JPH06152661A - Mapping method, decoding method and modem - Google Patents

Abstract

PURPOSE:To obtain a mapping method and a decoding method in which quantity of stored data is reduced and to obtain an inexpensive MODEM with less memory capacity. CONSTITUTION:A received 7-bit signal is divided into high-order 4 bits and low-order 3-bits by a divider 100, and the high-order 4 bits are used to generate an address accessing a basic coordinate table by a memory address generator 101 and a basic coordinate point is read from a memory 102. The low-order 3-bits are used for discriminating to which sub-set the signal belongs. In response to the discriminated sub-set, an adder 104 and a rotation device 105 apply mapping such as rotation to the basic coordinate point to obtain the reception signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention
More specifically, the present invention relates to a demodulation device (hereinafter abbreviated as a modem), and more particularly to a modem mapping method using a trellis code.

[0002]

2. Description of the Related Art With the recent development of information communication technology, higher speed data transmission is being realized. A typical example of this high-speed data transmission is a facsimile apparatus using a telephone line. A modem (so-called modem) is used for this data transmission, and the transmission method of the modem is V. 27 ter, V.I. 29, V.I. 17 is recommended.

These recommendations define the modulation method, transmission speed, etc., and the outline thereof is shown below.

V. 27ter: Phase modulation (PSK), 4
800 bps V. 29: Quadrature Amplitude Modulation (QAM), 9600b
ps V. 17: Quadrature Amplitude Modulation (QAM), 14400
bps bp is an abbreviation for bit / sec and represents the number of transmission bits per second.

As mentioned above, Recommendation V. In 17, 14.
Ultra-high-speed data transmission of 4 Kbps is possible.

Here, V. 17, V.I. The quadrature amplitude modulation method adopted in 29 etc. will be briefly described. Generally QA
The M signal is given by the following equation.

[0007] _{s (t) = a 1 (} t) sin (w c t) + a 2 (t) cos (w c t) ... (1) s (t): in modulation signal at a time t w _c: Carrier angular frequency a ₁ (t): Amplitude of in-phase component at time t a ₂ (t): Amplitude of quadrature component at time t As can be seen from equation (1), QAM and 90 This is a modulation method in which the amplitudes of different carrier waves (sin () and cos () of the formula (1), which are called in-phase wave and quadrature wave, respectively) are obtained and the addition signal is obtained. It is a combination of phase modulation.

In FIG. 17 is a graph in which the signal constellation of 17, that is, the amplitude a ₁ (t) of the in-phase component is plotted on the horizontal axis and the amplitude a ₂ (t) of the quadrature component is plotted on the vertical axis.

As can be seen from this figure, the phase component at each point corresponds to the phase-modulated phase, and the amplitude components on the vertical and horizontal axes correspond to the amplitude due to amplitude modulation.

The modem determines the point on the signal constellation according to the digital signal value (7-bit value in the case of V.17) to be transmitted (determination of a ₁ (t) and a ₂ (t) above).
However, a fixed period (generally called a symbol interval (reciprocal of baud rate). 1/2400 sec in V.17)
Equation (1) is calculated for each and sent to the line.

V. According to the 17th Recommendation, a trellis code is adopted as a convolutional code (modulation code) in order to realize reliable ultrahigh-speed transmission. This is to add and transmit a redundant signal according to a rule on the transmitting side.

Actually, in the 6-bit signal to be transmitted,
Add one redundant bit according to the trellis code rule. The rules of the trellis code are represented by a generally called trellis state transition diagram. Although we will not explain this in detail, the point of this rule is that the lower 2 bits and the redundant bit (7
8 subsets (a of the alphabet) formed by a total of 3 bits of the least significant bit when extended to bits
~ H).

FIG. 6 shows an expanded 7-bit transmission signal and a signal constellation indicating a signal point constellation, and FIG. 7 shows a transmission signal divided into a to h and a signal constellation indicating a signal point constellation. The relationship between the subsets a to h and the values of the lower 3 bits in the 7-bit signal is as follows.

Subset a: 000 Subset b: 010 Subset c: 100 Subset d: 110 Subset e: 011 Subset f: 101 Subset g: 111 Subset h: 001 Further, the subscript of the subset alphabet in FIG. 7 is the upper 4 bits of the 7-bit signal. The hexadecimal value of the bit.

In the case of constructing such a modem, a total of 7-bit data obtained by adding the redundant bit (1 bit) described above to the 6-bit transmission data input from the transmission terminal is used according to the correspondence of FIG. It is necessary to determine the point. This operation is generally called mapping.

For example, if the 7 bits input to the mapper is 0110001, the output of the mapper is the coordinate value (1.0, 2.0) (the subset is h and the upper 4 bits are hexadecimal values). Becomes 6, that is, h ₆ ).

[0017]

Conventionally, the mainstream of this mapping means is to have all points as a table in a memory element.

For example, in a memory such as a ROM, all the points are arranged so as to be double the 7 bits at which each address is transmitted, and the input 7-bit value is shifted to the left by 1 bit (value is 2 times), and two transmission coordinate values (x component, y component) are called from that address to map the transmission signal. In order to do such processing,
It is necessary to store all the coordinates of 128 signal points. This is x component for one point, y for the number of words in the memory.
Two words are needed to store the component, resulting in 256
A large-capacity memory element called word was required.

In a normal modem, the above digital signal processing is performed by software using a processor such as DSP (Digital Signal Processor).

Therefore, as the large-capacity storage element as described above, a large-capacity ROM is required outside the processor or a processor having a large-capacity ROM built therein. As described above, in the conventional means, a more expensive processor and an increase in the number of parts are required, which is also a cause of increasing the cost of each modem.

The present invention has been made in view of the above conventional example, and an object of the present invention is to provide a mapping method for performing mapping with a small memory capacity and a more inexpensive modulator / demodulator using the mapping method.

[0022]

[Means for Solving the Problems] and

In order to achieve the above object, the mapping method of the present invention has the following configuration.

A mapping method using a trellis code, wherein a selection step of selecting points in basic coordinates having a predetermined subset number of points out of all points at a desired communication speed; A step of expanding to a point on a predetermined signal constellation according to each subset.

Also, a mapping method using a trellis code, in which y in the basic coordinates having the number of points of a predetermined subset of all points at a desired communication speed is used.
A selection step of selecting a point in a region where the component is either positive or negative and a development step of expanding the selected point to a point on a predetermined signal constellation according to each subset.

The decoding method of the present invention has the following configuration.

