KR101284983B1 - Demapping method in wireless communication system using modulo operation - Google Patents

Abstract

The received signal obtained by performing a modulo operation on the input signal generates a maximum function value most likely to correspond to the candidate constellation point on the extended constellation, and uses the maximum function value to generate a log-likelihood ratio. ; LLR). In a wireless communication system performing a modulo operation, errors occurring during demapping can be reduced.

Demapping, Modulo Operations, DPC, Interference Preduction, Log Likelihood Ratio (LLR)

Description

Demapping method in wireless communication system using modulo operation

1 is a block diagram illustrating a transmitter and a receiver according to an embodiment of the present invention.

2 is a flowchart illustrating a demapping method according to an embodiment of the present invention.

3 is a diagram illustrating an example of generating an extended constellation according to an embodiment of the present invention.

4 illustrates an example in which candidate constellation points are arranged in the extension generation diagram generated in FIG. 3.

5 is an example of an extended constellation diagram in QPSK.

FIG. 6 is a diagram illustrating a received signal in which a modulo operation is performed on a symbol mapped to QPSK when noise is a Gaussian distribution.

FIG. 7 is a diagram illustrating coordinates on a constellation with respect to a received signal of a QPSK mapped symbol.

DESCRIPTION OF REFERENCE NUMERALS OF THE MAIN PARTS OF THE DRAWINGS

100: transmitter

200: receiver

The present invention relates to wireless communication, and more particularly, to a demapping method in a wireless communication system using modulo operation.

The demand for communication services such as the universalization of information communication services, the appearance of various multimedia services, and the emergence of high quality services are rapidly increasing. Various wireless communication technologies are being investigated in various fields to satisfy this demand.

Multiple input multiple output (MIMO) technology is attracting attention as a technology for enhancing link reliability and interference prevention with high frequency efficiency. In the MIMO technique, the signals transmitted through the multiple transmission antennas have independency, thereby increasing the diversity gain. MIMO technology can be divided into point-to-point communication such as single-user MIMO and point-to-mulitpoint communication such as multi-user MIMO.

In the one-to-many communication, an interference presubtraction scheme in a transmitter has recently been attracting attention. Signal A is called a signal to send to user A, and signal B is called a signal to send to user B. Signal A is first processed from the proper association with signal B to produce a noise-like signal (A '), which is added to signal B and sent to the channel. Receiving this signal, user B considers and decodes both the noise from the channel and the processed signal A 'in the original signal B as noise. User A can completely recover signal A from the processed noise A '. This eliminates the need for separate interference cancellation at the receiver, which can greatly reduce the complexity of the receiver. One such scheme is referred to as dirty paper coding (DPC).

The DPC may be referred to as an interference signal canceling technique in a transmitting end that prevents the receiving end from being affected by the interference signal when the transmitting end knows the interference signal in advance in the presence of an interference signal in addition to the noise signal in the channel. DPC is described in M. Costa, "Writing on Dirty Paper", IEEE Trans. on Inf. Theory, vol. IT-239, No. 3, May 1983. In this document, Costa shows that under DPC, the channel capacity is the same as in the absence of interference.

However, methods and apparatuses for effectively implementing DPC in one-to-many communication are not disclosed. In addition, although the MIMO scheme can achieve a high data rate, it is difficult to find an example using the DPC.

In addition, when a user signal is transmitted using a DPC, a receiver may perform a modulo operation. By performing the modulo operation, an error of the user signal may occur. In other words, an error may occur when the demapping becomes a symbol different from the constellation symbol transmitted by the transmitter. Therefore, it is necessary to reduce an error in demapping at the receiver for transmitted symbols transmitted using DPC.

SUMMARY OF THE INVENTION The present invention has been made in an effort to provide a demapping method and a receiver for reducing errors due to modulo operations.

The demapping method according to an aspect of the present invention generates a maximum function value that is most likely that a received signal obtained by performing a modulo operation on an input signal corresponds to a candidate constellation point on an extended constellation, and the maximum function value To generate a log-likelihood ratio (LLR).

