WO2009006812A1

WO2009006812A1 - Encoding modulation method,decoding method and apparatus of trellis coded modulation code

Info

Publication number: WO2009006812A1
Application number: PCT/CN2008/071349
Authority: WO
Inventors: Dongmei Fang
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2007-07-10
Filing date: 2008-06-18
Publication date: 2009-01-15
Also published as: CN101123439A

Abstract

Encoding modulation method, decoding method and apparatus of trellis coded modulation code are disclosed. The encoding method comprises steps of establishing a QAM constellation picture which is TCM constellation map picture with modulation level exceeding 4; making encoded bit groups of the constellation picture with I route and Q route to be arranged periodically and repeatedly in the I route and Q route; making un-encoded bit groups which correspond to encoded bit groups with period repeat arrangement to be different from each other; obtaining constellation index including the encoded and un-encoded bit groups according to the original bit information groups; determining a signal point as finally transmitted TCM signal in the constellation map. The symbols of parallel route are mapped evenly on the Q route of the constellation map because the encoded bit groups are arranged periodically and repeatedly, thus the performance of TCM code is improved.

Description

Coded modulation, decoding method and device for trellis coded modulation code

The invention belongs to the technical field of data coding, and in particular relates to a code modulation and decoding method and device for a trellis coded modulation code. Background technique

TCM (Trellis coded modulation) coding, or trellis coded modulation technology, is a "signal set space coding" that can encode and modulate without reducing the band utilization and power utilization. Combined, the redundancy of the signal set is used to obtain error correction capability.

The TCM code has two basic characteristics:

1. The number of signal points used in the constellation diagram is greater than the number of points required to encode the same type of modulation. These additional signal points provide redundancy for error correction coding.

2. Using a convolutional code to introduce a dependency between successive signal points, so only certain signal point sequences are allowed. These allowed signal sequences can be modeled as a grid structure, hence the name grid. Code modulation.

The basic idea of TCM is to use an extended signal set to provide controllable coding redundancy, and to design convolutional coding and multi-modulation mapping uniformly, that is, to perform hierarchical segmentation mapping on the transmitted signal point set to make the coded signal sequence free. Euclidean distance (the squared Euclidean distance is the square of the distance between two signal points, the cylinder is called the Euclidean distance, the free Euclidean distance is defined as the bifurcation from the same state at zero time, and the two at the subsequent time have a convergence The minimum Euclidean distance between sequences of infinitely long signal points.) Maximizes.

As shown in Figure 1, the TCM signal is generated as follows: At a time n, the original bit block of length k enters the encoder, and one of them enters the convolutional encoder bit (≤ k) through a code rate. The convolutional encoder is expanded into a coded ratio of the encoded bits. Special group, the m coded bits are used to select one of the subsets of the 2 ^k+m - ^k modulation signal set, and the remaining pass bits form an uncoded bit group for use in the subset Select one of the 1" signal points as the final transmitted TCM signal. When k, ² "parallel branch branches will appear in the grid map of the TCM code (the start and end states of the branch branch are The same, commonly referred to as ² ^ parallel paths), which causes a single Error Event to occur (corresponding to the Multiple Error Event). The signal points in the same subset correspond to the TCM coded output signal points on a parallel path, that is, the ² "signal points on the constellation map that are identical in coding are corresponding to a bundle of parallel paths. The TCM encodes the output signal point.

The division principle of the above subset (also called the set division principle) is of great significance in the construction of the TCM scheme. The so-called set segmentation is to divide a signal set into smaller subsets one after another, and to maximize the minimum Euclidean distance in the segmented subset.

The principle of set partitioning should follow the following two principles:

1. In the same level subset, the number of signal points and the Euclidean distance distribution contained in each subset should be consistent;

2. With the division of the subset, in the smaller subset, the minimum Euclidean distance between the signal points should increase step by step. When designing the TCM scheme, the modulation signal set with ² "^ signal points is divided into m stages. Let the subset ^Ω after the i-th stage division, the inner minimum Euclidean distance is ^Δ '· (^ ⁰ ' ¹ '"'' ^m )

(When =, the first subset contains only one signal point, and at this time ^Δ TM ^{= ∞} ), then there is W i'-^ .

