method using soft decision for quadrature amplitude modulation and an apparatus thereof - Google Patents

Abstract

PURPOSE: A soft decision method for demodulating a square QAM reception signal and a soft decision apparatus thereby are provided to improve the processing speed and reduce the manufacturing cost by using a condition probability vector equation method instead of a log likelihood ratio method. CONSTITUTION: A soft decision method for demodulating a square QAM reception signal includes a process for demodulating a square QAM reception signal, which is formed with an in-phase signal element and an orthogonal-phase signal element. The soft decision method includes a process for calculating a condition probability vector value as a soft decision value corresponding to a bit position of the hard decision by using a function including a conditional decision operation from an in-phase element value and an orthogonal-phase element value of the received signal.

Description

A demodulation method using soft decision for quadrature amplitude modulation and an apparatus

The present invention relates to soft crystal demodulation of an orthogonal amplitude modulated signal (hereinafter referred to as QAM). In particular, the present invention relates to a soft crystal demodulation method that improves the processing speed of a soft crystal by using a predetermined function and pattern in demodulating a received signal. It is about.

The QAM method changes the amplitude of a given carrier so that it can carry more signals, or bits, implied at one wavelength. The carrier uses cosine and sine waves, and the two waveforms are orthogonal to each other. So that there is no interference between them. In mathematical terms, it can be expressed as two numbers that do not interfere with each other: real and imaginary numbers. In other words, in the complex α + βi , the change in the value of α does not affect the value of β . For this reason, the cosine wave corresponds to α and the sine wave corresponds to β . Cosine waves are generally referred to as I-channels, and sine waves are referred to as Q-channels.

By combining the amplitudes of these two waves together to form a number of combinations, these combinations are placed on the complex plane with equal conditional probability, and the promise of these positions is called the QAM constellation diagram. Figure 1 shows an example of such a distribution diagram and the size can be seen that represents 16 combinations. In addition, each point that can be seen in Figure 1 refers to the distribution point (constellation point). Also, the combination of binary numbers written below each distribution point is called a symbol set at each point, that is, a bundle of bits.

In general, the QAM demodulator converts a signal coming into the I channel and the Q channel, that is, a received signal given by α + β i, into an original bit bundle according to the aforementioned previously weakened position, that is, a combination distribution. In this case, however, the received signal is not located at a predetermined position, that is, a combination distribution, in most cases due to the influence of noise interference, and for this reason, the demodulator must restore the signal changed by the noise to the original signal. However, it is often difficult to secure communication reliability for the demodulator to play such a role of noise canceling, and thus it is possible to implement a more effective and reliable communication system by transferring this role to the next stage, a channel decoder. However, in order to perform this process, as in a hard decision, bit quantization performed by a binary bit detector has a loss of information by mapping a demodulation signal having a continuous value to a two-level discrete signal. Without using a detector, an additional coding gain can be obtained by changing the similarity measure for the distance between the received signal and the promised distribution point from the Hamming distance to the Euclidean distance. .

As shown in FIG. 13, in order to modulate and transmit a signal encoded by a channel encoder, and to decode it through a soft decision coding process in a channel decoder of a receiver, a demodulator and a quadrature signal component are used. It must have a method of generating soft decision values corresponding to each output bit of the channel encoder from the received signal. There are two main ways of doing this: Nokia's simple metric procedure and Motorola's dual minimum metric procedure, both of which have output bits. We calculate the log likelihood ratio (LLR) for and use it as the input soft decision value of the channel decoder.

The simple matrix method is a mapping algorithm that transforms a complex LLR equation into a simple approximation equation. The LLR calculation is simple, but the degradation of performance due to LLR distortion caused by using the approximation equation is pointed out as a disadvantage. On the other hand, the dual minimum matrix method is a mapping algorithm that uses the LLR calculated using a more accurate approximation as the input of the channel decoder, and has the advantage of significantly improving the performance degradation caused by using the simple matrix method. It requires more computational complexity and has a problem that a significant increase in complexity is expected in hardware implementation.

An object of the present invention is to solve the problems of the prior art mentioned above, in the soft decision method for demodulating a quadrature amplitude modulation (QAM) received signal consisting of an in-phase signal component and a quadrature signal component From the quadrature and in-phase component values of the received signal, a condition probability vector value is obtained, which is the respective soft decision value corresponding to the bit position of the hard decision, using a function including a condition determination operation. As a result, the processing speed and the production cost of the actual hardware can be expected. In order to carry out this process, the following is a description of the conventional QAM combination distribution and its characteristic demodulation method. Combination distribution of QAM can be classified into three types according to the arrangement of bit bundles set at the distribution points. The first is distributed form as shown in FIG. 1, the second is distributed form as shown in FIG. 6, and the third is not included in the scope of this patent.

Features of the form shown in FIG. 1 can be summarized as follows, as seen in FIGS. 2 and 3. If the size of the QAM is 2 ²ⁿ , the number of bits set at each point is 2n. Among them, the conditional probability vector values corresponding to the first half, that is, bits 1 to n are determined by one of the received signals α and β . The conditional probability vector values corresponding to the demodulation and the last n + 1 th bit to the last 2 n th bit are demodulated by the other received signal, and the equations applied to the demodulation are the same in the first half and the second half. In other words, the result of the latter half can be obtained by substituting the received signal value corresponding to the latter half into the first half demodulation method. (This form is referred to as 'type 1' for convenience.)

The features of the form shown in FIG. 6 can be summarized as follows, as seen in FIGS. 7 and 8. If the size of QAM is 2 ²ⁿ , the number of bits set at each point is 2n, and the method of demodulating the conditional probability vector corresponding to the odd numbered bits is identical to the method of calculating the conditional probability vector corresponding to the next even numbered bits. do. Here, the received signal value for calculating the condition probability vector corresponding to the odd-numbered bits uses either α or β according to the given combination distribution, and the received signal value for the even-numbered bits uses the other. In other words, for the first and second conditional probability vector calculations, the demodulation method is the same and only the values of the received signals are different. (This type is referred to as 'type 2' for convenience.)

Other objects, features and advantages of the present invention will become apparent from the following detailed description of embodiments taken in conjunction with the accompanying drawings.

1 is a diagram illustrating a constellation point for explaining a soft crystal demodulation method according to a first embodiment of the present invention.

2 and 3 are diagrams for explaining the bit distribution in the combination distribution diagram shown in FIG.

4 is a diagram showing a function applied to a first probability vector of the first embodiment of the present invention.

5 is a diagram illustrating a function applied to a second probability vector of the first embodiment of the present invention.