According to another aspect of the present invention, a modulo operation is performed on an input signal to generate a received signal, and the received signal uses the plurality of constellation points representing the same bit string to obtain candidate function values for the constellation points. And a log-likelihood ratio (LLR) using the candidate function values.

According to another aspect of the present invention, a receiver performs a modulo operation on an input signal to generate a received signal, and a candidate function value indicating a possibility that the received signal corresponds to a candidate constellation point on an extended constellation. A demapper for selecting one function value to generate a log-likelihood ratio (LLR) and a decoder for decoding the symbols generated by the LLR.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Like reference numerals designate like elements throughout the specification.

The following techniques can be used in various communication systems. Communication systems are widely deployed to provide various communication services such as voice, packet data, and the like. This technique can be used for a downlink or an uplink. The downlink means communication from a base station (BS) to a user equipment (UE), and the uplink means communication from a terminal to a base station. A base station generally refers to a fixed station communicating with a terminal, and may be referred to in other terms such as a node-B, a base transceiver system (BTS), and an access point. A terminal may be fixed or mobile and may be referred to by other terms such as a mobile station (MS), a user terminal (UT), a subscriber station (SS), a wireless device,

1 is a block diagram illustrating a transmitter and a receiver according to an embodiment of the present invention.

Referring to FIG. 1, the transmitter 100 may include an encoder 110, a mapper 120, and a DPC device 130. The encoder 110 encodes the input information bit through the channel code. The channel code may include the function of correcting errors by the channel. As the encoding method by the encoder 110, various methods such as encoding in a block form and encoding in a trellis form may be used. Trellis-type encoding may include turbo coding or convolutional coding.

The mapper 120 maps the encoded data into a data symbol w representing a position on the signal constellation. The data symbol w is a symbol encoded and constellation mapped to data to be transmitted to each user. If there is no restriction on the modulation scheme performed by the mapper 120, it may be m-Phase Shift Keying (m-PSK) or m-Quadrature Amplitude Modulation (m-QAM). For example, m-PSK may be Binary Phase Shift Keying (BPSK), Quadrature Phase Shift Keying (QPSK), or 8-PSK. The m-QAM may be 16-QAM, 64-QAM, or 256-QAM.

The DPC device 150 generates a transmission signal M by performing DPC on the data symbol w input from the mapper 120. The DPC machine 150 may include an interference deductor 130 and a modulo operator 140. The interference deductor 130 subtracts the interference S known to the transmitter 100 for the mapped data symbol w. When the data symbol sent to the user is called w and the interference previously known by the transmitter 100 is S, the interference deductor 130 performs w-S to subtract the interference S.

For any interference S, it is necessary to change the shape of the transmitted signal set in order for the DPC to operate. This is because if the interference S is large, it may exceed the transmit power constraint. In consideration of this, a modulo operation for changing the shape of the transmission signal set is performed. The modulo operator 140 performs a modulo operation on the output w-S of the interference deductor and outputs the transmission signal M. FIG. For example, when a QPSK symbol has a form of ± 1 ± j in a constellation using QPSK, modulo operation allows -1 + j to be -5-3j, -5 + j, + 3 + j, +3+. 5j and the like operate as if they are the same symbol to be prepared for any interference S.

The transmission signal M output from the modulo operator 140 is represented by the following equation.

Where a and b are integers and L is half the size of the modulo box. The transmitter 100 deducts and transmits an interference signal in advance. In this case, since transmission power may increase due to interference, transmission power may be limited by modulo operation.

The receiver 200 may include a modulo operator 230, a demapper 220, and a decoder 210.

The modulo operator 230 performs a modulo operation on the input signal R to output the received signal Y. The input signal R is the sum of the noise N and the interference S as the transmitted signal M passes through the channel, resulting in R = M + S + N. Therefore, the input signal R is as follows.

When the modulo operation is performed on the input signal R, the following equation (3) can be obtained.

Where c and d are integers and L is half the size of the modulo box. The modulo operator of the receiver moves the received signal Y into a rectangular modulo box having a length of 2L each.