After the division of the subset of stages is completed, the number of encoded bits corresponding to the signal points in the same subset is . . . 3 is the same (it can be said that the same signal points corresponding to the bit groups after encoding correspond to the same subset ), and in order to distinguish ² "signal points in the same subset, the corresponding uncoded bit groups of each signal point must be different. Corresponding to the constellation map, the coding ratio The 1" signal points of the same group belong to the same subset, and the corresponding - uncoded bit groups must be different. On the constellation map, all subsets need to conform to the set partitioning principle.

2 is a schematic diagram of a set division of a 16QAM (quarature amplitude modulation) signal point set.

The grid diagram of the TCM optimal code should follow the following basic construction principles:

1. All modulated signals should have the same frequency of occurrence and should have as much regularity and symmetry as possible. This principle indicates that a good TCM code should have a regular structure. This is because the TCM scheme is actually a scheme for optimal segmentation of the signal space, and the modulation signal space is symmetric, so the optimal segmentation scheme should also Regular and symmetrical.

2. The corresponding signal of the branching branch starting from the same state shall belong to the same subset ^β after the first level set. Or ^β ι , which guarantees that the distance between different branches separated from the same state is greater than or equal to ^Δ ι.

3. The corresponding signal of the branch branch arriving at the same state shall belong to the same subset ^β after the first level set. Or ^β ι , which guarantees that the distance between different branches reaching the same state is greater than or equal to ^.

4. The parallel path corresponds to the subset after the segmentation. This will ensure that the distance between parallel paths is greater than or equal to ^Δ TM .

For the TCM code in the single-carrier (Wimax SCa) modulation mode with enhanced global microwave access interoperability, the code rate and modulation method are: 1/2 16QAM, 3/4 16QAM, 2/3 64QAM, 5/6 64QAM , 3/4 256QAM, 7/8 256QAM.

The convolutional encoder used in the TCM code under Wimax SCa is a (2, 1, 7) binary convolutional code generator with a code rate of 1/2 and a beam length of 7. Its structure is shown in Figure 3.

The 5/6 64QAM, 7/8 256QAM TCM code needs to be punctured after the convolutional encoder with a code rate of 1/2, that is, a part of the encoded bit is deleted, and the code rate is changed to 3/4. The punch pattern is X1Y1Y2X3, that is, every 3 groups of coded bits are punched, the first group is not punched, the second group is knocked off X, and the third group is knocked off Y.

Which for 2/3 64QAM, 5/6 64QAM, 3/4 256QAM, 7/8 256QAM The encoding and modulation process of the TCM code is as follows.

2/3 64QAM TCM encoder structure diagram shown in Figure 4.

A raw bit field of length 2 uluO enters the TCM encoder, and the TCM encoder part uses a convolutional code of 1/2. The input order of uluO is ul is the first arrival bit, and uO is the backward arrival bit. Ul is a pass-through bit, uO is 1/2 convolutional coding for entering the convolutional encoder bit, and a set of 3-bit constellation index b5b4b3 is obtained, which is mapped to the I path, which is related to the coordinates of the I path, where b4b3 is the encoded bit group. , b5 is an uncoded bit group. Then another set of original bit information group uluO is input to the encoder to obtain another set of 3-bit constellation index b2blb0, which is mapped to the Q path, which is related to the coordinates of the Q path, where blbO is the encoded bit group and b2 is the uncoded bit. group. Since it produces a 6-bit output every 4 input bits, the encoder should be called an encoder with a code rate of 4/6.