FIG. 6 is a diagram illustrating a constellation point for explaining a soft crystal demodulation method according to a second embodiment of the present invention. FIG.

7 and 8 are diagrams for explaining the bit distribution in the combination distribution diagram shown in FIG.

9 is a diagram showing a function applied to a first probability vector of the second embodiment of the present invention.

10 is a diagram showing a function applied to a second probability vector of the second embodiment of the present invention.

11 is a functional block diagram showing a condition probability vector determination process according to the present invention.

12 is a diagram showing the hardware configuration for soft decision of type 1 64-QAM according to the present invention;

13 is a block diagram illustrating a general digital communication system.

The present invention significantly improves the processing speed by applying the condition probability vector equation instead of the log likelihood ratio method, which is a soft decision demodulation method of a square QAM signal, which is mainly used in the industry. The demodulation method of the newly developed square QAM signal will be divided into the case of the first type and the second type, and an example thereof will be described through the first and second embodiments. Also, the final conditional probability vector value output range is output between any real number a and -a.

First of all, before we go into the explanation, some basic premise will be explained. The size of QAM can be characterized by Equation 1, and the number of bits set at each point of the distribution can be characterized by Equation 2. have.

[Equation 1]

2 ^2n- QAM, n = 2,3,4... …

[Equation 2]

Number of bits set at each point = 2n

As a result, the number of conditional probability vector values as the final output value is 2n.

First, a method of flexibly determining the received signal of the square QAM corresponding to the first type will be described. In the case of the first type, as mentioned in the description of the features of the first type, it is said that either the real part or the imaginary part of the received signal is used to calculate the condition probability vector corresponding to the first bit combination. for convenience, the first half and the second half demodulate using the value α β value range of the output, consequently, its understanding will forward to a value between -1 and 1 for convenience.

A method of calculating the condition probability vector corresponding to the first bit in the first type may be represented by Equation 3 and visualized this is shown in FIG. 4.

[Equation 3]

Β │ 2 ⁿ −1, the output is determined by sign ( β ).

Or ② | β | ≤ 1, the output is determined to be 0.9375 * β .

Or ③ If 1 <│ β | ≤ 2 ⁿ -1, the output is sign ( β ) It is determined by (│ β │-1) + 0.9375 * sign ( β ). Here, sign ( β ) means the sign of β value.

A method of calculating the condition probability vector corresponding to the second bit in the first type may be represented by Equation 4 and visualized this is shown in FIG. 5.

[Equation 4]

① 2 if ^{n -2 n (2-m)} ≤│ β │≤2 n -2 n (2-m) +1, the output is determined as (-1) ^{m + 1.}

Or 2 ^n-1 −1 ≦ β β <2 ^n-1 +1, the output is determined to be 0.9375 (2 ^n-1 − β ).

Or ③ 2 ^n-1 -2 ^{(n-1) (2-m)} + m ≤│ β | <2 ⁿ -2 ^{(n-1) (2-m)} + m -2,

The output is It is determined that (│ β │-2m + 1) +0.9735 (-1) ^{m + 1} +0.0625.

Here, m = 1 or 2, and n is a magnitude variable of QAM in Equation 1.

In the first type, a method of calculating a condition probability vector corresponding to the third to n−1th bits may be represented by Equation 5.

[Equation 5]

If m * 2 ^{n-k + 2} −1 ≤ β β ≤ m * 2 ^{n-k + 2} +1, the output is determined to be (-1) ^{m + 1} .

Or (2 L-1) * 2 ^{n-k + 1} -1 &lt; = beta | ≤ (2 L-1) * 2 ^{n-k + 1} +1,

The output is determined by (-1) ^{l + 1} 0.9375 {│ β |-(2ℓ-1) * 2 ^{n-k + 1} }.

Or ③ (P-1) * 2 ^{n-k + 1} +1 &lt; │ β | ≤ P * 2 ^{n-k + 1} -1,

The output depends on the value of P. If P is odd,

[(-1) ^{((p + 1) / 2) +1} * │ β │ + (-1) ^{(p + 1) / 2} {(P-1) * 2 ^{n-k + 1} +1}] + It is determined by (-1) ^{(p + 1) / 2} .

But if the value of P is even,

[(-1) ^{p / 2 + 1} * | β | + (-1) ^{p / 2} (P * 2 ^{n-k + 1} -1)] + (-1) ^{P / 2 + 1} .

Where m = 0,1..2 ^k-2 , and l = 1, 2, ..., 2 ^k-2 and P = 1, 2, ..., 2 ^k-1

Where k is a bit number and is an integer of 3 or more.

In the first type, a method of calculating the condition probability vector corresponding to the n th bit, the last bit of the first half, may be represented by Equation 6. This is a special case of Equation 5 above, where k = n and only the conditional expressions of ① and ② are applied.

[Equation 6]

① If ^{m * 2 2 -1 ≤ │ β} │≤m * 2 2 +1, the output is determined as (-1) ^{m + 1.}

Or (2ℓ-1) * 2 ¹ -1 <| β | | (2ℓ-1) * 2 ¹ +1,

The output is determined by (-1) ^{l + 1} 0.9375 {│ β |-(2ℓ-1) * 2 ¹ }.

Where m = 0,1..2 ^n-2 and l = 1, 2, ..., 2 ^n-2

The method of calculating the condition probability vector corresponding to the latter bits of the first type, that is, the bit numbers n + 1 to 2n, is obtained by substituting β by α in the method of obtaining the first conditional probability vector according to the characteristics of the first type. Can be. In other words, the condition in which all βs in Equation 3 are replaced with α becomes a conditional probability vector calculation equation corresponding to the first conditional probability vector, that is, the n + 1 th bit. The condition probability vector corresponding to the second + 2nd conditional probability vector, the second condition probability vector, can also be determined by substituting β with α in Equation 4, which is the condition for calculating the second conditional probability vector of the first half. The condition probability vectors corresponding to the bit numbers n + 3 to 2n can be determined by modifying Equations 5 and 6 as described above.

Next, a method of flexibly determining the received signal of the square QAM corresponding to the second type will be described. For ease of understanding, we will use α to determine the condition probability vector corresponding to the odd bits and β to determine the even bits.

A method of calculating the condition probability vector corresponding to the first bit in the second type may be represented by Equation 7 and visualized this is shown in FIG. 9.

[Equation 7]

If α 1 -2 ⁿ -1, the output is determined by -sign ( α ).

Or ⓑ | α | ≤ 1, the output is determined to be -0.9375 * α .

Or ⓒ 1 <| α | ≤ 2 ⁿ -1, the output is -sign ( α ) { (│ α │-1) + 0.9375}. Here, sign ( α ) means sign of α value.