The demapper 220 demaps the received signal Y. The demapper 220 may convert the symbol level information into the bit level by estimating the received signal as one point of coordinates in the constellation. The decoder 210 decodes the symbol demapped by the demapper 220 and restores the original information bits.

2 shows a demapping method according to an embodiment of the present invention.

Referring to FIG. 2, a modulo operation is performed on the received signal (S100). Modulo operation is performed on the input signal R to the receiver to output the received signal Y.

A candidate function value is generated for the received signal Y, which is a signal on which the modulo operation is performed, using the candidate constellation point on the extended constellation, and a maximum function value is selected from the generated candidate function values (S110). The candidate function value refers to a value indicating the likelihood that the received signal Y corresponds to the candidate constellation point on the extended constellation. An extended constellation is a constellation that includes an extended constellation as well as basic constellations on the basic constellation. The maximum value of the candidate function value is called a maximum function value, and the maximum function value is selected from the candidate function value.

3 shows an example of generating an extended constellation according to an embodiment of the present invention. Referring to FIG. 3, an example using a basic constellation of 16-QAM will be described. A constellation composed of a plurality of constellation points corresponding to symbols in the modulo box 300 is called a basic constellation, and the basic constellation in 16-QAM is 16 constellation points in the modulo box 300. It becomes the constellation which consists of.

The extended constellation includes the basic constellations and the extended constellations of the basic constellation. Extended constellation points refer to constellation points other than the basic constellation points. The modulo box 300 refers to an edge indicating an area where basic constellation points are located.

An example of generating an extended constellation point is to repeat the modulo box and place it adjacent to the modulo box 300 of the basic constellation. Eight modulo boxes can be arranged adjacent to one basic constellation, and the arrangement of constellation points included in the eight modulo boxes is the same as the arrangement on the basic constellation. Repeating the modulo box here means copying the constellation points inside the modulo box.

Herein, the constellation points included in the eight modulo boxes may be selected to be extended constellation points. For example, by repeatedly arranging four modulo boxes up, down, left, and right, a total of 16 extended constellation points can be obtained by selecting four constellation points close to the origin of constellation among constellation points located in each modulo box. have.

As another example, one of the basic constellation points, which is close to the constellation points near the edge, may be selected as the extended constellation point. Therefore, it is possible to select an extended constellation point located close to the modulo box of the basic constellation.

As another example, the basic constellation points may be repeatedly arranged vertically or horizontally. Repeated placement means that the basic constellation points are copied up, down, left, or right, with the same spacing and with the same spacing. Therefore, one or more extended constellation points can be selected from these copied basic constellation points.

As described above, by generating one or more extended constellation points, the extended constellation can be obtained by placing the constellation together with the basic constellation point. In generating extended constellations, various m-PSK or m-QAM basic constellations such as BPSK, QPSK, 8-QAM, 16-QAM, 64-QAM, 128-QAM, and 256-QAM can be used. In addition, in the basic constellation, the position of the constellation point on the constellation corresponding to the bit string may vary.

4 illustrates an example in which candidate constellation points are arranged in the extension generation diagram generated in FIG. 3. Referring to FIG. 4, the extended generation diagram includes basic constellation points on the basic generation diagram and extended constellation points located close to the modulo box 300. Eight modulo boxes are arranged adjacent to one basic constellation, and the arrangement of constellation points included in the eight modulo boxes is the same as the arrangement of constellation points on the basic constellation.

One or more constellation points that represent the same bit string are called candidate constellation points, and the candidate constellation points may include one or more constellation points on the extended constellation diagram.

Side constellations located on the side of the base constellation rather than on the corners may generate one extended constellation point. For example, the basic constellation point representing the ₀ bit string located on the right side may be generated by arranging an extended constellation point representing the ₁ bit string located on the left side. This is because even if the modulo box 300 is repeatedly arranged, the arrangement of constellation points in the modulo box 300 is the same as the arrangement on the basic constellation. Accordingly, the candidate constellation points representing the bit streams are the basic constellation points representing the _zero bit streams and the extended constellation points representing the _one bit stream.