Based on the obtained I-way constellation index b5b4b3 and the Q-way constellation index b2blb0, signal points are determined in the existing 64QAM constellation map and used as the final transmitted TCM signal. The existing 64QAM constellation map is shown in Figure 5. The signal points of the same group of the I-coded bit group b4b3 and the Q-channel coded bit group blbO are the same subset, for example, the I channel and the Q channel are (110, 001), (010, 001), (110, respectively) The signal points of 101, (010, 101) are the same subset, wherein the bit group b4b3 after the I channel coding is 10, and the bit group blbO is 01 after the Q channel coding, and the I channel and the Q channel are respectively (110, The signal points of 011), (010, 011), (110, 111), (010, 111) are the same subset, wherein the bit group b4b3 after the I channel coding is 10, and the bit group blbO after the Q channel coding is 11, through It is not difficult to find these two subsets that the TCM coded output signal points (the TCM coded output signal points on the parallel path, that is, the signal points in the same subset) on the two parallel paths are not uniformly mapped on the Q path. The Euclidean distance of the symbols (that is, the signal points in the same subset) is also long and short. The I and Q paths are a subset of the signal points of (110, 001), (010, 001), (110, 101), and (010, 101), respectively, and the Euclidean distance is compared with the I and Q paths respectively ( The Euclidean distance of the subset of signal points of 110, 011), (010, 011), (110, 111), (010, 111) is short. 5/6 64QAM TCM encoder structure diagram shown in Figure 6.

A raw bit information group u4u3u2ulu0 of length 5 enters the TCM encoder, and the TCM encoder part uses a convolutional code with a code rate of 3/4 (the original 1/2 convolutional code is punctured to make the code rate become 3/4) . The input order of u4u3u2ulu0 is u4, u3, u2, ul, u0. U2ulu0 performs 3/4 convolutional coding, and the output is c3c2clc0. The uncoded bit u4 and the encoded bit c3c2 are combined into a set of 3-bit constellation index b5b4b3, which is mapped to the I path, where b4b3 is the encoded bit group and b5 is the uncoded bit group. U3clc0 is a set of 3-bit constellation index b2blb0, mapped to the Q path, where blbO is the encoded bit group and b2 is the uncoded bit group.

Based on the obtained I-channel constellation index b5b4b3 and the Q-way constellation index b2blb0, the signal point is determined in the existing 64QAM constellation map and used as the final transmitted TCM signal. The existing 64QAM constellation map is shown in Figure 5.

The structure of the 3/4 256QAM TCM encoder is shown in Figure 7.

A raw bit block u2ulu0 of length 3 enters the TCM encoder, and the TCM encoder section uses a convolutional code of 1/2. The input order of u2ulu0 is: u2, ul, u0. U2ul is a pass-through bit, u0 performs 1/2 convolutional coding, and a set of 4-bit constellation index b7b6b5b4 is obtained, which is mapped to the I path, which is related to the coordinates of the I path, where b5b4 is the encoded bit group, and b7b6 is the uncoded bit group. . Then, a set of original bit information groups u2ulu0 is input to the encoder, and another set of 4-bit constellation index b3b2blb0 is generated, which is mapped to the Q path, which is related to the coordinates of the Q path, where blbO is the encoded bit group, and b3b2 is the uncoded bit. group. Since it produces an 8-bit output for every 6 input bits, the encoder should be called an encoder with a 6/8 bit rate.

Based on the obtained I-channel constellation index b7b6b5b4 and the Q-way constellation index b3b2blb0, the signal point is determined in the existing 256QAM constellation map and used as the final transmitted TCM signal. The current 256QAM constellation map is shown in Figure 8.

The structure of the 7/8 256QAM TCM encoder is shown in Figure 9.

A raw bit information group u6u5u4u3u2ulu0 of length 7 enters the TCM encoder. The TCM encoder section uses a convolutional code with a code rate of 3/4 (the original 1/2 convolutional code is punctured to make the code rate 3/4). The input order of u6u5u4u3u2ulu0 is u6, u5, u4, u3, u2, ul, u0. U2ulu0 performs 3/4 convolutional coding, and the output is c3c2clc0. The uncoded bit group u6u5 and the encoded bit group c3c2 are combined into a set of 4-bit constellation index b7b6b5b4, which is mapped to the I path, where b5b4 is the encoded bit group and b7b6 is the uncoded bit group. U4u3clc0 is combined into a set of 4-bit constellation index b3b2blb0, which is mapped to the Q path, where blbO is the encoded bit group and b3b2 is the uncoded bit group.

The decoding process of the TCM code can be roughly divided into the following two steps: subset decoding; soft decision Viterbi decoding.

Subset decoding

Among the ² " signal points of each subset of the constellation map of Fig. ^{2, the} signal point with the smallest Euclidean distance from the received signal symbol is found.