The method of calculating the condition probability vector corresponding to the second bit in the second type can be obtained by substituting β for all α in Equation 7, which calculates the first condition probability vector according to the feature of the second type.

A method of calculating the condition probability vector corresponding to the third bit in the second type may be represented by Equation (8).

[Equation 8]

If α x β ≥ 0

If 2 ⁿ -2 ^{n (2-m)} ≤│ α │2 ⁿ -2 ^{n (2-m)} +1, the output is determined as (-1) ^m .

Ⓑ or 2 when ^{n-1 -1≤│ α │ <2} n-1 +1, the output is determined as ^{0.9375 (│ α │-2 n} -1).

Or ⓒ 2 ^n-1 -2 ^{(n-1) (2-m)} + m≤│α│ <2 ⁿ -2 ^{(n-1) (2-m)} + m-2,

The output is It is determined that (│ α │-2m + 1) + 0.9735 (-1) ^m -0.0625.

If α x β <0, the formula is determined by substituting β for all α in the formula for α x β ≧ 0.

Thus, the method of obtaining the condition probability vector by dividing the case of α × β ≧ 0 and the case of α × β <0 may be another feature of the second type QAM. This feature is always applied when obtaining the condition probability vector corresponding to the third or more bits of the second type, and the intersubstitutive properties such as substituting β by α are also included in this feature.

The fourth case conditional probability vector than the expression QAM is 64-QAM size to obtain corresponding to the bit of the first equation (8) expression, to obtain the third conditional probability vector by the features of the second type α of the second type to the β, If each substituted in the β α can be obtained. However, if the size of the QAM is more than 256-QAM is represented by the equation (9).

[Equation 9]

If m * 2 ^{n-k + 3} -1 &lt; α ^{m &lt}; ^m * 2 ^{n-k + 3} +1, the output is determined to be (-1) ^{m + 1} .

Or ⓑ (2 L-1) * 2 ^{n-k + 2} -1 &lt; α a &lt; (2 L-1) * 2 ^{n-k + 2} +1,

The output is determined by (-1) ^{l + 1} {0.9375 | α | -0.9375 (2 ^l -1) * 2 ^{n-k + 2} }.

Or ⓒ (P-1) * 2 ^{n-k + 2} +1 <│ α | ≤ P * 2 ^{n-k + 2} -1,

The output depends on the value of P. If P is odd,

[(-1) ^{((p + 1) / 2) +1} * │ α │ + (-1) ^{(p + 1) / 2} {(P-1) * 2 ^{n-k + 2} +1}] + (-1) determined by ^{(p + 1) / 2} ,

If P is even,

It is determined as [(-1) ^{p / 2 + 1} * | α | + (-1) ^{p / 2} (P * 2 ^{n-k + 2-1} )] + (-1) ^{p / 2 + 1} .

Where k is the bit number and m = 0, 1,... , 2 ^k-3 , L = 1,2... , 2 ^k-3 , p = 1,2... , 2 ^k-2

The conditional probability vector corresponding to the fifth bit of the second type is represented by Equation 10 when the QAM size is 64-QAM, and when Equation 9 is applied when the QAM size is 256-QAM or more. do.

[Equation 10]

If α x β ≥ 0

If m * 2 ² -1 〈│ β │ ≤ m * 2 ² +1,

The output is determined by (-1) ^{m + 1} .

Ⓑ (2ℓ-1) * 2 ² -1 <│β│≤ (2ℓ-1) * 2 ² +1,

The output is determined by 0.9375 (-1) ^{l + 1} {| β |-(2ℓ-1) * 2 ² }.

Where m = 0, 1, 2 and l = 1, 2.

α x β if <0 is obtained when the substituted β of ⓐ and ⓐ expression according to the characteristic of the first type to α 2.

Of vector conditional probability that corresponds to the sixth bit of the second type calculation is β when the magnitude of QAM is 64-QAM the α of the first expression, equation 10 to obtain the fifth conditional probability vector by the features of the second type to the β It is obtained by substituting α with each other. However, if the size of the QAM is more than 256-QAM is represented by the equation (9).

The calculation of the condition probability vector corresponding to the seventh to nth bits of the second type is determined by Equation 9.

The calculation of the conditional probability vector corresponding to the n + 1 th bit of the second type is represented by equation (11), which is a special case of equation (9).

[Equation 11]

If m * 2 ² −1 ≤ α α ≤ m * 2 ² +1, the output is determined to be (-1) ^{m + 1} .

Or ⓑ (2ℓ-1) * 2 ¹ -1 <| α | | (2ℓ-1) * 2 ¹ +1,

The output is determined by (-1) ^{l + 1} {0.9375 | α | -0.9375 (2 ^l -1) * 2 ¹ }.

m = 0, 1,... 2 ^n-2 , L = 1,2... , 2 ^n-2

The conditional probability vector corresponding to the n + 2th type of the second type is obtained by substituting α for β and β for α in Equation (8).

Calculation of the conditional probability vector corresponding to n + 3 th to 2n-1 th of the second type can be obtained by substituting α in Equation 9 with β . However, the value of bit number k used at this time is assigned from 4 to n instead of n + 3 to 2n-1.

The 2nth conditional probability vector of the second type is obtained by substituting α in Equation 11 above with β .

Through the above process, the soft crystal demodulation of the square QAM is possible by using the received signal, that is, α + β i . However, the above-described method is arbitrarily ordered for better understanding in the method of selecting the received signal and substituting it into the discriminant, but in actual application, the character of α or β expressed in the formula is applied more universally. It should be noted that can be reversed as much as possible depending on the combination distribution of QAMs, and the range of output values can be asymmetric, such as between a and b, not just between a and -a. This broadens the versatility of the present invention and increases its significance. In addition, the above-described formulas may seem very complicated, but since this is a generalized formula for general application, it can be seen that the formula is a very simple formula through the embodiment.

First Embodiment

The first embodiment of the present invention corresponds to the first type, and the features of the first type are applied. In the first embodiment, 1024-QAM having a size of QAM is 1024. To select the order of the received signal, apply β to the first half and α to the second half.

Basically, the QAM in the two embodiments according to the present invention is determined by the following equation. Equation 1 determines the size of the QAM and Equation 2 represents the number of bits set at each point of the combined distribution according to the size of the QAM.

[Equation 1]

2 ²ⁿ -QAM, n = 2,3,4

[Equation 2]

Number of bits set at each point = 2n

Basically, the QAM size in the first embodiment of the present invention is determined by the following equation, and accordingly, the number of condition probability vector values, which are final output values, is 2n.