The corner constellation point located at the corner of the basic constellation point may generate three extended constellation points. For example, a default constellation point representing a _zero bit string located in the upper right corner is a _one bit string located in the upper left, a _two bit string located in the lower right and a lower left corner. Each extended constellation point representing ₃ bit strings may be generated. Thus, the candidate constellation points representing the bit streams have a base constellation point representing the _zero bit stream and three scalability representing the _one bit string, the _two bit strings, and the _three bit strings. It is a shop.

Among the basic constellation points, the constellation points other than the corner constellation point or the peripheral constellation point become inner constellation points and do not generate extended constellation points. For example, a basic constellation point representing a (0000) ₀ bit string located near the origin on the constellation does not produce an extended constellation point. This is because the arrangement of constellation points in the modulo box 300 is the same as the arrangement on the basic constellation while repeatedly arranging the modulo boxes. Therefore, the candidate constellation point representing the (0000) bit string is one basic constellation point representing the (0000) ₀ bit string.

As such, the candidate constellation points are one or more constellation points representing the same bit string, and may include basic constellation points and extended constellation points. However, when there is no extended constellation point representing a specific bit string, only the basic constellation point may be a candidate constellation point.

5 is an example of an extended constellation diagram in QPSK. Referring to FIG. 5, in a typical QPSK constellation diagram, the length and width of the module are mapped to one of four points in the modulo box 300 having a length of 2L. The basic constellation in QPSK is a constellation of four basic constellation points in the modulo box 300.

The symbols mapped on the basic constellation map are mapped to the real part I and the imaginary part Q by 1 bit, respectively, and correspond to a total of 2 bits of information. For example, consider the case where the real axis affects the preceding bit and the imaginary axis affects the later bit. In other words, if the preceding bit is zero, the real value of the signal has A, and if it has 1, it has -A. On the other hand, if the trailing bit is zero, the imaginary value of the signal is A, and if it is 1, it has -A. The basic constellation point located on the basic constellation map corresponds to a specific bit string, and the same constellation point may correspond to another basic constellation point on the basic constellation in some cases.

The extended constellation includes the basic constellations and the extended constellations of the basic constellation. The extended constellation of FIG. 5 is generated by repeating the modulo box of the basic constellation and arranging adjacent to the modulo box of the basic constellation, and then selecting the extended constellation points disposed close to the constellation points of the basic constellation. .

The extended constellation has candidate constellation points, which are one or more constellation points corresponding to the same bit string. In the basic constellation diagram, one constellation point corresponding to the bit string to be demapped in the modulo box 300 is located. In the extended constellation diagram, one or more constellation points corresponding to the bit string to be demapped are located. For example, if the bit string to be demapped is (00), then the candidate constellation points in the extended constellation are four (00) ₀ , (00) ₁ , (00) ₂ and (00) ₃ . If the bit string to be demapped is (11), the candidate constellation points in the extended constellation are (11) ₀ , (11) ₁ , (11) ₂ and (11) ₃ .

The candidate function value (P _i) by using the candidate sex shops on such an extension also aqueous phase can be calculated by the following equation. Candidate function value (P _i) is the received signal the operation is performed in the module represents the potential corresponding to a particular bit string numerically.

Where σ ² is the noise variance, y is the received signal on which the modulo operation is performed, and h is the channel response. Candidate function value in equation 4 (P _i) can be calculated by substituting the coordinate (α) of the candidate _i which satisfies the property store bits of α. For example, if the bit string α to be demapped is (00), the candidate constellation points of (α) _i are four constellation points of (00) ₀ , (00) ₁ , (00) ₂ and (00) ₃ , (00) ₀ is the basic constellation point and (00) ₁ , (00) ₂ and (00) ₃ are three extended constellation points. So these by the four sex shops by the equation (4) can generate the four candidate function value (P _i).