2. Soft decision Viterbi decoding

In the result of the subset decoding, a soft decision metric is obtained for ² TM coded bit groups, and the metric value is usually a subset decoded signal point corresponding to each coded bit group (the subset and the received symbol) The Euclidean distance of the signal point with the smallest Euclidean distance or its linear transformation. Substituting the above soft decision metric into the Viterbi decoder for decoding operation, an estimated value (decoding result) of the original bit block can be obtained.

A defect of the prior art TCM codes with modulation orders greater than 4 (such as 2/3 64QAM, 5/6 64QAM, 3/4 256QAM, 7/8 256QAM TCM codes) is that their coding and modulation are based on existing The constellation map determines the signal point, determines the final transmitted TCM signal, and the TCM coded output signal points (that is, the signal points in the same subset) on each parallel path are not uniformly mapped on the Q path, which makes the Q path have children. The Euclidean distance of the set is long, the Euclidean distance of some subsets is short, and the shortest is the adjacent constellation. The distance between the points, as shown in FIG. 10A, the encoded bit groups of the four signal points marked are the same, and are the signal points in the same subset, and the distance of the Q path of the signal points in the subset is 1 ( The Euclidean distance between adjacent signal points in the constellation map is 1). As shown in another subset of the constellation map shown in FIG. 10B, the distance of the Q path of the signal points in the subset is 7. As can be seen from the comparison of Fig. 10A and Fig. 10B, it is apparent that the subset of the Q-path has a long Euclidean distance, and some of the subsets have a short Euclidean distance, which deteriorates the overall performance of the TCM code. The decoding process is decoded according to the existing constellation diagram. If the Euclidean distance of the TCM encoded output signal point on the parallel path is short, the decoding error is high, and this does not conform to the TCM described above. The modulation signal in the optimal code construction principle should have as much regularity and symmetry as possible. Summary of the invention

In order to solve the problem that the Euclidean distance of the subset on the Q-path is long, the Euclidean distance of some subsets is short, and the shortest distance is the distance between adjacent constellation points, so that the overall performance of the TCM code is deteriorated, the embodiment of the present invention A coding and modulation method for a TCM is provided, including:

Establishing a quadrature amplitude modulation QAM constellation with a modulation order greater than 4 as a constellation map of the TCM, so that the coded bit groups mapped on the constellation map I and Q are periodically repeated on the I and Q paths, respectively. Arranging, and making uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle different from each other;

Obtaining a constellation index including the encoded bit group and the uncoded bit group according to the original bit information group;

A signal point is determined in the constellation map as a final transmitted TCM signal based on the constellation index.

The embodiment of the invention further provides a method for decoding a trellis coded modulation code (TCM), which includes:

Establishing a quadrature amplitude modulation QAM constellation with a modulation order greater than 4 as a constellation map of the TCM, so that the coded bit groups mapped on the constellation map I and Q are in the I The path and the Q path are repeatedly arranged in a cycle, and the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle are different from each other;

Decoding the signal to be decoded according to the constellation map.

In order to solve the problem that the Euclidean distance of the subset on the Q-path is long, the Euclidean distance of some subsets is short, and the shortest distance is the distance between adjacent constellation points, so that the overall performance of the TCM code is deteriorated, and the present invention is implemented. The example also provides a code modulation device for a TCM, including:

a storage module: configured to store a quadrature amplitude modulation QAM constellation with a modulation order greater than 4 as a constellation map of the TCM, and map the encoded bit groups on the constellation map I and Q on the I path and Q Repeatedly arranged on the road, respectively, and the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle are different from each other;

a constellation index obtaining module: configured to obtain a constellation index including the encoded bit group and the uncoded bit group according to the original bit information group;

The TCM signal confirmation module is configured to determine a signal point in the constellation map as a final transmitted TCM signal according to the constellation index.