When n is 5 using Equations 1 to 2, that is, 2 ^{2 * 5} -QAM = 1024-QAM according to Equation 1, and the number of bits set in each distribution point is 2x5 = 10 bits according to Equation 2 Will be described. First of all, before entering the formula application, it should be recalled that if the formula for the first five bits of the total ten bits is known by the characteristics of the first type, the formula for the last five bits can also be known immediately.

First of all, the first conditional probability vector is

① | β |> 2 ⁵ -1, the output is determined by sign ( β ).

Or 2 | β | 1, the output is determined to be 0.9375 * β .

Or ③ 1 <│ β is │≤2 ⁵ -1, the output is sign (β) { (│ β │ -1) + 0.9375}.

The second conditional probability vector is k = 2, m = 1,2.

If 0≤│ β │≤1, the output is determined as 1,

Or 2 ⁵ -1≤│ β | ≤2 ⁵ , the output is -1.

Or 2 ⁴ −1 ≦ β β <2 ⁴ +1, the output is determined to be 0.9375 (2 ⁴ − β β ).

Or 1≤│ β │ <2 ⁴ -1,- (│ β │-1) + 1,

If 2 ⁴ +1 ≤│ β │ <2 ⁵ -1,- (│ β │-3) -0.825 is determined

The third (ie k = 3, m = 0,1,2, l = 1,2, p = 1,2,3,4) conditional vector

If m * 2 ⁴ -1 ≤ β ≤ m * 2 ⁴ +1, the output is determined to be (-1) ^{m + 1} .

In this case, substituting m = 0,1,2,

If -1≤│ β │≤1, it is determined to be -1.

Or ²⁴ is -1≤│ β │≤2 ⁴ +1, is determined to one.

Or 2 ⁵ -1≤│ β | ≤2 ⁵ +1, it is determined as -1.

Or (2ℓ-1) * 2 ³ -1 <| β | ≤ (2ℓ-1) * 2 ³ +1,

The output is substituting l = 1,2 at (-1) ^{l + 1} 0.9375 {│ β |-(2ℓ-1) * 2 ³ }

If 2 ³ -1 &lt; β │ ≤ ^{2 3} + 1, the output is determined to be 0.9375 (│ β │-2 ³ ),

Or 3 * 2 ³ -1 &lt; beta &amp;le;&amp;le; ³ * 2 ³ +1, the output is determined to be -0.9375 ( beta --3 * 2 ³ ).

Or ③ (P-1) * 2 ³ +1 <| β | ≤ P * 2 ³ -1 and substituting P = 1,2,3,4 depending on whether P is odd or even

1 <│ is β │≤2 ³ -1, the output is (│ β │-1) -1,

Or 2 ³ +1 <| β | <2 ⁴ -1, the output is (│ β │-2 ⁴ + 1) + 1,

Or 2 ⁴ +1 <| β | < ³ * 2 ³ -1, the output is (2 ⁴ + 1-│ β │) + 1,

Or 3 * 2 ³ +1 <| β | <2 ⁵ -1, the output is Is determined by (2 ⁵ + 1-│ β |) -1.

Then fourth (ie k = 4, m = 0,1,2,3,4, l = 1,2,3,4 p = 1,2,3,4,5,6,7,8). ) The formula for the conditional probability vector is

If -1 ≤ β ≤ 1, the output is determined to be -1.

Or 2 if ³ -1≤│ β │≤2 ³ +1, the output is determined to one.

Or 2 is ⁴ -1≤│ β │≤2 ⁴ +1, the output is determined as -1.

Or 3 * 2 ³ −1 ≦ β β ³ * 2 ³ +1, the output is determined to be 1.

Or 2 ⁵ -1≤│ β | ≤2 ⁵ +1, it is determined as -1.

If or ^{2 2 -1 <│ β │≤2 2} +1, the output is ^{0.9375 {│ β │-2 2} } are determined,

Or 3 * 2 ² -1 &lt; beta &amp;le;&amp;le; 3 * 2 ² +1, the output is determined to be -0.9375 { beta -3 * 2 ² }.

Or 5 * ² 2 -1 <│ is β │≤5 * ² 2 +1, the output is determined as 0.9375 {│ β │-5 * 2 2}.

Or 7 * 2 ² -1 &lt; beta &lt; 7 * 2 ² +1, the output is determined to be -0.9375 { beta -7 * 2 ² }.

Or 1 <| β | ≤ ² 2-1, the output is {│ β │-1} -1,

Or 2 ² +1 <| β | < ^{2 3} -1, the output is {│β│-2 ³ +1} +1,

Or 2 ³ +1 <│ β | ≤3 * 2 ² -1, the output is {2 ³ + 1-│ β │} +1,

Or 6 * 2 ² +1 <│ β | ≤7 * 2 ² -1, the output is {6 * 2 ² + 1-│ β |} +1.

Or 7 * 2 ² +1 <| β | ≤ ^{2 5} -1, the output is {2 ⁵ -1-│β│} -1.

Then the fifth (that is, k = 5, m = 0, 1, 2 ... 7, 8, l = 1, 2, 3 ... 7, 8).

If -1 ≤ β β ≤ 1, the output is determined to be -1.

Or 2 ² −1 ≦ β β ≦ ^{2 2} + l, the output is determined to be 1.

Or 3 * 2 ² −1 ≦ β β 3 * 2 ² + l, the output is determined to be -1.

Or if ^{7 * 2 2 -1≤│ β │≤7 *} 2 2 + l, the output is determined to one.

Or 2. If the ⁵ -1≤│ β │≤2 ⁵ +1, the output is determined as -1.

Or 1 if <│ β │ <3, the output is determined as 0.9375 (│ β │-2) ,

Or 5 <If │ β │ <7, the output is determined as -0.9375 (│ β │-6) .

Or 9 <│ β │ <11, the output is determined to be 0.9375 (│ β │-10),

Or 25 &lt; β &lt; 27, the output is determined to be 0.9375 (│ β -26).

Or 29 is <│ β │ <31, the output is determined as -0.9375 (│ β │-30) .

Then, from 610th condition calculation of the probability vector is use of the equation for replacing the β to α in the fifth conditional probability vector calculation, from the first, depending on the nature of the first type.

Second Embodiment

The second embodiment of the present invention corresponds to the second type, and the features of the second type are applied. In the second embodiment, 1024-QAM having a QAM size of 1024 will be given. In order to select the received signal, α is first selected and applied.

As in the first embodiment, Equation 1 determines the size of the QAM and Equation 2 represents the number of bits set at each point of the combined distribution according to the size of the QAM.