를 만족하는 경우에 후보 함수 값(P_i)는 최대가 될 수 있다. 후보 함수 값(P_i)이 최대가 된다는 것은 전송기에서 전송된 성상 심벌이 해당 후보 성상점에 대응되는 특정 비트열일 가능성이 가장 크다는 것을 의미한다. 따라서, 확장 성상도에서의 하나 이상의 후보 성상점들 중에서 식

를 가장 근사하게 만족시키는 후보 성상점에 의한 후보 함수 값(P_i)을 구하면, 그 때의 성상점이 특정 비트열일 가능성이 가장 크게 된다.In other words, if the bit string α to be demapped is (00), the candidate constellation points corresponding thereto may be four constellation points (00) ₀ , (00) ₁ , (00) _2, and (00) ₃ . The candidate function value P _i is obtained by substituting the values of (α) _i , which are coordinates on the constellations for the candidate constellation points, respectively. here,

Candidate function value (P _i) in the case of satisfying may be a maximum. The candidate function value (P _i) is that the maximum means that the constellation symbols transmitted from the transmitter the greater the specific bit ten days potential corresponding to the candidate sex shops. Thus, one of the one or more candidate constellation points in the extended constellation

To ask the candidate value of the function (P _i) by the candidate property store which is most approximate satisfied, the constellation points are the specific bit ten days likely the greatest at that time.

The maximum function value P _max , which is the maximum among the generated candidate function values P _i , may be obtained by the following equation.

Referring back to FIG. 2, a log-likelihood ratio (hereinafter referred to as LLR) is calculated using the maximum function value P _max (S120). The output from the demapper 220 may be expressed as a reliability or probability value for 1 and 0 in the input bit string rather than a binary number. Therefore, the following equation can be used to calculate the LLR representing the reliability or probability of 1 or 0 for the k-th bit as a ratio.

는 α 의 실수 값에 맵핑되는 k-번째 비트가 가리키는 것이 0 인 심벌이고,

는 α의 실수 값에 맵핑되는 k-번째 비트가 가리키는 것이 1 인 심벌이다. 수학식 6에서의 분모는 성상 심벌의 좌표 중 하나인 α가 맵핑되는 k-번째 비트가 0 이 되는 최대 함수 값들의 합이며, 분자는 성상 심벌의 좌표 중 하나인 α가 맵핑되는 k-번째 비트가 1 이 되는 최대 함수 값들의 합이다. 따라서 분모가 커지면 LLR은 낮아지고, 분자가 커지면 LLR은 높아진다. LLR이 낮다는 의미는 성상 심벌의 좌표 중 하나인 α가 맵핑되는 k-번째 비트가 0 이 될 확률이 높다는 의미이고, LLR이 높다는 의미는 성상 심벌의 좌표 중 하나인 α가 맵핑되는 k-번째 비트가 1 이 될 확률이 높다는 의미이다.Where α is one of the coordinates of the constellation symbol on the constellation map,

Is a symbol whose 0 is the k-th bit mapped to the real value of α,

Is a symbol whose k-th bit is mapped to the real value of α. The denominator in Equation 6 is the sum of the maximum value of the function whose k-th bit to which α, one of the coordinates of the constellation symbol is mapped, becomes 0, and the numerator is the k-th bit to which α, which is one of the coordinates of the constellation symbol, is mapped. Is the sum of the maximum function values, where 1 is 1. Therefore, the larger the denominator, the lower the LLR, and the larger the numerator, the higher the LLR. Low LLR means that the k-th bit to which α, one of the constellation symbol's coordinates is mapped, has a high probability of being zero. High LLR means the k-th to which α, one of the constellation symbol's coordinates is mapped. This means that the bit is likely to be one.

The maximum function value P _max in Equation 6 may be calculated using Equation 5. A maximum function value P _max that satisfies Equation 5 may be calculated using candidate constellation points on the extended constellation corresponding to a specific bit string.

For example, in order to calculate the LLR of the first bit when demapping a QPSK mapped symbol, the following equation may be used.

In calculating the LLR of the first bit, the sum of the maximum function values (P _max ) for (00) and (01), which is the bit string where the first bit is zero in the denominator, and the first bit in the numerator is 1. Calculate by taking the logarithm of the sum of the maximum function values (P _max ) for the bit strings (10) and (11).

