KR100339855B1 - A coded modulation system based on convolutional coding and constellation control and the coded modulation method - Google Patents

Abstract

The present invention relates to an encoding modulation system using constellation control according to a convolutional code and a code rate. In particular, the present invention uses a convolutional encoder in which only four types of code sequences are generated and codes the generated four code sequences as 4-PSK phase shift keying) signal constellation. Signs with close Hamming distances between the code sequences are located close to the signal constellation. Convolutional codes and constellation control are used to place the distances far from the signal constellation between codes having large Hamming distances. An encoding modulation system and a coding modulation method thereof.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coding modulation system using convolutional codes and constellation control, and a coded modulation system using the convolutional coding and constellation control,

The present invention relates to a coding modulation system that adjusts the position of a signal constellation according to a hamming distance between a convolutional code having a trellis structure and a code sequence so as to generate only four types of code sequences having a specific rule, And a coding modulation method therefor.

The channel coding is to enable detection and correction of errors by transmitting a certain amount of additional information to the transmitted information in order to reduce the error probability of the digital communication system. The channel code is largely divided into a block code and a convolutional code. The fundamental difference between a block code and a convolutional code is the presence or absence of a memory.

In the case of a block code, block information obtained by cutting the information series into blocks of a predetermined length by k bits is encoded into an n-bit codeword having the same length, whereas the convolutional code is based on a coding method in which a codeword composed of n bits is dependent on the current information composed of k bits Not only is it affected by past information. The length of a convolutional code that affects one information bit in a coded sequence is called a constraint length. The convolutional code can be characterized by the length of the constraint field and the coding rate. Since the convolutional code has a complicated structure compared with the block code, it is difficult to encode and decode it, and the correcting ability of the convolutional error is lower than that of the block code. However, when the code rate is the same, the ability to detect and correct irregular and independent errors at arbitrary bit positions is superior to the block code.

A communication system using a channel code can obtain a low bit error probability due to its powerful error correction capability. However, there is a problem that transmission bandwidth increases due to use of additional information due to channel coding. The additional information is a bit transmitted in addition to the information bits in order to achieve purposes such as detection and correction of errors occurring in a transmission channel in addition to information bits, and communication. There is a trellis coded modulation (TCM) scheme which combines channel coding and phase shift keying (hereinafter referred to as PSK modulation) to solve the problem of bandwidth increase due to the use of a channel coding scheme.

The basic concept of the TCM scheme is that one information bit is encoded into n coded bits and then n bits are transmitted as 2 ⁿ -PSK signals. Although n coded bits are required to transmit one information bit, since the transmission symbol rate is the same due to 2 ⁿ -PSK modulation, the transmission bandwidth is not increased and error correction according to coding improves the bit error rate (BER) performance . However, the TCM scheme exponentially increases the number of symbols required for configuration as the code rate increases. That is, 2 ^R signal gratings are required when the coding rate is 1 / R. As a result, the distance between adjacent symbols becomes closer and the probability of error occurrence between symbols increases, so that the performance improvement sharply decreases. Therefore, there is a disadvantage that power must be increased to obtain desired system performance. In addition, although the error correction capability can be increased by increasing the length of the confinement field, the structure of the transceiver is complicated

SUMMARY OF THE INVENTION Accordingly, the present invention has been made keeping in mind the above problems occurring in the prior art, and it is therefore an object of the present invention to provide a convolutional encoder capable of generating only four code sequences having a specific rule regardless of a coding rate, Modulated by the PSK signal and transmitted.

In addition, the constellation of the 4-PSK signal is adjusted according to the hamming distance between the code sequences, and the error probability between code sequences having a Hamming distance close to the signal constellation is made large according to the code rate and the required error probability characteristic. Another purpose is to control the signal constellation so that the error probability between distant signals is small so that the error correction capability is maximized.

Another object of the present invention is to maximize the code gain in decoding by optimizing the signal constellation so that the error probability is the same when the cumulative hamming distances are the same by adjusting the probability of error occurrence between code sequences.

According to an aspect of the present invention, there is provided a coding and modulation system for adjusting a signal constellation according to a Hamming distance between convolutional code sequences, thereby changing a probability of error occurrence between signals.

The feature of the convolutional encoder of the coding and modulation system is that there are always four convolutional codes generated irrespective of the code rate and the constraint field size required by the system.

1 is a block diagram showing an encoding modulation system using convolutional codes and constellation control according to the present invention;

2 is a signal constellation diagram according to the present invention.