[Equation 1]

2 ²ⁿ -QAM, n = 2,3,4

[Equation 2]

Number of bits set at each point = 2n

Basically, the QAM size in the second embodiment of the present invention is determined as shown in the above equation, and thus the number of condition probability vector values, which are final output values, is 2n.

When n is 5 using Equations 1 to 2, that is, 2 ^{2 * 5} -QAM = 1024-QAM according to Equation 1, and the number of bits set at each distribution point is 2x5 = 10 bits according to Equation 2 Will be described.

First, the first conditional probability vector is calculated

│ α │ If 2 ⁵ -1, the output is determined by -sign ( α ).

Or | α | ≦ 1, the output is determined to be -0.9375 α .

Or 1 <│ α | ≤2 ⁵ -1, the output is -sign ( α ) { (│ α │-1) + 0.9375}.

The second conditional probability vector equation is then written as the substituted form of the first equation.

Ⓐ │ β │ If 2 ⁵ -1, the output is determined by -sign ( β ).

Ⓑ | β | ≤ 1, the output is determined as -0.9375 β .

Ⓒ 1 <│ β │≤2 ^{5 If} -1, the output is -sign ( β ) { (│ β │ -1) + 0.9375}.

The third conditional probability vector is then

(1) when αβ ≥ 0,

If 2 ⁵ -2 ^{5 (2-m)} ≤│ α │≤2 ⁵ -2 ^{5 (2-m)} +1, the output is determined as (-1) ^m .

In this case, m = 1,2, so substituting this

If 0≤│ α │≤1, the output is determined as -1,

Or 2. If the ⁵ -1≤│ α │≤2 ^5, the output is determined to one.

Or ⓑ 2 ⁴ -1 ≤ α | <2 ⁴ +1, the output is determined to be 0.9375 (│ α │-2 ⁴ ).

Or ⓒ 2 ⁴ -2 ^{4 (2-m)} + m≤│ α │ <2 ⁵ -2 ^{4 (2-m)} + m-2,

The output is (│ α │-2m + 1) +0.9735 (-1) ^m -0.0625,

If you substitute m = 1,2 here,

1≤│ α │ <2 ⁴ -1 (│ α │-1) -1,

Or 2 ⁴ + 1≤│ α | <2 ⁵ -1, (│ α │-3) + 0.825.

(2) when αβ <0,

In this case, α is substituted by β in the equations ⓐ, ⓑ, and ⓒ in the method of determining the output of the third conditional probability vector described above ( αβ ≥ 0).

The fourth (ie k = 4, m = 0,1,2, l = 1,2, p = 1,2,3,4) conditional vector calculation

(1) when αβ ≥ 0,

If m * 2 ⁴ -1 &lt; α | &lt; m * 2 ⁴ +1, the output is determined as (-1) ^{m + 1} .

In this case, substituting m = 0,1,2,

If -1 &lt; α &lt; 1, it is determined as -1.

Or ^24-1 is <│ α │≤2 ⁴ +1, is determined to one.

Or 2 ⁵ -1 <│ α │≤2 ⁵ +1, is determined to be -1.

Or ⓑ (2 L-1) * 2 ³ -1 <│ α | ≤ (2 L-1) * 2 ³ +1,

The output is substituting l = 1,2 at (-1) ^{l + 1} {0.9375│ α | -0.9375 (2ℓ-1) * 2 ³ }

If 2 ³ -1 &lt; α &lt; 2 ³ + 1, the output is determined to be 0.9375 (│ α | -2 ³ ),

Or 3 * 2 ³ -1 &lt; α | &lt; ³ * 2 ³ + 1, the output is determined to be -0.9375 (│ α --3 * 2 ³ ).

Or ⓒ (P-1) * 2 ³ +1 <│ α | ≤ P * 2 ³ -1, and P is odd,

The output is [(-1) ^{((p + 1) / 2) +1} * │ α │ + (-1) ^{(p + 1) / 2} {(P-1) * 2 ³ +1] + (-1) ^{( p + 1) / 2} . But if the value of P is even,

The output is [(-1) ^{p / 2 + l} * | α | + (-1) ^{p / 2} (P * 2 ³ -1)] + (-1) ^{P / 2 + 1} .

Where p = 1,2,3,4,

If 1 <│ α | ≤ 2 ³ -1, the output is {│ α │-1} -1,

Or 2 ³ +1 <│ α | ≤2 ⁴ -1, the output is {│ α │-2 ⁴ +1} +1,

Or 2 ⁴ +1 <│ α | ³ * 2 ³ -1, the output is {2 ⁴ + 1-│ α │} +1,

If 3 * 2 ³ +1 <│ α | ≤2 ⁵ -1, the output is {2 ⁵ + 1-│ α |} -1.

(2) When αβ <0,

In this case, α is substituted by β in the equations ⓐ, ⓑ, ⓒ in the method of determining the output of the fourth conditional probability vector described above ( αβ ≥ 0).

Then the fifth (ie k = 5, m = 0,1,2,3,4, ℓ = 1,2,3,4) equation

(1) When αβ ≥ 0,

If m * 2 ³ -1 &lt; α | &lt; m * 2 ³ +1, the output is determined as (-1) ^{m + 1} .

In this case, substituting m = 0,1,2,3,4,

If -1 &lt; α &lt; 1, the output is determined to be -1.

Or 2 ³ -1 &lt; alpha &lt; 2 ³ + 1, the output is determined to be one.

Or 2 ⁴ -1 &lt; alpha &lt; = 2 ⁴ +1, the output is determined to be -1.

Or 3 * 2 ³ -1 &lt; alpha ³ &amp; le; ³ &amp; le;

Or 2 ⁵ -1 &lt; α &lt; 2 ⁵ +1, the output is determined to be -1.

Or ⓑ (2ℓ-1) * 2 ² -1 <│ α | ≤ (2ℓ-1) * 2 ² +1

The output is substituting l = 1,2,3,4 at (-1) ^{l + 1} 0.9375 {│ α | -0.9375 (2ℓ-1) * 2 ² }

If the ^{2 2 -1 <│ α │≤2 2} +1, the output is determined as ^{0.9375 {│ α │-2 2} },

Or 3 * 2 ² -1 &lt; α 3 &lt; 3 * 2 ² +1, the output is determined to be -0.9375 { -α --3 * 2 ² }.

Or 5 * ² 2 -1 <│ is α │≤5 * ² 2 +1, the output is determined as 0.9375 {│ α │-5 * 2 2}.

Or the 7 * ² 2 -1 <│ is α │≤7 * ² 2 +1, the output is determined as -0.9375 {│ α │-7 * 2 2}.