The embodiment of the invention further provides a decoding device for a trellis coded modulation code (TCM), which comprises:

Decoding module: configured to perform decoding processing on the signal to be decoded according to the constellation map. It can be seen from the specific implementation provided by the above invention that it is precisely because the post-encoding bit group period is repeatedly arranged and the corresponding uncoded bit groups are different from each other, so that the symbols of the parallel path are uniformly mapped on the constellation map, so that the TCM code is Overall performance has improved. DRAWINGS 1 is a schematic diagram of a prior art TCM code modulation structure;

2 is a schematic diagram of a 16QAM set partitioning in the prior art;

3 is a schematic diagram showing the structure of a prior art code rate 1/2 convolutional encoder;

4 is a schematic structural diagram of a prior art 2/3 64QAM TCM encoder;

5 is a 64QAM constellation map of a prior art TCM;

6 is a schematic structural diagram of a prior art 5/6 64QAM TCM encoder;

7 is a schematic structural diagram of a prior art 3/4 256QAM TCM encoder;

8 is a 256QAM constellation map of a prior art TCM;

9 is a schematic structural diagram of a prior art 7/8 256QAM TCM encoder;

10A is a schematic diagram of a prior art parallel path distance;

10B is a schematic diagram of a prior art parallel path distance;

FIG. 11 is a flowchart of a method according to a first embodiment of the present invention;

12A is a 64QAM constellation map of a TCM according to a first embodiment of the present invention; FIG. 12B is a 64QAM constellation map of a TCM according to a first embodiment of the present invention; FIG. 12C is a 64QAM of a TCM according to a first embodiment of the present invention; FIG. 13A is a 256QAM constellation map of a TCM according to a first embodiment of the present invention; FIG. 13B is a 256QAM constellation map of a TCM according to a first embodiment of the present invention; FIG. 14 is a second embodiment of the present invention. A schematic diagram of determining a signal point with the smallest Euclidean distance during decoding;

15 is a schematic diagram of comparison of 2/3 64QAM TCM code BER simulation results according to an embodiment of the present invention;

16 is a schematic diagram of comparison of 5/6 64QAM TCM code BER simulation results according to an embodiment of the present invention;

FIG. 17 is a schematic diagram showing comparison of BER simulation results of 3/4 256QAM TCM codes according to an embodiment of the present invention; FIG.

FIG. 18 is a comparison of BER simulation results of 7/8 256QAM TCM codes according to an embodiment of the present invention. Schematic diagram

FIG. 19 is a schematic structural diagram of a device according to a third embodiment of the present invention; FIG.

FIG. 20 is a schematic structural diagram of a device according to a fourth embodiment of the present invention. detailed description

The preferred modulation mapping method of the TCM coding structure of 2/3 64QAM, 5/6 64QAM, 3/4 256QAM, 7/8 256QAM proposed by the embodiment of the present invention is not only a kind of only one kind, but a certain mapping rule is satisfied. A collection of modulation mapping methods. This mapping rule is described below.

2/3 64QAM, 5/6 64QAM, 3/4 256QAM, 7/8 256QAM in the prior art

The number of bits of the encoded bit group of the TCM code is four (the encoded bit group refers to one or more output bits after convolutional coding), and is mapped to the I and Q paths on the constellation, respectively. Each channel maps 2 bits. Let the modulation order be M, then the number of bits of the uncoded bit group is (M-4), and the number of bits mapped on each path of I/Q is (M-4)/2 (uncoded bit group refers to Is a pass-through bit, one or more output bits that are not convolutionally encoded).

The sequence of coded bit groups mapped on each I/Q of the path needs to satisfy the characteristics of periodic repetition. For repeated coded bit groups, the corresponding uncoded bit groups need to be different. The order of uncoded bit groups has no special requirements.

This rule ensures that the TCM coded output signal points on the parallel path are uniformly symmetrical on the I and Q paths, and conforms to the principle of maximizing the Euclidean distance between the TCM coded output signal points on the parallel path.

The encoded bit group sequence mapped in the I path and the Q path needs to satisfy the Gray code mapping rule respectively, that is, only one bit between the two adjacent coded bit groups in the sequence is different.

The coding and modulation method of the trellis coded modulation code that satisfies the above mapping rule provided by the embodiment of the present invention includes: Establishing a QAM constellation diagram with a modulation order greater than 4 as a constellation map of the trellis coded modulation code, so that the coded bit groups mapped on the constellation map I and Q are periodically repeated on the I and Q paths, respectively. Arranging, and making uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle different from each other; obtaining a constellation index including the encoded bit group and the uncoded bit group according to the original bit information group; The signal point is determined in the constellation map as the finally transmitted trellis coded modulation code signal.