In other words, by substituting the maximum function value P _max of each symbol into Equation 7, the LLR value for the first bit can be calculated. LLR goes low if the first bit is likely to be zero, and LLR goes high if the first bit is likely to be 1.

To calculate the LLR of the second bit can be calculated using the following equation.

In calculating the LLR for the second bit, the sum of the maximum function values (P _max ) for (00) and (10), where the first bit is 1 in the denominator, and the first bit in the numerator. Calculate by taking the logarithm of the sum of the maximum function value (P _max ) for the bit strings (01) and (11). If the second bit is likely to be zero, the LLR of Equation 8 is low, and if the second bit is likely to be one, the LLR of Equation 8 is high.

Therefore, by calculating the LLR (b _R , ₁ ) and LLR (b _R , ₂ ) it is possible to calculate the ratio of the reliability or probability to be 0 or 1 for the first bit and the second bit. Therefore, the demapper may demap the received signal on the constellation by LLR (b _R , ₁ ) and LLR (b _R , ₂ ).

As described above, the demapping method according to an embodiment of the present invention obtains the maximum function value P _max using candidate constellation points of the extended constellations, and demaps the LLR value based on this. By introducing the expanded constellation can more accurately calculate a candidate value of the function (P _i) is calculated by the equation (4). In addition, an error caused by moving a constellation symbol into a modulo box by a modulo operation may be reduced by obtaining a maximum function value P _max by applying one or more candidate constellation points. Therefore, in the LLR calculation using the maximum function value P _max , a relatively accurate LLR value can be calculated.

Therefore, the receiver can demap the constellation symbols transmitted from the transmitter by extracting the position of the received signal on the constellation derived by modulo operation relatively accurately. Therefore, the performance deterioration can be reduced and the performance can be improved in the demapper.

Table 1 shows Signal to Noise Ratio (SNR) according to MCS level in order to satisfy the 10% Frame Error Rate (FER). Here, the conventional technique is an SNR value (dB) required for demapping by a general demapping method by receiving a signal on which the DPC is performed, and the proposed technique is a demapping method according to an embodiment of the present invention. It represents the SNR value (dB) required at the time of demapping.

MCS Level Conventional Technique (dB) Proposed Technique (dB) Difference (dB) MCS 0 (QPSK, 1/8 Coding) 3.4603 2.3091 1.1512 MCS 1 (QPSK, 1/6 Coding) 4.0801 2.6205 1.4596 MCS 2 (QPSK, 1/4 Coding) 5.1633 3.4794 1.6839 MCS 3 (QPSK, 1/3 Coding) 4.4745 4.0797 1.6442 MCS 4 (QPSK, 3/7 Coding) 6.6984 4.6930 2.0055 MCS 5 (QPSK, 1/2 Coding) 6.3100 5.3589 2.2005 MCS 6 (QPSK, 5/9 Coding) 7.8616 5.5531 2.3085 MCS 7 (QPSK, 5/8 Coding) 6.9962 6.3201 1.9255 MCS 8 (QPSK, 7/10 Coding) 9.0454 6.6986 2.3467 MCS 9 (QPSK, 3/4 Coding) 9.8067 7.3389 2.4678 MCS 10 (16QAM, 4/9 Coding) 12.7948 8.0465 4.7483 MCS 11 (16QAM, 1/2 Coding) 13.5300 8.5738 4.9562 MCS 12 (16QAM, 13/24 Coding) 14.2228 9.2461 4.9767 MCS 13 (16QAM, 5/8 Coding) 15.4059 10.2171 5.1888 MCS 14 (16QAM, 2/3 Coding) 16.0129 10.2391 5.7739 MCS 15 (16QAM, 3/4 Coding) 17.3048 11.0832 6.2215 MCS 16 (16QAM, 5/6 Coding) 18.0120 12.7021 5.3099 MCS 17 (64QAM, 3/5 Coding) 22.6151 13.8120 8.8031 MCS 18 (64QAM, 5/8 Coding) 21.9743 14.3230 7.6513 MCS 19 (64QAM, 17/25 Coding) 22.4824 15.3110 7.1714 MCS 20 (64QAM, 3/4 Coding) 23.3498 16.0875 7.2623 MCS 21 (64QAM, 5/6 Coding) 23.4619 17.6534 5.8086

Referring to Table 1, 22 MCS levels have different coding schemes and modulation schemes. The rate is determined by the MCS level, which is the level for a predefined modulation and channel coding combination. The MCS level is determined according to the received SNR, and the MCS level having the highest efficiency is selected according to the SNR.