A sub-set decoding method using a trellis code, in which a receiving point corresponding to each sub-set is converted into a basic coordinate having the number of points of one subset of all points at a desired communication speed; Determination step of determining the sign of the y component of the received reception point, a mapping step of mapping the reception point converted by the determination value of the determination step to a region where the y component is positive or negative, and the mapping step. For the received point, an evaluation step of evaluating the y component of the basic coordinates in either the positive or negative region, and the judgment value evaluated by the evaluation step, the y component is determined to be negative or negative according to the sign judgment result of the judgment step. A reverse mapping step for reverse mapping to the positive region.

A sub-set decoding method using a trellis code, in which a receiving point corresponding to each sub-set is converted into basic coordinates having the number of points of one subset of all points at a desired communication speed. , A determination step of determining the signs of the x component and the y component of the converted reception point, and the conversion reception point is subjected to area determination in one quadrant of the basic coordinates based on the determination value determined in the determination step. It is characterized by

The modulation / demodulation device of the present invention has the following configuration.

A modulation / demodulation apparatus using a trellis code, wherein selecting means for selecting a point within the basic coordinates having a predetermined subset number of points out of all points at a desired communication speed, and the selected point are And expanding means for expanding to a point on a predetermined signal constellation according to each subset.

Further, the modulation / demodulation apparatus using the trellis code, the conversion means for converting the reception points corresponding to each subset into basic coordinates having the number of points of one subset of all points at a desired communication speed, Determining means for determining the sign of the y component of the converted receiving point, mapping means for mapping the receiving point converted by the determination value in the determining step to a region where the y component is positive or negative, and the mapping step. With respect to the mapped reception point, the y-component of the basic coordinates is evaluated in either positive or negative area, and the judgment value evaluated in the evaluation step is set to a negative y-component according to the sign judgment result in the judgment step. Or a reverse mapping means for reverse mapping to a positive region.

Further, the modulation / demodulation apparatus using the trellis code, selecting means for selecting a point within the basic coordinates having the number of points of a predetermined subset of all points at a desired communication speed, and the selected point Is expanded to a point on a predetermined signal constellation according to each subset.

Further, the modulation / demodulation apparatus using the trellis code, the conversion means for converting the reception points corresponding to each subset into basic coordinates having the number of points of one subset of all points at a desired communication speed, A determination means for determining the signs of the x component and the y component of the converted reception point, and determining the area of the converted reception point in one quadrant of the basic coordinates based on the determination value determined in the determination step. Is characterized by.

[0033]

【Example】

[Embodiment 1] As described in the above-mentioned prior art, from FIG. 7, the subsets e, c, and g are equal to those obtained by rotating the subset a by 90 degrees, 180 degrees, and 270 degrees, respectively, and the subsets d, h, and b. Subset f 90 degrees,
It turns out that it is equivalent to the one rotated 180 degrees and 270 degrees.
This is V. This is a point arrangement determined in order to guarantee the rotation invariance of 90 degrees, 180 degrees, and 270 degrees of 17 modems, and this rule is used in this embodiment.

As can be seen from this rule, the subsets e, c and g and the subsets d, h and b can be easily obtained by rotating the subsets a and f. Therefore, when mapping, only the subsets a and f need to be focused. Here, the basic point arrangement used for mapping is further considered from the relationship of the point arrangement of the subsets a and f. This basic point arrangement is shown in FIG. Each point of the basic point arrangement is named z,
The subscript is the hexadecimal value of the upper 4 bits. The coordinates of the subset a (x _a , y _a ), the coordinates of the subset f (x
The relationship between _f , y _f ) and the coordinates (x _z , y _z ) of the basic point arrangement is shown below.

[0035] _{(x a, y a) =} (x z, y z -1) ... (2) (x f, y f) = (x z -1, y z) ... (3) mapping according to this embodiment Are all performed on this basic point arrangement.

Next, this embodiment will be described in detail with reference to FIGS.

FIG. 1 is a block diagram of a mapping apparatus according to the present embodiment, FIG. 2 is a point table on the arrangement of basic points stored in a memory, and FIG. 3 is a mapping procedure by the apparatus of FIG. It is a flowchart.

First, in a convolutional decoder (not shown), 7 bits (code value) to which redundant bits are added are divided by the upper 4 bits / lower 3 bits divider 100 (step S301). Specifically, the logical sum of the input 7 bits and hexadecimal 3 is output as the lower 3 bits, and the logical sum result of the input 7 bits and hexadecimal 78 is shifted to the right by 3 bits. The result is output as the upper 4 bits.

Next, the upper 4 bits are input to the address generator (101) of the storage element (memory such as ROM) 102 in which each point of the basic point arrangement shown in FIG. 4 is stored, and the address of the memory is determined. (Step S30
2). FIG. 2 shows a memory table in which the basic points are stored. In the table, each point is stored in the order of Z _{0 to} Z _f by using two words of x component and y component, respectively.
The value written on the left side of the figure is the memory address. Therefore, if such a memory arrangement is used, a value obtained by doubling the upper 4-bit value (which takes a value from 0 to f) input to the memory address generator is used as the memory address, and the data from the storage element 102 is stored. If read, the point on the basic point arrangement is determined from the upper 4 bits (step S3).
03). That is, the memory address generator 101 outputs a double value of the input value (more specifically, a 1-bit left shift value of the input value). By the steps up to this point, the corresponding point in the basic point arrangement of code values is determined.

Next, the lower 3 bits divided by the upper 4 bits / lower 3 bits divider are input to the subset determiner 103. The subset determiner 103 receives the input 3
It is determined which subset of a to h the bit value corresponds to, and as a result, a conversion multiplier (x
The component, the y component added value, and the rotation angle) are output. Specifically, the subset is defined as a subset a,
It is divided into two groups e, c, g (group of subset a) and subsets f, d, h, b (group of subset a) (step S304). Next, a conversion constant (x component, y component added value) from the coordinates of the basic point arrangement to the coordinates of the subset a or the subset f is determined according to each group (steps S305 and S306). In the case of the subset a group, the x component and y component addition values are 0.0 and -1.0, respectively, and in the case of the subset f group, the x component and y component addition values are respectively -1.0 and -1.0.
It is 0.0. That is, in the case of the group of the subset a, the constants for converting the basic points into the points of the subset a are selected,
In the case of the group of the subset f, the basic point is set to the subset f
Select the constant to be converted to the point of. As described above, the conversion of the other subsets into points is carried out for each subset a.
Alternatively, it is rotated by 90 degrees, 180 degrees, or 270 degrees of the point of f. The next step is to determine this rotation angle. Here, when the subset is a or f, 0 degree is selected as the rotation angle (steps S307 and S3).
08), if the subset is e or d, 90 degrees is selected as the rotation angle (steps S309 and S310), and if the subset is c or h, 180 degrees is selected as the rotation angle (steps S311 and S312) and the subset is g. In the case of b, select 270 degrees as the rotation angle (step S
313 / S314), and outputs the result. In this way, the subset determiner 103 outputs the x-component / y-component added value and the rotation angle.