The decoding method of the trellis coded modulation code that satisfies the above mapping rule provided by the embodiment of the present invention includes:

Establishing a QAM constellation diagram with a modulation order greater than 4 as a constellation map of the trellis coded modulation code, so that the coded bit groups mapped on the constellation map I and Q are periodically repeated on the I and Q paths, respectively. Arranging, and making the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle different from each other; performing decoding processing on the signals to be decoded according to the constellation map.

The first embodiment provided by the present invention is a coding and modulation method of a trellis coded modulation code (TCM). The method flow is as shown in FIG. 11, and includes:

Step 101: Establish a constellation map of the 5/6 64QAM TCM as shown in FIG. 12A. The four bits of the encoded bit group in the constellation map are respectively mapped to the combination of b4b3 of the I path and blbO, b4b3 or blbO of the Q path. There are 2 ² = 4 in total, so the period of the bit sequence repetition after encoding on each road is 4. The two uncoded bits are mapped to b5 of the I path and b2 of the Q path, respectively. For repeated b4b3, the value of b5 needs to be different. For the same reason, for repeated blbO, the value of b2 needs to be different. The preferred mapping method therein is as follows.

I/Q path coordinate value 7A 5A 3A A -A -3A -5A -7A (A is the constellation diagram power normalization factor)

I road b4b3 10 00 01 11 10 00 01 11

I road b5 0 0 0 0 1 1 1 1 (may also be 0 1 1

0 1 0 0 1 , no special order required) Q road blbO 01 00 10 11 01 00 10 11

Q road b2 0 0 0 0 1 1 1 1 (can also be 00 111100, no special order required)

Step 102: Receive a raw bit information group "10011" (corresponding respectively

U4u3u2ulu0) Enter the 5/664QAM TCM encoder. The input order of "1001 Γ is the first "1" first arrival, the first "0" second arrival, the second "0" third arrival, the second "1" fourth arrival The third "1" finally arrived.

Step 103: The TCM encoder encodes the original bit information group "10011" to obtain a constellation index including the encoded bit group and the uncoded bit group. The TCM encoder section uses a convolutional code with a code rate of 3/4 (the original 1/2 convolutional code is punctured to make the code rate 3/4). For "01 卷 3/4 convolutional coding, if the initial value of the six registers of the convolutional encoder is 0, the output is "111 0" (corresponding to c3c2clc0 respectively), and the order is the first "1" first. Output, second "1" second, third "1" third, "0" last. Uncoded bit group "1" (u4) combined with encoded bit group "11" (c3c2) The 3-bit constellation index "ill" (b5b4b3) is mapped to the I path, where the last two "1"s (b4b3) are the encoded bit groups, and the first "1" (b5) is the uncoded bit group. "(u3clc0) is a set of 3-bit constellation index "010" (b2blb0) mapped to the Q path, where the last two bits "10" (blbO) are the encoded bit group, the first "0" (b2) ) is an uncoded bit group.

Step 104: According to the obtained I-way constellation index "111" and the Q-way constellation index "010", the constellation map of the 5/664QAM TCM is established in step 101 (as shown in FIG. 12A), and the I-way coordinate is determined to be -7. The Q path coordinates are 3 signal points and the signal point is used as the final transmitted TCM signal.

According to the technical principle described above, the constellation map of 5/664QAM TCM can also be obtained as shown in FIG. 12B and FIG. 12C, which indicates that the constellation map of 5/664QAMTCM is not unique, but satisfies the foregoing description. A collection of technical principles.

According to the technical principle described above, 3/4256QAM, 7/8 can also be obtained. The constellation map of 256QAM, and determine the final transmitted TCM signal, will not be described here. The constellation map in the preferred scheme of 256QAM is shown in FIG. 13A, and the constellation map of the same 256QAM is not the only one shown in FIG. 13A, and can also be as shown in FIG. 13B, which satisfies the technical principle described above. set.

The technical solution described in the embodiments of the present invention makes signal points in the same subset on the constellation map, especially on the Q road, and the minimum Euclidean distance of the signal points in the same subset is greater than 4 in the prior art. The minimum Euclidean distance of 1, the overall performance of the TCM code is improved.