3 is a diagram illustrating a structure of an encoder trellis according to the present invention.

Description of the Related Art

10: convolutional encoder 11: shift register (memory)

12: adder 13: combine vector machine

14: memory device 15: switch

20: Signaling upper part 30: Demodulation and judgment part

40: Viterbi decoding unit

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The present invention relates to an encoding modulation system using a convolutional code and an encoding modulation method using the modulation system, the encoder and the encoder having a trellis structure in which the number of code sequences outputted from the convolutional encoder is always four and the minimum free distance is maximum, And a distance between the 4-PSK signals to be formed according to the hamming distance between the code sequences is adjusted and transmitted.

FIG. 1 is a block diagram of a coding and modulation system using convolutional codes and constellation control according to the present invention. Referring to FIG. 1, A convolutional encoder (10) having a trellis structure for always having a maximum of four free paths and having a minimum free distance; a convolutional encoder (10) for converting the signal constellation And a Viterbi decoding unit 40 according to the same trellis structure as the demodulating and judging unit 30 and the transmitting unit in the receiving unit.

The transmitter generates a codeword using the convolutional encoder 10 and then controls the constellation of the 4-PSK signals according to the required code rate and SNR.

The demodulation and determination section 30 converts the signal transmitted from the transmission section back into a codeword suitable for the required code rate. The converted codeword is decoded to information bits via the Viterbi decoding unit 40, and the trellis structure of the Viterbi decoding unit 40 has the same trellis structure as the transmitting unit.

The convolutional encoder 10 of FIG. 1 shows that the length of the constraint field is 4 and the code rate is 1/4, and the code sequence output using the applied equations will be described. A general convolutional encoder consists of N storage devices, an adder, and a switch that outputs a codeword. The convolutional coder uses N storage devices, which give correlation between output sequences to make the current information series affected by past information series. The combinatorial vector indicates whether the storage device and the adder are connected. The adder reflects the influence of the memory output on the correlated output series through the coupling vector. The switch determines the output code sequence of the convolutional encoder.

First, Equation (1) defines a connection vector when a code rate is 1 / R to indicate convolutional coding.

Where g denotes the combined vector and R denotes the average length of the output code sequence according to one input bit.

Equation (2) represents the joint vectors of the encoder having a constraint length of 4 when R = 2, that is, when the code rate is 1/2.

The shape of the joint vector can be adjusted according to the length and coding rate of the constraint field to be used in the system. The minimum free distance of the code to be output depends on the form of the joint vector. Depending on the length of the confinement field, the shape of the coupling vector may vary. However, in the encoder having an arbitrary constraint length, the structure of the joint vector that outputs four code sequences at all times irrespective of the code rate and the maximum free distance of the output code sequence is maximized is j = R in Equation 1 So that the shape of the other joint vector is the same.

An example of the case where R = 4 is shown in Equation (3).

When a convolutional code is generated using the combination vector of Equation (3) in the convolutional encoder of FIG. 1, the code rate is 1/4, the trellis structure is as shown in FIG. 3, and the structure of the output code is as shown in Equation 4.

{0000}, {0001}, {1110}, {1111}

When the code rate is 1/4, the four codes to be output are a code sequence composed of only four consecutive '0' s and a code sequence composed of only three consecutive '0s' A code sequence composed of three consecutive '1' s followed by one '0', and a code sequence composed of four consecutive '1' s. Equation (5) shows a structure of four code sequences to be outputted when the code rate is 1 / R, and a code sequence consisting of R consecutive '0' and a code sequence consisting of R-1 consecutive '0' 1, a code sequence consisting of R-1 consecutive 1's and one consecutive 0's, and a code sequence consisting of R consecutive 1's.

2, the Hamming distance between s ₀ and s ₁ , s ₂ and s ₃ signals is always '1', and s ₀ , s ₂ And s ₁ and s ₃ have hamming distances R-1, s ₀ , s ₃ and s ₂ , s ₁ , respectively. If, when arranged in a conventional QPSK signal constellation constant distance between the s ₁ from the code sequence s ₂ or s ₀ at s _0, and thus independent of the Hamming distance between the code sequence error probability is constant.

In the present invention, the signal constellation method depends on the SNR and the code rate. In Fig. 2, an error occurrence to s ₀ and s ₁ is equal to a Hamming weight difference of '1' when decoded,

The error probability from s ₀ to s _{1 is found} as in Equation (6).