This x-component / y-component added value is input to the adder 104 together with the x-component and y-component of the basic point coordinates read from the memory. The adder 104 adds the x component added value to the x component of the basic point coordinates and the y component added value to the y component of the basic point coordinates, and outputs the result (step S315). By this operation, the coordinate of the point of the subset a or f corresponding to the basic point coordinate value corresponding to the code value is determined.

This adder output and the subset discriminator 10
The reversal angle from 3 is input to the rotator 105. Rotator 1
05 rotates the adder output to the right by 0 °, 90 °, 180 °, and 270 °, respectively, according to the rotation angle of the subset determiner output, and outputs the result (step S306). Rotation of 0 degrees, 90 degrees, 180 degrees and 270 degrees to the right respectively
The signals R, R ', R ", and R"' can be expressed by the following equations, respectively.

R = (x, y): right 0 degree rotation R ′ = (y, −x): right 90 degree rotation R ″ = (− x, −y): right 180 degree rotation R ″ ′ = (− y, x): rotated right 270 degrees The rotation angle of the subset determiner 103 output is the subset a
Or, it is a rotation angle with respect to the coordinates of f, and if the rotation is performed at this rotation angle, the sub-area a to which the lower 3 bit values originally belong is
Mapping points (coordinates) for h are determined.

As described above, according to the present invention, the number of data points used for mapping is 16 from the conventional 128 points.
It is possible to perform mapping (expansion) from 16 points to 128 points by a simple means of reducing to points and moving coordinates and rotating points. Therefore, it is possible to provide a mapping unit that significantly reduces the storage capacity required to store the mapping points as compared with the related art, and that can be expanded to 128 points by a simple means.

If the mapping method described above is used for a modem, the required memory capacity can be small and an inexpensive modem can be realized.

Further, in the present embodiment, only the case where the bit rate is 14.4 Kbps has been shown, but the present invention is not limited to this, and for example, V.50. 1 recommended in 17
It can be easily applied to the cases of 2.0K, 9.6K and 7.2Kbps.

The number of points at 12.0K, 9.6K and 7.2Kbps is 64, 32 and 16, respectively.
240 points in total including 128 points at 4.4 Kbps
The number of points is the same as the number of points, and the number of words in memory is 4.
It will be 80 words.

Applying the method described in the above embodiment,
The number of points at 12.0K, 9.6K, and 7.2Kbps is 8, 4, and 2 points, respectively, in the same way that the number of points at 14.4Kbps is 16, and as a result, the number of words in the memory is Reduce from 480 words to 60 words.

Further, in the present embodiment, the case where it is applied only to the mapping used in the transmission system of the modem has been described, but it can be applied to the reception system. In the reception type, it is necessary to decode the data encoded in the transmission system, and in order to obtain the error signal, the point on the signal constellation may be obtained in reverse from this decoded value. This operation can also be said to be one mapping, and as a means thereof, the present invention can be applied as it is. In this case, the receiving unit can also reduce the data point storage capacity in the same manner. Furthermore, since the stored data points are the same as the transmitting unit, the transmitting and receiving unit can share the data, and the data storage capacity as a modem system can be significantly increased. Can be reduced.

[0050]

[Other Examples]

[Second Embodiment] As a second embodiment, a mapping method having the same configuration as that of the first embodiment will be described. Focusing on the sign of the y component of the basic point arrangement in FIG. 4, the following can be seen.

When the sign of the y component is negative: the least significant bit of the upper bits is 0 (even number) When the sign of the y component is positive: the least significant bit of the upper bits is 1 (odd number) The upper bit value is x It is a value obtained by adding 1 to the corresponding bit when the axisymmetric y component is negative.

Therefore, the point actually used for mapping may be only when the y component of FIG. 4 is positive (upper half of FIG. 4) or when the y component is negative (lower half of FIG. 4). All mapping according to this embodiment is performed on this basic point arrangement (lower half of FIG. 4).

Next, this embodiment will be described with reference to FIGS.

FIG. 8 is a block diagram of a mapping apparatus according to this embodiment, FIG. 9 shows a point table on the basic point arrangement stored in the memory, and FIG. 10 shows a mapping procedure by the apparatus of FIG. It is a flowchart.

First, in a convolutional decoder (not shown), 7 bits (code value) to which redundant bits are added are divided by the upper 4 bits / lower 3 bits divider 100 (step S1001). Specifically, the input 7 bits and 16
The logical OR of the decimal 3 is output as the lower 3 bits, and the logical OR result of the input 7 bits and the hexadecimal 78 is shifted to the right by 3 bits and output as the upper 4 bits.

Next, for the upper 4 bits, the even / odd of the bit value is judged by the even / odd judgment unit 801 (S100).
2). This even-odd decision unit 801 enables the evenization instruction signal when the input is odd, outputs -1 as the multiplication constant, and disables the evenization instruction signal when the input is even, and outputs +1 as the multiplication constant. To do.

The even numbering unit 802 has a function of making the input even.
Of the even-odd decision unit 801
Even if the even instruction signal is enabled (even
Set to 0, S1003 to S1004), even numbered fingers
When the indicating signal is disabled, nothing is output (S1).
005). Use this output directly as a memory address
To be FIG. 9 shows a table of the memory 803. memory
Inside, there are even points of the upper 4 bits of the basic point arrangement
But Z₀ , Z₂ , Z_Four , Z₆ , Z₈ , Z_a , Z _c , Z_e 
Are stored in this order. The value at the left end of the figure is the memory
Indicates an address. Therefore, the output of the evenizer 802 is
If you use the address of the memory 103 directly,
Mapping corresponding to each code value when 4 bits are even
A value is obtained (S1006).

Further, the y component output from the memory 803 is the multiplication constant (+1 or −) from the even / odd decision unit 801.
1) is multiplied. That is, the sign of the y component is inverted only when the upper 4 bits are odd (S1007).

In this way, the coordinates of the basic point are determined by the x component of the memory output and the y component of the multiplier output.