A second embodiment of the present invention provides a trellis coded modulation code (TCM) decoding method, which is exemplified by a 5/6 64QAM TCM code, and includes:

Step 201: Establish a constellation map of the 5/6 64QAM TCM. As shown in FIG. 12A, the specific scheme is the same as step 101.

After the constellation map is established according to step 201, according to the constellation diagram, the decoding signal is subjected to subsequent decoding processing by using the prior art. As a preferred decoding processing scheme, subset decoding and soft decision Viterbi decoding may be employed.

Step 202: Sub-set decoding, in the four signal points of each subset of the 16 subsets of the constellation map of Fig. 12A, find the signal point with the smallest Euclidean distance from the received symbol. The schematic diagram of determining the signal point with the smallest Euclidean distance during decoding is shown in Fig. 14. The circle is the position of the received signal point, and the subset corresponding to the bit group "0000" (b4b3blb0) after encoding is the four signals in Fig. 14. Point, then the upper left signal point is the subset decoding result.

Step 203: Soft-decision Viterbi decoding, in the result of the subset decoding, obtains the Euclidean distance of the 16 encoded bit groups as its soft decision metric. By substituting the soft decision metric value into the Viterbi decoder for decoding operation, an estimated value (decoding result) for the original bit block can be obtained.

The technical principle of the decoding process of 2/3 64QAM, 3/4 256QAM, 7/8 256QAMTCM code is the same as that of the 5/6 64QAM TCM code described above, and will not be described here. The decoding process is performed according to the constellation diagram provided by the embodiment of the present invention, so that the signal points in the same subset are uniformly mapped on the constellation map, especially on the Q road, and the minimum Euclidean distance of the signal points in the same subset is 4 is larger than the minimum Euclidean distance 1 in the prior art, and the minimum Euclidean distance of the TCM coded output signal point on the parallel path is larger than the minimum Euclidean distance of the prior art, and the decoding error occurs under a certain signal to noise ratio. The probability is less than the probability of decoding errors in the prior art, and the overall performance of the TCM code is improved.

The TCM code described in the embodiment of the present invention can be applied to the TCM code under Wimax SCa as an improved solution. It is also applicable to other communication systems that use such TCM codes.

BER (bit error rate) simulation results of 2/3 64QAM, 5/6 64QAM, 3/4 256QAM, 7/8 256QAM TCM codes defined in the prior art and preferred TCM codes described in the embodiments of the present invention The comparison is shown in Fig. 15, Fig. 16, Fig. 17, and Fig. 18. Where Wimax represents the TCM code defined in the Wimax protocol, MY represents the preferred TCM code described in the embodiment of the present invention, the channel is the AWGN channel, and Es/Nt is the ratio of the symbol energy to the noise power in the unit bandwidth, that is, normalized. Signal to noise ratio.

It can be seen that compared to the 2/3 64QAM, 5/6 64QAM, 3/4 defined in the original Wimax protocol.

The performance of the preferred TCM code described in the embodiments of the present invention is significantly improved by the 256QAM, 7/8 256QAM TCM code. At BER = 10- ^6, there is a gain of 5dB ~ 9dB. It can be seen that, under the same signal-to-noise ratio, the technical solution described in the embodiments of the present invention improves the overall performance of the TCM code compared with the prior art.

Moreover, the TCM code designed by the embodiment of the present invention can maintain the definition of the original Wimax protocol.

The coding part of the TCM code has the same structure, and only needs to change the constellation map to obtain significant gain, which is compatible with the original protocol.

The third embodiment of the present invention is a code modulation device of a TCM, and its structure is as shown in FIG. 19, including:

The storage module 301 is configured to store a square QAM constellation diagram with a modulation order greater than 4 as

a constellation map of the TCM, mapping the encoded bits on the I and Q paths of the constellation map The groups are repeatedly arranged periodically on the I path and the Q path, and the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle are different from each other;

a constellation index obtaining module 302: configured to obtain, according to the original bit information group, a constellation index including the encoded bit group and the uncoded bit group;

The TCM signal confirmation module 303 is configured to determine a signal point in the constellation map according to the constellation index as a final transmitted TCM signal.