Where d ₁ represents the distance between s ₀ or s ₂ and s ₃ .

from s ₀ to s ₃ has an error difference between the Hamming weight in R s ₀ s ₃ to the error probability is the same as equation (7).

Where d ₃ represents the distance from s ₀ to s ₂ or s ₁ to s ₃ .

The constellation for obtaining the optimal code gain is obtained by Equations (8) and (9). Equation 8 is obtained by adjusting the signal position on the lattice diagram so that the probability of occurrence of one error between s ₀ and s ₃ signals is equal to the probability of occurrence of successive R errors between s ₀ and s ₁ signals, Thus making the probability of error equal. In Equation (8), d ₁ to d ₃ represent the distances between the signals used in the transmission in FIG. 1, E _S in Equation (9) is the energy of the transmitted 4-PSK signal and N ₀ is the noise power spectral density power spectral density.

As R increases, the distance between s ₀ and s ₁ decreases according to equation (8), so that the probability of error between s ₀ and s ₁ becomes higher. However, when R is large, the minimum free distance of the convolutional code becomes large. Since the minimum free distance d _free has the relationship of the number of correctable code bits t and Equation 10, R increases and d _free increases.

When the length of the constraint field is 3, the trellis structure of the encoder has four states, and it goes through three branches to start from an arbitrary state and return to its original state. When the coding rates are 1/2, 1/4, 1/8, 1/16, ..., 1 / R, the minimum free distance of the convolutional codes used in the present invention is expressed by Equation (11).

The convolutional code used in the present invention is constructed such that the Hamming distance of divided branches is R and the hamming distance of the gathering branches is R as a trellis structure and the minimum free distance is maximized so that when the length of the constraint field is 3, If the rate is 1 / R, the minimum hamming distance can be obtained by Equation (12).

As shown in FIG. 1, if the length of the constraint field is 4, the minimum free distance for each code rate must be increased to at least four branches in order to start from an arbitrary state and return to the original state as shown in FIG. 3 do. When the coding rate is 1/2, 1/4, 1/8, 1/16, ..., 1 / R, the minimum free distance is expressed by Equation (13).

The minimum free distance increases according to the length of the confinement field. The codeword used in the present invention is designed such that the Hamming distance of the branching branch and the branching branch is R. However, when the state is increased, R and R-1 are not allotted together. Therefore, the minimum free distance of the convolutional encoder used in the present invention is not expressed by a general expression as shown in Equation (12).

When encoding modulation method according to the invention compared to conventional QPSK constellation scheme, s _0, s ₁ or s _2, s _3, while the s _0, s ₂ or s _1, it becomes higher the probability of error, between the signal s ₃ signal The probability of occurrence of an error between the two becomes low. However, since the cumulative hamming distance is increased by 1 in the case of errors between s ₀ , s ₁ or s ₂ and s ₃ , the sum of Hamming distances accumulated in t symbol errors is t, _It is possible to correct the symbol error between ₀ ↔ s ₁ , s ₂ ↔ s ₃ . On the other hand, since the cumulative hamming distance is increased by R-1 between s ₀ , s _2, or s ₁ and s ₃ signals, the error occurrence probability is lowered to maximize the error correction capability as a whole.

Since the distance between transmission signals calculated from Equations (8) and (9) is determined according to the noise power spectral density and the coding rate, the constellation method used in the present invention can be adjusted according to the SNR and the coding rate required by the system.

As described above, in the coding modulation method for adjusting and transmitting the position of the 4-PSK signal constellation according to the Hamming distance between the convolutional code sequences according to the present invention, only four codes are always generated irrespective of the coding rate and 4-PSK modulation The M-PSK modulation method in which the signal constellation increases exponentially due to an increase in the coding rate is not necessary because the bandwidth does not increase as in the TCM method.

Also, since the coding modulation method according to the present invention uses a signal constellation method that can optimize the SNR and the coding rate required by the system, if the information of the channel is known to some extent, the coding gain according to coding can be maximized have.

Also, the coded modulation scheme can be used in high speed broadband data communication because it can be used in all communication systems and methods requiring error correction.