Next, the lower 3 bits divided by the upper 4 bits / lower 3 bits divider are input to the subset determiner 103. The subset determiner 103 receives the input 3
It is determined which subset of a to h the bit value corresponds to, and as a result, a conversion multiplier (x
The component, the y component added value, and the rotation angle) are output. Specifically, the subset is defined as a subset a,
It is divided into two groups e, c, g (group of subset a) and subsets f, d, h, b (group of subset a) (step S1008). Next, a conversion constant (x component, y component added value) from the coordinates of the basic point arrangement to the coordinates of the subset a or the subset f is determined according to each group (steps S1009 and S101).
0). In the case of the group of subset a, the x component and y component addition values are 0.0 and -1.0, respectively, and in the case of the group of the subset f, the x component and y component addition values are respectively-.
It is 1.0 and 0.0. That is, in the case of the group of subset a, the constants for converting the basic points into the points of the subset a are selected, and in the case of the group of the subset f, the constants for converting the basic points into the points of the subset f are selected. As described above, the conversion of the points of the other subsets into the points of the subsets a or f is performed by 90 degrees, 180 degrees, and 27 degrees, respectively.
It will be rotated 0 degrees. The next step is to determine this rotation angle. Here, when the subset is a or f, 0 degree is selected as the rotation angle (step S101).
1 · S1012), if the subset is e or d, 90 ° is selected as the rotation angle (steps S1013 · S10).
14) If the subset is c or h, select 180 degrees as the rotation angle (steps S1015 and S1016),
When the subset is g or b, 270 degrees is selected as the rotation angle (steps S1017 and S1018), and the result is output. Thus, the subset determiner 103
The x-component / y-component added value and the rotation angle are output.

This x-component / y-component added value is input to the adder 104 together with the x-component and y-component of the basic point coordinates read from the memory. The adder 104 adds the x component addition value to the x component of the basic point coordinates and the y component addition value to the y component of the basic point coordinates, and outputs the result (step S1019). By this operation, from the basic point coordinate value corresponding to the code value, the corresponding subset a or f
The coordinates of the point are determined.

This adder output and the subset determiner 10
The reversal angle from 3 is input to the rotator 105. Rotator 1
05 rotates the adder output to the right by 0 °, 90 °, 180 °, and 270 °, respectively, according to the rotation angle of the subset determiner output, and outputs the result (step S1020).
The rotations of 0 °, 90 °, 180 °, and 270 ° to the right can be expressed by the following formulas as signals R, R ′, R ″, and R ″ ′, respectively.

R = (x, y): right 0 degree rotation R ′ = (y, −x): right 90 degree rotation R ″ = (− x, −y): right 180 degree rotation R ″ ′ = (− y, x): rotated right 270 degrees The rotation angle of the subset determiner 103 output is the subset a
Or, it is a rotation angle with respect to the coordinates of f, and if the rotation is performed at this rotation angle, the sub-area a to which the lower 3 bit values originally belong is
Mapping points (coordinates) for h are determined.

As described above, according to the present invention, the number of data points used for mapping is reduced from the conventional 128 points to 8 points, and the coordinate points are moved from 8 points to 128 points by simple means such as point rotation. Mapping (expansion) can be done. Therefore, it is possible to provide a mapping unit that significantly reduces the storage capacity required to store the mapping points as compared with the related art, and that can be expanded to 128 points by a simple means.

Further, in the present embodiment, only the case where the bit rate is 14.4 Kbps is shown, but the present invention is not limited to this, and for example, V.V. 1 recommended in 17
It can be easily applied to the cases of 2.0K, 9.6K and 7.2Kbps.

However, 12.0 Kbps and 9.6 Kbp
In the case of s, in order to make the basic point arrangement symmetrical with respect to the x axis, it is necessary to rotate the finally obtained rotator output by 45 degrees to the left and double the amplitude.

Furthermore, 12.0 Kbps, 9.6 Kbp
The rule of the x-axis symmetry of the basic point arrangement of s, 7.2 Kbps is as follows.

[0068] Therefore, the storage area (sign of the y-axis) of the basic points required for mapping differs depending on the bit rate. Also, 12.0K
In the case of bps, the upper 4-bit value cannot be divided into an even number and an odd number depending on the sign of the y-axis component. Therefore, the even / odd decision unit 801 in FIG. Needs to be changed to an adder. The function of the anti-period 801 in this configuration is to output −1 as a multiplication constant when the input is 4 or more and −4 as an addition constant to the adder 802, and when the input is 3 or less, +1 as a multiplication constant. 0 is output as the addition constant. Furthermore, means for doubling the output of the adder 802 is also required.

12.0 Kbps, 9.6 Kbps, 7.
The number of points for 2 Kbps is 64, 32, 16 respectively.
Thus, the total number of points is 240 points including 128 points of 14.4 KBPS, and the number of words in the memory is 480 words.

When the present invention is used, the number of points at 14.4 Kbps becomes 8 as well as 12.0 Kbps.
The number of points at 9.6 Kbps and 7.2 Kbps is 4, 2 and 1 point, respectively, and as a result, the required number of words in the memory is reduced from 480 words in the past to 30 words.

Further, in the present embodiment, the case where it is applied only to the mapping used in the transmission system of the modem has been described, but it can be applied to the reception system. In the reception type, it is necessary to decode the data encoded in the transmission system, and in order to obtain the error signal, the point on the signal constellation may be obtained in reverse from this decoded value. This operation can also be said to be one mapping, and as a means thereof, the present invention can be applied as it is. In this case, the receiving unit can also reduce the data point storage capacity in the same manner. Furthermore, since the stored data points are the same as the transmitting unit, the transmitting and receiving unit can share the data, and the data storage capacity as a modem system can be significantly increased. Can be reduced. [Third Embodiment] As a third embodiment, a subset decoding method for decoding a received signal in a modem will be described.

Hereinafter, the same V. Take 17 as an example. FIG. It is a figure showing the signal constellation of 17. Also,
FIG. 6 shows an expanded (redundant) 7-bit transmission signal and a signal constellation representing a signal point constellation, and FIG. 7 shows a signal constellation representing a transmission signal and a signal point constellation divided into a to h. .

On the receiving side, the operation of minimizing the error probability of the transmission data is performed based on the redundant signal and the code or rule. On the receiving side, this operation, that is, Viterbi decoding is used as a trellis decoding means. Viterbi decoding selects the point (judgment point) closest to the reception point for each subset, accumulates the distance between the reception point and the judgment point, and calculates a plurality of calculation values corresponding to the number of states of the convolutional decoder. This is a means for comparing and selecting the path corresponding to the smallest value of the calculation value, and after the lapse of a plurality of blocks, the minimum calculation value is traced back to the past and combined.

Among the above procedures, the calculation of the judgment point is called subset decoding, the accumulation / comparison of accumulated values / pass selection is called ACS operation, and the decoding traced back in the past is called path tracing. In this way, redundant bits and the regularity of the code are used to combine the signals with the highest probability of transmission and reduce transmission errors.