The fourth embodiment of the present invention provides a trellis coded modulation code (TCM) decoding device, and the structure thereof is as shown in FIG. 20, including:

The storage module 401 is configured to store a square QAM constellation diagram with a modulation order greater than 4 as a constellation map of the TCM, and the coded bit groups mapped on the constellation map I and Q paths are respectively on the I road and the Q road. The cycle is repeatedly arranged, and the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle are different from each other;

The decoding module 402 is configured to perform decoding processing on the signal to be decoded according to the constellation map. Specifically, in practical applications, the constellation diagram used in this embodiment can be combined with the present invention.

The spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and variations of the inventions

Claims

Rights request

A coding and modulation method for a trellis coded modulation code, comprising: establishing a quadrature amplitude modulation QAM constellation diagram with a modulation order greater than 4 as a constellation map of a trellis coded modulation code, so that the mapping is performed in the The coded bit groups on the I-channel and the Q-channel of the constellation map are repeatedly arranged periodically on the I-channel and the Q-channel, respectively, and the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle are different from each other;

A signal point is determined in the constellation map as the final transmitted trellis coded modulation code signal based on the constellation index.

2. The method according to claim 1, wherein only one bit is different between any two adjacent coded bit groups on the constellation map I or Q path.

3. The method of claim 2, wherein the constellation map comprises a 64QAM constellation map of a grid coded modulation code.

The method according to claim 3, wherein the original bit information block is processed by a 64QAM trellis coded modulation code encoder having a code rate of 2/3, and the encoded bit group and the uncoded are obtained. The constellation index of the bit group, or

The original bit information block is processed by a 64QAM trellis coded modulation code encoder with a code rate of 5/6 to obtain a constellation index including the encoded bit group and the uncoded bit group.

5. The method of claim 2, wherein the constellation map comprises a 256QAM constellation map of a grid coded modulation code.

The method according to claim 5, wherein the original bit block is processed by a 256QAM trellis coded modulation code encoder having a code rate of 3/4, and the bit group and the uncoded group are obtained. The constellation index of the bit group, or

The original bit information group is processed by a 256QAM trellis coded modulation code encoder having a code rate of 7/8 to obtain a constellation index including the coded bit group and the uncoded bit group.

7. A method for decoding a trellis coded modulation code, comprising: Establishing a quadrature amplitude modulation QAM constellation with a modulation order greater than 4 as a constellation map of the trellis coded modulation code, so that the coded bit groups mapped on the constellation map I and Q are in the I path and Q Repeatedly arranging the cycles on the way, respectively, and making the uncoded bit groups corresponding to the coded bit groups repeatedly arranged in each cycle different from each other;

Decoding the signal to be decoded according to the constellation map.

8. The method according to claim 7, wherein only one bit is different between any two adjacent coded bit groups on the constellation map I or Q path.

9. The method of claim 8, wherein the constellation map of the trellis coded modulation code comprises a 64QAM constellation map of trellis coded modulation codes.

10. The method according to claim 8, wherein the constellation map of the trellis coded modulation code comprises a 256QAM constellation map of trellis coded modulation codes.

A coded modulation device for a trellis coded modulation code, comprising: a storage module: configured to store a quadrature amplitude modulation QAM constellation diagram with a modulation order greater than 4 as a constellation map of a trellis coded modulation code, The coded bit groups mapped on the constellation map I and Q paths are repeatedly arranged periodically on the I path and the Q path, and the constellation index acquisition module is repeated with each cycle: used to obtain the coding according to the original bit information group. Constellation index of the latter bit group and the uncoded bit group;

A trellis coded modulation code signal confirmation module is configured to determine a signal point in the constellation map according to the constellation index as a trellis coded modulation code signal for final transmission.

12. A trellis coded modulation code decoding apparatus, comprising: a storage module: configured to store a quadrature amplitude modulation QAM constellation diagram with a modulation order greater than 4 as a constellation map of a trellis coded modulation code, The coded bit groups mapped on the I and Q paths of the constellation map are repeatedly arranged periodically on the I and Q paths, and the decoding module is repeated with each period: used to decode according to the constellation map The signal is decoded.