Claims (16)

A convolutional encoder having a trellis structure in which the number of code sequences to be output is always 4 and the minimum free distance is maximized irrespective of the code rate and the constraint field used from the input information bits; A signal-receiving portion for adjusting the signal constellation according to the hamming distance between the code sequences, The receiver includes a demodulation and decision section for converting the constellation-adjusted 4-PSK transmission signal into a code sequence corresponding to a code rate to be used, And a Viterbi decoding unit according to the same trellis structure as that of the transmission unit. The method according to claim 1, The convolutional encoder has four code symbols output when the code rate is 1 / R, a code sequence composed of R consecutive '0's, and a code sequence composed of R-1 consecutive' 0's, A code sequence composed of R-1 consecutive '1's and a succeeding' 0 ', and a code sequence composed of R consecutive' 1's are generated. Controlled Coded Modulation System. 3. The method according to claim 1 or 2, Wherein the four codewords are converted into a 4-PSK signal by a signal upper part. The method according to claim 1, The signal constellation part positions the codes having close Hamming distances between the code sequences output from the convolutional encoder to a distance in the 4-PSK signal constellation and sets the distance from the 4-PSK signal constellation diagram Wherein the convolutional code and the constellation control are used. 5. The method of claim 4, Wherein the signal constellation unit adjusts the distance between the constellation-shaped signals so that the error probability is the same if the cumulative Hamming distance is the same, according to the SNR and the coding rate required by the system. 6. The method of claim 5, The signal constellation section may be configured such that the constellation of the signal (Where d ₁ is the distance between s ₀ and s ₁ or s ₂ and s ₃ signals, d ₃ is the distance between s ₀ and s ₃ or s ₁ and s ₂ signals, and No is the noise power spectrum And R is an average length of an output code sequence according to one input bit). The encoding / modulation system using the convolutional code and constellation control. The method according to claim 1, Wherein the convolutional encoder has a trellis structure in which the output codes of four codes are path weights and the hamming distance of the divided branches is R and the summing distance of the gathering branches is R and the minimum free distance is maximized An Encoding Modulation System Using Convolutional Code and Constellation Control. The method according to claim 1, In the demodulation unit, the demodulated four 4-PSK signals are divided into a code sequence composed of R consecutive 0s and a code sequence composed of Rs consecutive 0s, , A code sequence consisting of R-1 consecutive '1's, followed by a' 0 'code sequence, and R consecutive' 1 'code sequences. An Encoding Modulation System Using Constellation Control. A step of causing the convolutional encoder to have a trellis structure in which the number of code sequences to be output is always 4 and the minimum free distance is maximized irrespective of the code rate and the constraint field used from the input information bits; Adjusting a 4-PSK signal constellation according to a Hamming distance between four code sequences output to have the trellis structure; A demodulating step of demodulating the code sequence whose signal constellation is adjusted to a code sequence corresponding to a code rate used; And performing Viterbi decoding on the demodulated code sequence according to the same trellis structure as that of the transmission unit. 10. The method of claim 9, When the code rate is 1 / R, the four code sequences have a code sequence composed of R consecutive '0's, a code sequence composed of R-1 consecutive' 0's and one '1' And a code sequence composed of R-1 consecutive '1' and one consecutive '0', and R consecutive '1' code sequences. 11. The method according to claim 9 or 10, Wherein the four code sequences are configured as a 4-PSK signal by a signal constellation step. 10. The method of claim 9, In the 4-PSK signal constellation diagram, the codes having close Hamming distances are positioned close to each other in the 4-PSK signal constellation diagram by the signal constellation step, And the code is modulated by using the convolutional code and the constellation control. 13. The method of claim 12, Wherein the distance between signals on the constellation is adjusted so that error probabilities are equal if the cumulative Hamming distance is equal to the SNR and coding rate required by the system. 14. The method of claim 13, If the cumulative hamming distances on the constellation are the same, (Where d ₁ is the distance between s ₀ and s ₁ or s ₂ and s ₃ signals, d ₃ is the distance between s ₀ and s ₃ or s ₁ and s ₂ signals and No is the noise power spectrum And R is an average length of an output code sequence according to one input bit). The encoding modulation method using convolutional code and constellation control. 10. The method of claim 9, The four codes output by the code sequence output stage are path weighting. The branching branches have a trellis structure in which the hamming distance is R and the gathering branches have Hamming distance R, and the minimum free distance is maximized And an encoding control method using constellation control. 10. The method of claim 9, The demodulated four 4-PSK signals are composed of a code sequence consisting of R consecutive '0' and R '1 consecutive' 0 ', each of which has a length of R according to a coding rate A code sequence consisting of R-1 consecutive '1's, followed by a code sequence composed of' 0's, and R consecutive '1's. Used coding modulation method.