The SNR value in the proposed scheme to satisfy 10% frame error rate at all MCS levels is lower than the SNR value in the conventional scheme. It is shown that the proposed technique according to an embodiment of the present invention under the condition of satisfying the same frame error rate may have a lower SNR value than that of the conventional technique.

For example, at the MCS 0 level, an SNR of about 3.46 dB is required for the conventional scheme, and an SNR of 2.31 dB is required for the proposed scheme. Therefore, it can be seen that the performance improvement of about 1.15 dB is achieved by the proposed technique. In the MCS 17 level, the proposed technique produces about 8.80 dB in SNR compared to the conventional scheme. On average, a performance improvement of about 4.23 dB occurs.

As such, the SNR for satisfying the 10% frame error rate is relatively lower than that of the conventional technique in the case of the proposed technique. In order to satisfy the same frame error rate, it can be seen that a relatively higher SNR value is required in the case of performing the conventional scheme. In other words, it can be seen that the performance deterioration can be reduced in the case of demapping by the proposed technique, compared to the case of demapping by the conventional technique.

This is because the function value is calculated using only the basic constellation point on the basic constellation in case of demapping by performing the existing technique. The function value P by performing the DPC can be obtained by the following equation.

Equation 9 uses the received signal y that is moved into the modulo box when performing a modulo operation on the input signal. However, the received signal obtained by performing the modulo operation may be moved to a position other than the position on the constellation at the time of transmission while the noise is included. In other words, while moving the received signal into the modulo box by modulo operation, the function value P may be changed by Equation 9 by moving to a position other than the position of the constellation symbol during transmission. Therefore, this may cause a significant error in obtaining the LLR value.

Hereinafter, an error caused in demapping transmission symbols transmitted by the DPC will be described in detail.

로서 약 0.707의 값을 가진다. 여기서 가로 축은 성상도 상의 실수 값을 나타내며, 세로 축은 모듈로 연산이 수행된 수신 신호이다.FIG. 6 shows a received signal in which a modulo operation is performed on a symbol mapped to QPSK when noise is a Gaussian distribution. In the transmitter, when the symbol of (00) is transmitted and the SNR is 5 dB, the distribution of the real value of the received signal after the modulo operation is obtained by simulation. Where the value of A is

It has a value of about 0.707. Here, the horizontal axis represents a real value in constellation and the vertical axis represents a received signal on which a modulo operation is performed.

Referring to FIG. 6, the (00) symbol may include a noise of a Gaussian distribution while being transmitted. Modulo operation is performed on the received signal containing noise. After the modulo operation is performed, the noise no longer has a Gaussian distribution. This is because a received signal that is out of the modulo box by a modulo operation is forced into the modulo box.

Accordingly, a received signal having a Gaussian distribution is obtained in the range of ± 0.5 based on the value of A, about 0.707, and if it is out of this range, it no longer has a Gaussian distribution. It can also be seen that the right end is forcibly cut off by modulo operation.

As described above, when the real value of the signal received by the modulo operation is A, the LLR value is highest, and when the real value of the received signal is outside A, the LLR value is lowered like the Gaussian distribution. However, it can be seen that the left portion does not have a Gaussian distribution and thus has a tail that increases again after the LLR of the received signal becomes zero.

Figure 7 shows the coordinates on the constellation for the signal receiving the QPSK mapped symbol. The transmitter is a result of real and imaginary values of data obtained after performing a modulo operation on a received signal when the symbol of (00) is transmitted and the SNR is 5 dB. Here, the horizontal axis is the coordinate of the real value on the constellation, and the vertical axis is the coordinate of the imaginary value on the constellation.