Next, subset decoding will be described. FIG. 15 shows the signal constellation including the receiving point. Subset decoding is to select points (a total of 8 points) in the subset for each of a to h closest to the reception point and store the distance between the selection point and the reception point. In FIG. 15, assuming that x is the reception point, the selected signals are a _b , b ₉ , and
_{c c, d 2, e 2} , f f, g 5, becomes h _c, the reception point x
And the distance to the above point are stored.

As described above, in the subset decoding, the distance between the receiving point and all the signal points (128 points in the case of V.17) is calculated, this distance is compared, and the shortest distance in each subset is calculated. I had selected a point. Therefore, 12
Calculation, comparison, and selection of distances were required as many as eight times, and a considerable processing time was required for the calculation. Further, in order to perform such an operation as fast as possible, it is necessary to store all the coordinates of 128 signal points, which requires a large-capacity storage element.

In a normal modem, the above digital signal processing is performed by software using a processor such as DSP. Therefore, the enormous processing time as described above requires a faster processor or a plurality of processors, and a large-capacity ROM as a processor having a large-capacity ROM as a large-capacity storage element or a large-capacity ROM outside the processor. .

In such a conventional method, a more expensive processor and an increase in the number of parts are required, which is also a cause of raising the cost of the modem alone.

In the present embodiment, by utilizing the regularity of the signal arrangement of the subset and providing the means for performing the subset decoding with only a small number of basic point arrangements, it is possible to operate at a higher speed and with a smaller number of storage elements. Subset decoding will be described. <Subset Composite Procedure> From FIGS. 7 and 15, the subsets e, c, and g are 90 degrees and 1 respectively of the subset a.
The subsets d, h, and b are equal to those rotated by 80 degrees and 270 degrees, and the subsets d, h, and b are 90 degrees and 180 degrees, respectively.
It turns out that it is equal to the one rotated by 270 degrees. This is V. It is a point arrangement determined in order to guarantee the rotation invariance of 90 degrees, 180 degrees, and 270 degrees of 17 modems, and this rule is used.

As can be seen from this rule, the subsets e, c, g and the subsets d, h, b can be easily obtained by rotating the subsets a, f. Therefore, in the case of subset decoding, it suffices to focus only on the subsets a and f. Here, the basic points used for the subset decoding will be further considered from the relationship of the point arrangement of the subsets a and f. This basic point arrangement is shown in FIG. Each point of the basic point arrangement is designated as z, and its subscript is the hexadecimal value of the upper 4 bits. The coordinates of the basic constellation (x _z, y _z) with a subset a of the coordinates (x _a, y _a), showing the coordinates of the subset _f (x f, y f) the relation between below. _{(X z, y z) =} (x a, y a +1) ... (4) (x z, y z) = (x f + 1, y f) ... (5) In addition, the basic constellation of FIG. 4, Is on the x-axis and y
It can be seen that the upper 4-bit value when the component is positive is equal to 1 when the y component is negative. That is, as long as the y component has a positive or negative point, it can be easily extended to other points (y component is negative or positive). The subset decoding according to the present embodiment is performed only on the quadrant (the upper half or the lower half of FIG. 4) in which the y component of the basic point arrangement is positive or negative. Subset decoding using a pointer whose y component is only negative will be described below.

Next, this embodiment will be described in detail with reference to FIGS.

FIG. 11 is a block diagram of the subset decoding of this embodiment, and FIG. 12 is a flow chart showing the subset decoding operation of this embodiment.

First, the received signal does not always become the normal signal constellation point due to the influence of the line noise or the line distortion, and the reception point (R = (x
_R , y _R )) are input, and 0 degree, 90 degree, and 18 degrees, respectively.
Signals R, R ', R''rotated 0 degrees and 270 degrees counterclockwise,
Rotators 1101, 1102, 11 to obtain R ′ ″
At 03, it rotates 90 degrees to the left (step S1201). The signals R, R ′, R ″, and R ′ ″ are expressed by the following equations.

[0084] _{R = (x R, y R} ) R '= (- y R, x R) R''= (- x R, -y R) R''' = (y R, -x R) further , The rotated signals R, R ', R ", R'" are
Only the Y-axis component is incremented by 1 by the Y-axis movers 1108, 1109, 1110, 1111 and the subset a of the reception points a,
e, c, g coordinates on the basic point arrangement, R _a , R _e , R _c ,
R _g is obtained, and similarly, the rotated signal is incremented by +1 only in the X-axis component by the X-axis movers 1104, 1105, 1106, 1107, and the subsets f, d, h, b of the reception points are obtained.
The coordinates R _f , R _d , R _h , and R _b on the basic point arrangement of are calculated (step S1202). The actual calculation formula is as follows.

[0085] _{R a = (x R, y} R +1) ... Y component of _{R +1 R e = (- y} R, x R +1) ... R ' of the Y component _{+1 R c = (- x R} , -y R +1) ... R '' Y component of _{+1 R g = (y R,} -x R +1) ... R 'Y component _{of''+1 R f = (x} R +1, y R) ... R X component +1 R _{d = (- y R + 1,} x R) ... R 'X component of _{+1 R h = (- x R} + 1, -y R) ... R' X component of _{'+1 R b = (y R} + 1, -x R) ... X component of R '''+ 1 By the above operation, eight coordinates on the basic point arrangement of the reception points corresponding to the subsets a to h can be obtained.

Next, in order to determining region y components performed in the negative region, it determines the sign of the y component of the signal R _a to R _h (coordinates) at y component code determiner 1112 (step S120
3). Code determination, if y component of the signal R _a to R _h is negative, the S _ya to S _yh is a judgment value corresponding to the +1, if y component of the signal R _a to R _h is positive, corresponding The judgment values S _{ya to} S _yh are set to -1, and they are output.

[0087] Then, the signal R _a y component and the sign determination value S _ya to S _yh of to R _h is, y is multiplied by the component multipliers 1114 signal R _a '~R _h' is calculated (step S120
4). This operation is expressed by the formula below.

[0088] _{R a '= (x a,} y a * S ya) However, R _a = (x _{a,
y a) R b '= (} x b, y b * S yb) However, R _b = (x _b,
y _b ) R _c '= (x _c , y _c * S _yc ), where R _c = (x _c ,
y _c ) R _d '= (x _d , y _d * S _yd ) where R _d = (x _d ,
y _d ) R _e '= (x _e , y _e * S _ye ) where R _e = (x _e ,
y _e ) R _f '= (x _f , y _f * S _yf ), where R _f = (x _f ,
y _f ) R _g '= (x _g , y _g * S _yg ), where R _g = (x _g ,
y _g ) R _h '= (x _h , y _h * S _yh ), where R _h = (x _h ,
In y _h) That, the above operations, the y-component of R _a to R _h are reversed sign of the y component only when positive, mapping y component of the fundamental constellation of R _a to R _h in the negative region Is completed.