Or ⓒ (P-1) * 2 ² +1 〈│ α │ ≤P * 2 ² −1 and substituting p = 1,2,3 ................. 7,8 according to when P is odd and even ,

If 1 <│ α │≤2 ² -1, the output is {│ α │-1} -1,

Or 2 ² +1 <│ α | ≤2 ³ -1, the output is {│α│-2 ³ is determined as +1} +1,

Or 2 ³ +1 <│ α | ≤3 * 2 ² -1, the output is {2 ³ + 1-│ α │} +1,

Or 3 * 2 ² +1 <│ α | ≤2 ⁴ -1, the output is {2 ⁴ -1-│ α │} -1.

Or 2 ⁴ +1 <│ α | ≤5 * 2 ² -1, the output is {│ α │-2 ⁴ -1} -1,

Or 5 * 2 ² +1 <│ α | ≤6 * 2 ² -1, the output is {│ α │-6 * 2 ² + 1} + 1 is determined.

Or 6 * 2 ² +1 <│ α | ≤7 * 2 ² -1, the output is Is determined by [6 * 2 ² + 1-│ α |} +1.

Or 7 * 2 ² +1 <│ α | ≤2 ⁵ -1, the output is {2 ⁵ -1-│ α │} -1.

(2) When αβ <0,

In this case, it is determined by substituting α by β in the equations ⓐ, ⓑ, ⓒ in the method of determining the output of the fifth conditional probability vector described above ( αβ ≥ 0).

Subsequently, the sixth conditional probability vector (that is, k = 6, m = 0,1,2 ... 7,8, l = 1,2,3, ... 7,8),

(1) When αβ ≥ 0

If m * 2 ² -1 ≤ α α ≤ m * 2 ² +1, the output is determined as (-1) ^{m + 1} .

At this time, m = 0, 1, 2 ... 7, 8 is applied.

In other words, when α -1≤│ │≤1, the output is determined as -1.

Or 2 ² -1 ≦ αα ≦ ^{2 2} +1, the output is determined to be 1.

If or ^{3 * 2 2 -1≤│ α │≤3 *} 2 2 +1, the output is determined as -1.

Or if ^{7 * 2 2 -1≤│ α │≤7 *} 2 2 +1, the output is determined to one.

Or 2. If the ⁵ -1≤│ α │≤2 ⁵ +1, the output is determined as -1.

Or ⓑ (2ℓ-1) * 2-1 <│ α │ <(2ℓ-1) * 2 + 1,

The output is (-1) ^{ℓ + 1} + {0.9375 | α │-0.9375 (2 ℓ-1) * 2}, and substitute ℓ = 1, 2, 3 ... 7, 8,

If 1 &lt; α α &lt; 3, the output is determined to be 0.9375 (│ α | 2),

Or 5 <If │ α │ <7, the output is determined as -0.9375 (│ α │-6) .

Or 9 <│ α │ <11, the output is determined to be 0.9375 (│ α │-10),

Or 25 &lt; α α &lt; 27, the output is determined to be 0.9375 (│ α -26).

Or 29 is <│ α │ <31, the output is determined as -0.9375 (│ α │-30) .

(2) When αβ <0,

In this case, in the method of determining the output of the sixth conditional probability vector described above ( αβ ≥ 0), α is substituted by β in the equation of ⓐ, ⓑ.

Then, from the 7th 10th condition calculation of the probability vectors for α in a calculation of the sixth conditional probability vector from the third to the β, is use of the equation for replacing the β to α.

11 is a diagram illustrating a function probability vector determination process according to the present invention.

12 is a diagram illustrating a hardware configuration for determining the condition probability vector of the first type 64-QAM according to the present invention. Those skilled in the art may freely modify the hardware configuration within the scope of the present invention.

The invention has been described above in connection with preferred embodiments. However, this is for the purpose of illustration only and not for the purpose of limiting the invention, in practice those skilled in the art will readily be able to understand the modifications within the scope of the invention.

The configuration of the present invention has been described above. The formula for the approximation graph of the present invention presented in the construction process seems to be a fairly complicated formula. However, this complexity is not due to the approximation, but only because it is generalized to be applicable to QAMs of all sizes. In other words, once the size of the QAM is determined, the approximation used is linear linear. This fact has been sufficiently explained in Embodiments 1 and 2 of the present invention, and it seems complicated in the case of 1024QAM where the size of QAM is relatively large, and the actual hardware configuration can be composed of only a few comparators. FIG. 12 illustrates the hardware configuration for determining the condition probability vector of Type 1 64-QAM according to the present invention. The present invention supports this fact. The present invention provides a soft crystal demodulation of a square QAM signal mainly used in industry. By applying the linearized condition probability vector equation instead of the log likelihood ratio method, a significant improvement in processing speed and a reduction in manufacturing cost can be expected.

Claims (20)