Next, R _a 'to R _h ' are the area determiner 111.
3, the y component of the basic point arrangement is subjected to region determination in a negative region, and the determined determination point (decoded value) and signal R
_The errors E _{a to} E _h are calculated from the difference _{a to} R _h (step S1205).

Details of this area determination will be described with reference to FIGS. 13 and 14. FIG. 13 is a diagram showing a judgment area in the area where the y component is negative, and Z ₀ , Z ₂ , and
_{Z 4, Z 6, Z 8} , Z a, Z c, has eight points Z _e, represents the boundary line determination area by a chain line. FIG. 14 is a flowchart showing the flow of area determination.

First, when a reception point that is affected by noise or the like and can take an arbitrary point in the negative y component is input, the x component and the y component are compared. | X | ≦ 2 and y ≧ −
In the case of 4, Z ₆ is selected as the judgment point (S1402),
When | x | ≦ 2 and y <−4, Z _c is selected as the determination point (S1404), | x | ≦ 6 and y ≧ −4, and
If x ≧ 0, Z ₄ is selected as the determination point (S140
7), | x | ≦ 6 and y ≧ −4, and x <0,
Z ₂ is selected as the determination point (S1408).

Further, in the cases other than Z ₆ , Z _c , Z ₄ and Z ₂ above, and when x ≧ 0 and y ≧ −x + 2, Z ₈ is selected as the judgment point (S1411) and x ≧ 0. And y <
In the case of −x + 2, Z _e (1412) is selected as the determination point.

Further, in the case other than Z ₆ , Z _c , Z ₄ and Z ₂ above, and when x <0 and y ≧ x + 2, Z ₀ is selected as the judgment point (S1414) and x <0 and y <x
In the case of +2, Z _a is selected as the determination point (S141
5).

Finally, the difference between the reception point and the judgment point is incorrect.
It is calculated as a difference (S1416). For example, in FIG.
If * is the reception point of subset a, Z is the decision point_Four But
As an error, (x_a '-4, y_a ’+2) is calculated
To be done. Where (X_a ’, Y _a ′) Is the receiving point (R_a ’)
Coordinates, (4, -2) is Z_Four Are the coordinates of.

Further, eight kinds of receiving points R _{a to} R _h can be considered. For example, when the current receiving point corresponds to R _a , the lower 3 bit value of the subset a is 000, and Z ₄ If it is determined that the upper 4 bits are 0100
Therefore, the decoded value is 0100000.

[0096] Upon obtaining this way the error, followed by the determination value from the area decision unit 1113 y components _{(H a} ~H _h) is, the y component multipliers 1115 y component code determiner 111
2 outputs (S _{ya to} S _yh ), and when the upper 4 bits of the corresponding code value is the code determination value −1,
It is incremented by 1 (step S1206). For example, in the above example, when the y component of _Ra is positive (S _ya = -1), the final decoded value is 0101000, and thus the final decoded value is obtained.

Area determination (decoding) and error calculation are similarly performed for other subsets.

The decoded value and the error are processed by the Viterbi decoding and the equalizer (neither is shown).
Therefore, the subset decoding only needs to store the above-mentioned decoded value and the error, and is stored in a memory such as a RAM.

As described above, according to the decoding method of the present embodiment, the number of data points used for area determination is reduced from the conventional 128 points to 8 points, and the simple operation of code determination makes
The area can be easily determined. Therefore, it is possible to provide the subset decoding in which the storage capacity required for storing the determination data points is significantly reduced as compared with the conventional art, by only a simple operation of code determination.

In this embodiment, the bit rate is 14.
Although only the case of 4 kbps is shown, the present invention is not limited to this, and V. 12.0k recommended in 17
It can be easily applied to the cases of 9.6k and 7.2kbps.

However, in the case of 12.0 k and 9.6 kbps, it is necessary to rotate the finally output rotator 45 degrees to the left and double the amplitude in order to make the basic point arrangement the x-axis. There is.

Furthermore, 12.0k, 9.6k, 7.2k
The conversion constants from the upper 4-bit value in which the y component of the basic point arrangement of bps is negative to the upper 4-bit value in which the y component is positive are +4, -1, and +1 respectively. Therefore, the value finally added to the upper 4 bits differs depending on the bit rate.

V. 17 modems 12.0k, 9.6k,
The number of points of 7.2 kbps is 64, 32,
The total number of points is 16, and the total number of points is 240 points including 128 points of 14.4 kbps, and the number of words in the memory to be stored as specific points is 480 words.

According to the present embodiment, the number of points at 14.4 kbps is 8 as well as 12.0 k and 9.6.
The number of points for k and 7.2 kbps is 4, 2 and 4, respectively.
This is 1 point, and as a result, the number of words in the memory is reduced from the conventional 480 words to 30 words.

As described above, according to this embodiment,
Compared to the conventional method, subset decoding can be done in 1/3 the calculation time. The storage capacity at the judgment point can be reduced to 1/16. For example,
V. In the case of 17 modems, the memory capacity, which conventionally required 480 words, can be reduced to 30 words.

Therefore, the modem using this embodiment is
It can be realized with a simpler configuration than the conventional one. When the modem is implemented by software, it can be implemented by a processor that has a lower speed and a smaller memory capacity than before, and a lower cost modem can be provided. [Fourth Embodiment] In the fourth embodiment, similarly to the third embodiment, a description will be given of subset composition which can reduce the required memory capacity and can be processed by a low-speed processor.

From FIGS. 7 and 15, the subsets e, c,
g is 90 degrees, 180 degrees, and 27 degrees of the subset a, respectively.
Equal to rotated 0 degrees, and the subsets d, h, b are
It can be seen that it is equal to the subset f rotated by 90 degrees, 180 degrees, and 270 degrees, respectively. This is V. It is a point arrangement determined in order to guarantee the rotation invariance of 90 degrees, 180 degrees, and 270 degrees of 17 modems, and this rule is used.

As can be seen from this rule, the subsets e, c, g and the subsets d, h, b can be easily obtained by rotating the subsets a, f. Therefore, in the case of subset decoding, it suffices to focus only on the subsets a and f. Here, the basic points used for the subset decoding will be further considered from the relationship of the point arrangement of the subsets a and f. This basic point arrangement is shown in FIG. Each point of the basic point arrangement is designated as z, and its subscript is the hexadecimal value of the upper 4 bits. The coordinates of the basic constellation (x _z, y _z) with the subset a coordinate (x _a, y _a), showing the coordinates of the subset _f (x f, y f) the relation between below. _{(X z, y z) =} (x a, y a +1) ... (4) (x z, y z) = (x f + 1, y f) ... (5) According to the present embodiment, the subset composite all It is done on this basic point.