In the soft decision method for demodulating a received signal of a quadrature quadrature amplitude modulation (QAM) consisting of an in-phase signal component and a quadrature signal component, a condition determination operation is performed from the quadrature component value and in-phase component value of the received signal. A method for demodulating a soft decision of an orthogonal amplitude modulation using a function included to obtain a condition probability vector value corresponding to each soft decision value corresponding to a bit position of a hard decision The method of claim 1, wherein the condition probability vector for half of all the bits is the same as the method for determining the condition probability vector for the other half of the bits, and is determined by substituting the quadrature component values and the in-phase component values. Flexible demodulation method of quadrature amplitude modulation The method according to any one of claims 1 to 3, wherein the first conditional probability vector of the first type is determined by the following Equation 12 by selecting one of the received values α and β according to the form of the combined distribution diagram. A flexible determination method for demodulating a quadrature amplitude modulation (QAM) received signal, characterized in that [Equation 12] ① │ Ω │≥2 ⁿ -1, the output is determined by a * sign ( Ω ). Or ② | Ω | ≤ 1, the output is determined as a * 0.9375 * Ω . Or ③ 1 is <│ Ω │≤2 ⁿ -1, The output is a * sign ( Ω ) { (│ Ω │-1) + 0.9375}. [Where Ω is the selected reception value, either α or β , 'sign ( Ω )' represents the sign of the selected reception value, 'a' is any real number set according to the desired output range, and α is I ( Receive value of real part) channel, β is received value of Q (imaginary part) channel] The method according to any one of claims 1 to 4, wherein the second conditional probability vector of the first type is determined by the received value selected in the first conditional probability vector and the following Equation (13). Flexible Decision Method for Demodulation of Amplitude Modulation (QAM) Received Signal [Equation 13] ① 2 ⁿ -2 ^{n (2-m)} ≤│ Ω │≤2 ⁿ -2 ^{n (2-m)} +1, the output is determined as a * (-1) ^{m + 1} . ② If 2 ^n-1 −1≤│ Ω │ <2 ^n-1 +1, The output is determined by a * 0.9375 (2 ^n-1- | Ω |). ③ If 2 ^{n-1-2 (n-1) (2-m)} + m≤│ Ω │ <2 ^n- 2 ^{(n-1) (2-m)} + m-2, -a * { Determined as ( ? Ω -2m + 1) +0.9735 (-1) ^{m + 1 +} 0.0625}. Where Ω is the selected reception value, n is the variable that determines the magnitude of the QAM, ie 2 ²ⁿ , and 'a' is any real number set according to the desired output range, m = 1, 2] The conditional probability vector of the third to n-1th type of the first type is determined by the received value selected in the first conditional probability vector and the following Equation (14). A flexible determination method for demodulating a quadrature amplitude modulation (QAM) received signal, characterized in that [Equation 14] ① If ^{m * 2 n-k + 2} -1≤│ Ω │≤m * 2 n-k + 2 +1, the output is a * (- 1) determined by the ^{m + 1.} If or ② (2ℓ-1) * 2 n-k + 1 -1 <│ Ω │≤ (2ℓ-1) * 2 n-k + 1 +1, The output is determined by a * (-1) ^{l + 1} 0.9375 {│ Ω |-(2ℓ-1) * 2 ^{n-k + 1} }. Or ③ (P-1) * 2 ^{n-k + 1} +1 &lt; │ &amp; le; P * 2 ^{n-k + 1} -1, The output is according to the value of P, if P is odd, a * { [(-1) ^{((p + 1) / 2) +1} * │ Ω │ + (-1) ^{(p + 1) / 2} {(P-1) * 2 ^{n-k + 1} +1}] + (-1) determined by ^{(p + 1) / 2} }. If P is even, a * { [(-1) ^{p / 2) +1} * Ω Ω + (-1) ^{p / 2} (P * 2 ^{n-k + 1} -1)] + (-1) ^{P / 2 + 1} }. [Wherein, Ω is the selected reception value, n is the variable that determines the magnitude of QAM, that is, 2 ²ⁿ , and m = 0,1, ..., 2 ^k-2 , ℓ = 1,2, ..., 3 ^k-2 , and p = 1,2, ..., 2 ^k-1 and k is the number of conditional probability vector (k = 3,4, ..., n-1)] The orthogonality of claim 1 or 2, wherein the nth conditional probability vector of the first type is determined by the reception value selected when the first conditional probability vector is determined and the following Equation (15). Flexible Decision Method for Demodulation of Amplitude Modulation (QAM) Received Signal [Equation 15] ① If ^{m * 2 2 -1≤│ Ω │m *} 2 2 +1, the output is a * (- 1) determined by the ^{m + 1.} Or (2 L-1) * 2 ¹ -1 &lt; Ω &lt; (2 L-1) * 2 ¹ +1, The output is determined by a * (-1) ^{l + 1} 0.9375 {│ Ω |-(2ℓ-1) * 2 ¹ }. Where Ω is the selected reception value, n is the variable that determines the magnitude of the QAM, ie 2 ²ⁿ , and m = 0,1..2 ^n-2 , and ℓ = 1, 2..3 ^n-2 ] The conditional probability vector of any one of claims 1 and 2, wherein the conditional probability vector of the n + 1st to 2nth type of the first type is a received value that is not selected when the first conditional probability vector is determined and the equations (12) to (15). Determination method for demodulating a quadrature amplitude modulation (QAM) received signal, characterized in that by sequentially using each (where the number k of the condition probability vector included in Equation 14 is sequentially from n to 3 to n-1) Instead of +3 to 2n-1) The method according to any one of claims 1 to 3, wherein the first conditional probability vector of type 2 is determined by the following Equation 16 by selecting one of the received values α and β according to the combination distribution type. A flexible determination method for demodulating a quadrature amplitude modulation (QAM) received signal, characterized in that [Equation 16] Ω Ω ≥2 ⁿ -1, the output is determined by -a * sign ( Ω ). Or ⓑ Ω Ω ≤ 1, the output is determined as a * 0.9375 * Ω . Or ⓒ 1 <│ Ω │≤2 ⁿ -1, the output is -a * sign ( Ω ) { (│ Ω │-1) + 0.9375}. [Where Ω is the selected reception value, 'sign ( Ω )' is the sign of the selected reception value, n is the variable that determines the size of QAM, ie 2 ²ⁿ , and 'a' is an arbitrary value set according to the desired output range. Real number, α is the received value of the I (real) channel, β is the received value of the (imaginary) channel] The second conditional probability vector of the second type is calculated by substituting a selected reception value with an unselected reception value in a method of obtaining the first conditional probability vector of the second type. A flexible determination method for demodulating an orthogonal amplitude modulation (QAM) received signal, characterized in that Claim 1 or 2 wherein selection according to any one of the preceding, in accordance with the three types of first conditional probability vector is received α value and β combined distribution to any one of the second type of and to, when the α * β ≥0 When [Equation 17] is used, and when [alpha] * β <0, Quadrature Amplitude Modulation (QAM) reception is characterized by replacing the selected reception value in [Equation 17] with an unselected reception value. Flexible Decision Method for Demodulation of Signal [Equation 17] If 2 ⁿ -2 ^{n (2-m)} ≤│ Ω │2 ⁿ -2 ^{n (2-m)} +1, the output is determined ^as a * (-1) ^m . Or ⓑ it is ^{2 n-1 -1≤│ Ω │ <} 2 n-1 +1, the output is determined as a * 0.9375 (│ Ω │- 2 n-1). Or ⓒ 2 ^n-1 -2 ^{(n-1) (2-m)} + m≤│ Ω │ <2 ⁿ -2 ^{(n-1) (2-m)} + m-2, The output is a * { Determined as (│ Ω │-2m + 1) +0.9735 (-1) ^m -0.0625}. Where Ω is the selected reception value, n is the variable that determines the magnitude of the QAM, ie 2 ²ⁿ 'a' is any real number set according to the desired output range and α is I (real part) Reception value of channel, β is reception value of Q (imaginary part) channel, m = 1,2] The method according to claim 1 or 2, wherein the fourth conditional probability vector is α * β ≥ in a method of obtaining a third conditional probability vector of type 2 when the size of QAM of type 2 is less than or equal to 64-QAM. When 0 and α * β <0, a soft decision method for demodulating a quadrature amplitude modulation (QAM) received signal, wherein the received values are calculated by substituting unused received values. The fifth conditional probability vector of claim 1 or 2, wherein when the size of the QAM of type 2 is 64-QAM, the fifth conditional probability vector selects one of the reception values α and β according to the shape of the combined distribution diagram. In the case of α * β ≥ 0, Equation 18 is used, and in the case of α * β <0, the selected reception value in Equation 18 is replaced with an unselected reception value. Flexible Decision Method for Demodulation of Quadrature Quadrature Amplitude Modulation (QAM) Received Signals Equation 18 If ^{ⓐ m * 2 n-1 -1} <│ Ω │≤m * 2 n-1 +1, the output is a * (- 1) determined by the ^{m + 1.} Ⓑ is ^{(2ℓ-1) * 2 n} -1 -1 <│ Ω │≤ (2ℓ-1) * 2 n-1 +1, The output is determined by a * (-1) ^{ℓ + 1} {0.9375│ Ω │-0.9375 (2ℓ-1) * 2 ^n-1 } Where Ω is the selected reception value, n is the variable that determines the magnitude of the QAM, ie 2 ²ⁿ 'a' is any real number set according to the desired output range and α is I (real part) The reception value of the channel, β is the reception value of the Q (imaginary part) channel, m = 0,1,2 and ℓ = 1,2] The sixth conditional probability vector according to any one of claims 1 and 2, wherein when the size of QAM of type 2 is 64-QAM, α * β ≥ in a method of obtaining the fifth conditional probability vector of type 2 When 0 and α * β <0, a soft decision method for demodulating a quadrature amplitude modulation (QAM) received signal, wherein the received values are calculated by substituting unused received values. The conditional probability vector of the fourth to nth terms of any of the received values α or β is determined by the combination distribution diagram when the QAM of type 2 is 256-QAM or more. If it is selected according to the form and α * β ≥ 0, it is determined by Equation 19 below, and if α * β 〈0, the received value not selected in Equation 19 is determined using the selected formula value. A flexible determination method for demodulating a quadrature amplitude modulation (QAM) received signal for demodulating a quadrature amplitude modulated (QAM) received signal [Equation 19] If m * 2 ^{n-k + 3} −1 〈│ Ω │ ≦ m * 2 ^{n-k + 3 +} 1, the output is determined as a * (-1) ^{m + 1} Or ⓑ (2 L-1) * 2 ^{n-k + 2} -1 &lt; Ω ≤ (2 L-1) 2 ^{n-k + 2} +1, The output is ^{a * (- 1) ℓ +} 1 determined by {0.9375│ Ω │-0.9375 (2ℓ -1) * 2 n-k + 2}. Or ⓒ (P-1) * 2 ^{n-k + 2} +1 <│ Ω │≤P * 2 ^{n-k + 2} -1, The output is according to the value of P, if P is odd, a * { [(-1) ^{((p + 1) / 2) +1} * │ Ω │ + (-1) ^{(p + 1) / 2} {(P-1) * 2 ^{n-k + 2} +1}] + (-1) determined by ^{(p + 1) / 2} }, If P is even, a * { Determined by [(-1) ^{p / 2 + 1} * Ω Ω + (-1) ^{p / 2} (P * 2 ^{n-k + 2} +1)] + (-1) ^{p / 2 + 1} }. Where k is the conditional probability vector number (4,5…, n), Ω is the selected reception value, n is the variable that determines the magnitude of the QAM, ie 2 ²ⁿ , and 'a' is an arbitrary value set according to the desired output range. Where α is the received value of the I (real) channel, β is the received value of the Q (imaginary) channel, and m = 0,1,. , 2 ^k-3 , L = 1,2... , 2 ^k-3 , p = 1,2... , 2 ^k-2 ] 3. The method according to any one of claims 1 and 2, wherein the n + 1th conditional probability vector is determined from the fourth to nth conditional probability vectors of type 2 when the size of QAM of type 2 is 256-QAM or more. When the received value is α * β ≥ 0, it is determined by Equation 20 below, and when α * β <0, the received value not selected in Equation 20 is replaced by the selected equation. A flexible determination method for demodulating a quadrature amplitude modulation (QAM) received signal, characterized in that [Equation 20] If m * 2 ² -1 ≤│Ω│≤ m * 2 ² +1, the output is determined as a * (-1) ^{m + 1} . Or ⓑ (2ℓ-1) * 2 ¹ -1 <│ Ω │ <(2ℓ-1) * 2 ¹ +1, The output is determined by a * (-1) ^{l + 1} {0.9375 | Ω | -0.9375 (2ℓ-1) * 2 ¹ }. [Where Ω is the selected reception value, n is the variable that determines the size of QAM, ie 2 ²ⁿ , 'a' is any real number that is set according to the desired output range, and α is the reception value of I (real) channel, β is the reception value of the Q (imaginary part) channel, and m = 0, 1,... , 2 ^n-2 , L = 1,2... , 2 ^n-2 ] The conditional probability vector of any one of claims 1 and 2, wherein the size of the QAM of type 2 is greater than or equal to 256-QAM. A soft decision method for demodulating an orthogonal amplitude modulation (QAM) received signal, which is identical to the method for obtaining conditional probability vectors The conditional probability vector of n + 3 th to 2n-1 th is the magnitude of QAM of type 2 when the QAM of type 2 is 256-QAM or more? In case of 256-QAM or more, when α * β ≥ 0 and α * β 〈0 in the fourth to nth conditional probability vectors, the received values are calculated by substituting the received values. Flexible Decision Method for Demodulation of Quadrature Quadrature Amplitude Modulation (QAM) The conditional probability vector of claim 2, wherein the second n-th conditional probability vector is greater than or equal to 256-QAM when the size of QAM of type 2 is greater than or equal to 256-QAM. Demodulating the Quadrature Amplitude Modulation (QAM) Received Signal when the conditions of the probability probability vector are determined by substituting the received values for the values of α * β ≥ 0 and α * β <0, respectively. Ductile Determination Method A device for demodulating a quadrature amplitude modulation (QAM) received signal comprising an in-phase signal component and a quadrature signal component, using a function including condition determination operation from quadrature and in-phase component values of the received signal. And a condition probability vector decision unit for obtaining condition probability vector values which are respective soft decision values corresponding to the bit positions of the hard decision. 20. The method of claim 19, wherein the condition probability vector determiner is the same as the method of determining the condition probability vector for the other half of the bits, replacing the quadrature component values and the in-phase component values, respectively. Flexible crystal demodulation device for orthogonal amplitude modulation, characterized in that for determining