Next, this embodiment will be described in detail with reference to FIGS.

FIG. 16 is a block diagram of the subset decoding of this embodiment, and FIG. 17 is a flow chart showing the subset decoding operation of this embodiment.

First, the received signal does not always become the normal signal constellation point due to the influence of the line noise and the line distortion, and the reception point (R = (x
_R , y _R )) are input, and 0 degree, 90 degree, and 18 degrees, respectively.
Signals R, R ', R''rotated 0 degrees and 270 degrees counterclockwise,
Rotators 1601, 1602, 16 to obtain R ′ ″
At 03, it rotates 90 degrees to the left (step S1701). The signals R, R ′, R ″, and R ′ ″ are expressed by the following equations.

[0112] _{R = (x R, y R} ) R '= (- y R, x R) R''= (- x R, -y R) R''' = (y R, -x R) further , The rotated signals R, R ', R ", R'" are
Only the Y-axis component is incremented by 1 by the Y-axis movers 1608, 1609, 1610, 1611, and the subset a of the reception points a,
e, c, g coordinates on the basic point arrangement, R _a , R _e , R _c ,
R _g is obtained, and similarly, the rotated signal is incremented by only the X-axis component by the X-axis movers 1604, 1605, 1606, 1607, and the subsets f, d, h, b of the reception points are obtained.
The coordinates R _f , R _d , R _h , and R _b on the basic point arrangement of are obtained (step S1702). The actual calculation formula is as follows.

[0113] _{R a = (x R, y} R +1) ... Y component of _{R +1 R e = (- y} R, x R +1) ... R ' of the Y component _{+1 R c = (- x R} , -y R +1) ... Y component of R '' +1 R _g = (y _R , -x _R +1) ... Y component of R '''+1 R _f = (x _R +1, y _R ) ... X component of R +1 R _{d = (- y R + 1,} x R) ... R 'X component of _{+1 R h = (- x R} + 1, -y R) ... R' X component of _{'+1 R b = (y R} + 1, -x R) ... X component of R '''+ 1 By the above operation, eight coordinates on the basic point arrangement of the receiving points corresponding to the subsets a to h can be obtained.

[0114] Next, it is determined x-component of the signal R _a to R _h, of the sign of the y component y component code determiner 1612 (step S1703~S1705). The code determination, i.e., is to determine the affiliation quadrant signals R _a to R _h (first quadrant through fourth quadrant), for example, respectively as the determination result for the first quadrant through fourth quadrant 0 (binary 00), 1
(Binary 01), 3 (binary 11), and outputs the 2 (binary 10) as the determination result signal S _a to S _h. Then the signal R _a
˜R _h and signals S _a ˜S _h are input to the area determiner in pairs, and the area determination is performed in the quadrants determined by the signals S _a ˜S _h , and the determined determination points and the signals R _a ˜R are obtained. _The errors E _{a to} E _h are calculated from the difference of _h (step S170).
6-S1708). For example, when the code determination result is 0, that is, the first quadrant is selected,
Will be explained.

FIG. 18 is a diagram showing a judgment area in the first quadrant, where Z ₅ , Z ₇ , Z ₉ and Z are the judgment points.
_{It has d} and Z _f , and the boundary line of the judgment area is represented by a one-dot chain line. FIG. 18 is a flowchart showing the flow of area determination in the first quadrant.

First, when a reception point which can be any point in the first quadrant due to noise or the like is input, its x component,
When y components are compared and x ≦ 2 and y ≦ 4 (step S1901), Z ₇ is selected as a determination point (S190).
2), if x ≦ 2 and y> 4 (S1903), Z _f is selected as the determination point (S1904), x ≦ 6 and y ≦ 4.
In the case of (S1905), Z ₅ is selected as the judgment point (S1905).
1906), and x> 6 and y> 4 and y ≦ x−2 (S1907), Z ₉ is selected as the determination point (S190).
8) In all other cases, Z _d is selected as the determination point (S1409). Finally, the difference between the reception point and the judgment point is calculated as an error (S1910).

For example, when * in FIG. 18 is the reception point of the subset a, Z ₅ is selected as the decision point and (x
_a -4, y _a -2) is calculated. Where (X _a , y
_a ) is the coordinates of the receiving point (R _a ), and (4, 2) is the coordinates of Z ₅ .

Further, eight kinds of receiving points R _{a to} R _h can be considered. For example, when the current receiving point corresponds to R _a , the lower 3 bit value of the subset a is 000, and Z ₅ If it is determined that the upper 4 bits are 0101
Therefore, the decoded value is 0101000. In this way, the final decoded value is obtained.

Area determination (decoding) and error calculation are similarly performed in the other quadrants.

The decoded value and the error are processed by the Viterbi decoding and the equalizer (neither is shown).
Therefore, the subset decoding only needs to store the above-mentioned decoded value and the error, and is stored in a memory such as a RAM.

As described above, according to the decoding method of the present embodiment, the number of data points used for area determination is reduced from 128 points in the prior art to 16 points, and the area determination is limited to one quadrant including the receiving point. The area determination of 16 points can be reduced to the area determination of 5 points. Therefore, as compared with the related art, it is possible to significantly reduce the storage capacity required to store the decision data points, and to provide the subset decoding that enables faster area decision.

In the present embodiment, only the case where the bit rate is 14.4 kbps is shown, but the present invention is not limited to this, and for example, V.50. 12.0 recommended in 17
It can be easily applied to the cases of k, 9.6k, and 7.2 kbps.

However, in the case of 12.0 k and 9.6 kbps, in order to target the basic point arrangement in the first quadrant to the fourth quadrant, the reception point is set to the right 45 before or after step S1701 in FIG. Rotate a degree and change the amplitude to route 2
It is necessary to add a means for doubling. This processing is performed by the following formula.

(X ', y') = ((1/2). (X +
y), (1/2) · (−x + y)) As described above, according to this embodiment, compared to the conventional case,
Subset decoding in 1/3 operation time. The storage capacity at the judgment point can be reduced to 1/8. Therefore, the modem using this embodiment can be realized with a simpler configuration than the conventional one. When the modem is implemented by software, it can be implemented by a processor that has a lower speed and a smaller memory capacity than before, and a lower cost modem can be provided.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0126]

As described above, according to the mapping method and the modulation / demodulation device of the present invention, the amount of data for storing the mapping points required for mapping can be reduced, and a more inexpensive modulation / demodulation device can be provided. effective.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a mapping unit according to a first embodiment.

FIG. 2 is a diagram showing a table arrangement of a memory used in the first embodiment.

FIG. 3 is a flowchart showing the operation of the mapping means according to the first embodiment.

FIG. 4 is a diagram of a signal constellation of a basic point used in an example.

FIG. 5: It is a figure of 14.4 Kbps signal constellation of 17 recommendation.

FIG. 6 is a diagram showing correspondence between code values and subset values and signal constellations.

FIG. 7 is a diagram showing correspondence between code values, subset values, and signal constellations.

FIG. 8 is a schematic configuration diagram of a mapping unit according to the second embodiment.

FIG. 9 is a diagram showing a table arrangement of a memory used in the second embodiment.

FIG. 10 is a flowchart showing the operation of the mapping means according to the second embodiment.

FIG. 11 is a schematic configuration diagram of a subset decoding unit according to the third embodiment.

FIG. 12 is a flowchart showing the operation of the subset decoding means according to the third embodiment.

FIG. 13 is a diagram illustrating a determination region in a region where y component <0 according to the third embodiment.

FIG. 14 is a flowchart showing a determination operation in a region of y component <0 according to the third embodiment.

FIG. 15 is a diagram showing a relationship between a reception point and a signal constellation.

FIG. 16 is a schematic configuration diagram of a subset decoding means according to the fourth embodiment.

FIG. 17 is a flowchart showing the operation of the subset decoding means according to the fourth embodiment.

FIG. 18 is a diagram showing a determination area in a first quadrant of the fourth embodiment.

FIG. 19 is a flowchart showing the determination operation in the first quadrant of the fourth embodiment.

[Explanation of symbols]

 Reference numeral 100 is an upper bit / lower bit divider, 101 is a memory address generator, 102 is a memory (storage element), 103 is a subset determiner, 104 is an adder, 105 is a rotator.

Claims (14)

[Claims] 1. A mapping method using a trellis code, comprising a selecting step of selecting a point within a basic coordinate having a point number of a predetermined subset of all points at a desired communication speed, and the selected point. Is expanded to a point on a predetermined signal constellation according to each subset, and a mapping step. 2. The selecting step stores a point in a basic coordinate in a memory, determines a read position of the memory from an uncoded bit, and reads a point on the basic coordinate corresponding to the uncoded bit. The mapping method according to claim 1, wherein: 3. The step of determining the subset from the encoded bits, obtaining the shift amounts of the x and y components from the basic coordinates, and obtaining the rotation angle from the basic coordinate point to the subset in the expanding step. 2. The mapping according to claim 1, further comprising: an adding step of adding the read basic coordinate points and the shift amount; and a rotating step of rotating the addition result of the adding step at the rotation angle. Method. 4. A modulation / demodulation apparatus using a trellis code, comprising: selecting means for selecting a point within a basic coordinate having a predetermined subset number of points out of all points at a desired communication speed; and the selected point. And expanding means for expanding to a point on a predetermined signal constellation according to each subset. 5. A mapping method using a trellis code, wherein the y component in the basic coordinates having the number of points of a predetermined subset of all points at a desired communication speed is either positive or negative. And a developing step of expanding the selected points to points on a predetermined signal constellation according to each subset. 6. The selecting step stores in memory a point whose y component in the basic coordinate is either positive or negative.
Determining the address of the memory by adding an even-odd or predetermined threshold value of uncoded bits, reading the point from the memory according to the address, and using the y component of the read point as the determination result. 6. The mapping method according to claim 5, further comprising a step of multiplying with a corresponding multiplication value. 7. The step of determining the subset from the encoded bits, obtaining the shift amounts of the x and y components from the basic coordinates, and obtaining the rotation angle from the basic coordinate point to the subset in the expanding step. An adding step of adding the read basic coordinate point and the shift amount,
The mapping method according to claim 5, further comprising a rotation step of rotating the addition result of the addition step at the rotation angle. 8. A mapping method using a trellis code, wherein the y component in the basic coordinates having the number of points of a predetermined subset of all points at a desired communication speed is either positive or negative. The modulation / demodulation device is characterized by comprising: selection means for selecting the points of (1) and expansion means for expanding the selected points to points on a predetermined signal constellation according to each subset. 9. A subset decoding method using a trellis code, comprising a conversion step of converting reception points corresponding to each subset into basic coordinates having a point number of one subset of all points at a desired communication speed. , The determination step of determining the sign of the y component of the converted reception point, and the reception point converted by the determination value of the determination step
A mapping step of mapping a component into a positive or negative region, an evaluation step of evaluating the reception point mapped by the mapping step in a region where the y component of the basic coordinates is either positive or negative, and the evaluation step And a reverse mapping step of reverse mapping the determined determination value to the negative or positive region of the y component according to the code determination result of the determination step. 10. The conversion step includes receiving points of 0 degrees and 90 degrees.
Rotated 180 degrees, 270 degrees counterclockwise, and a value determined in advance from the positional relationship between the basic coordinates and the two subsets, x component and y
10. The subset decoding method according to claim 9, wherein the components are shifted. 11. A modulation / demodulation apparatus using a trellis code, wherein the conversion means converts a reception point corresponding to each subset into basic coordinates having the number of points of one subset of all points at a desired communication speed, The judging means for judging the sign of the y component of the converted receiving point and the receiving point converted by the judgment value in the judging step are
Mapping means for mapping a component to a positive or negative region, evaluation means for evaluating the receiving point mapped by the mapping process in a region where the y component of the basic coordinates is either positive or negative, and the evaluation process An inverse mapping unit that inversely maps the determined determination value into the negative or positive region of the y component according to the sign determination result of the determination step. 12. A subset decoding method using a trellis code, comprising: a conversion step of converting reception points corresponding to each subset into basic coordinates having the number of points of one subset of all points at a desired communication speed. Determining the sign of the x component and the y component of the converted reception point, and determining the area of the converted reception point in one quadrant of the basic coordinates by the judgment value determined by the judgment step. A subset decoding method comprising: 13. The conversion step comprises setting a reception point at 0 degrees and 90 degrees.
Rotated 180 degrees, 270 degrees counterclockwise, and a value determined in advance from the positional relationship between the basic coordinates and the two subsets, x component and y
10. The subset decoding method according to claim 9, wherein the components are shifted. 14. A modulation / demodulation apparatus using a trellis code, wherein the conversion means converts a reception point corresponding to each subset into a basic coordinate having the number of points of one subset of all points at a desired communication speed. A conversion unit for determining the signs of the x component and the y component of the converted receiving point, and determining the converted receiving point in one quadrant of the basic coordinates based on the determination value determined in the determination step. A modulation / demodulation device characterized by the above.