로서 약 0.707이다. 하지만, 모듈로 연산에 의하여 수신된 신호는 (A, A) 좌표를 기준으로 원의 형태에서 일부분이 절단되어 다른 쪽에 이동된다. 이는 모듈로 연산에 (A, A) 좌표를 기준으로 원의 형태로 수신된 신호들 일부가 모듈로 박스 외부에 있어, 이 를 모듈로 연산을 수행하면 도 7의 테두리인 모듈로 박스 내부로 수신된 신호의 좌표가 이동되기 때문이다.Referring to FIG. 7, in the case where the noise has a Gaussian distribution, the coordinates of a symbol obtained after performing a modulo operation at the receiver have a shape of a circle based on (A, A) coordinates to which (00) symbols are mapped. to be. Where the value of A is

As about 0.707. However, the signal received by the modulo operation is cut off a part of the circle based on the (A, A) coordinates and moved to the other side. This is because some of the signals received in the form of circles based on the (A, A) coordinates are modulo boxes outside the modulo operation. This is because the coordinates of the acquired signal are shifted.

In the distribution of the received signal, it can be seen that the right part and the upper part are cut off in the shape of a circle based on the (A, A) coordinates. The cut portion of the right portion is moved to the upper left portion by modulo operation, and the cut portion of the upper portion is moved to the lower right portion by modulo operation based on the coordinates (A, A). For example, received signals located outside the upper right corner of the modulo box may be moved to three points by modulo operation. In other words, received signals located outside the upper right corner by modulo operation may be moved to the lower right, upper left and lower left.

As such, when noise is included in the symbol on the constellation transmitted from the transmitter, the receiver moves the received symbol to the modulo box by modulo operation. However, while performing a modulo operation, the symbol is moved to a position different from the position of the symbol on the transmitted signal, so that the receiver can demap the symbol to a completely different symbol. Obtaining LLRs in these situations can cause performance degradation.

However, according to the demapping method according to an embodiment of the present invention, the extended constellation point is introduced as well as the basic constellation points using the extended constellation. With this, by using the Castle candidate store the received signal to obtain a candidate value of the function (P _i) representing the likelihood of the candidate corresponding to the gender store, the maximum value among selects the maximum value of the function (P _max). Therefore, even if the received signal is moved to another position in the modulo box by modulo operation, the maximum function value P _max can be calculated relatively accurately by using the corresponding candidate constellation point. This reduces errors in the LLR values and reduces performance degradation of the de-mapped bit stream.

The present invention may be implemented in hardware, software, or a combination thereof. (DSP), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a microprocessor, and the like, which are designed to perform the above- , Other electronic units, or a combination thereof. In the software implementation, the module may be implemented as a module that performs the above-described function. The software may be stored in a memory unit and executed by a processor. The memory unit or processor may employ various means well known to those skilled in the art.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention. You will understand. Therefore, it is intended that the present invention covers all embodiments falling within the scope of the following claims, rather than being limited to the above-described embodiments.

As described above, according to the present invention, it is possible to reduce the loss due to modulo operation in demapping, thereby reducing the reception error due to data transmission.

Claims (10)

Generating a maximum function value that is most likely to correspond to a candidate constellation point on the extended constellation by performing a modulo operation on the input signal; And Generating a log-likelihood ratio (LLR) using the maximum function value; The extended constellation includes a basic constellation point of the basic constellation and at least one extended constellation point. The method of claim 1, further comprising demapping a constellation point to which the received signal is mapped to a bit string based on the LLR. The method of claim 1, wherein the extended constellation point is Demapping method characterized in that the constellation point is arranged to repeat the basic constellation point. The method of claim 1, wherein the extended constellation point is The de-mapping method is characterized in that the modulo box including the basic constellation point is repeatedly positioned adjacent to the modulo box and selected from constellation points disposed close to the basic constellation points of the basic constellation. . The method of claim 1, wherein the candidate constellation point is And a constellation point representing the same bit string on the extended constellation diagram. The method of claim 1, wherein the maximum function value is A demapping method using the following equation.

Where σ ² is the noise variance, y is the received signal, h is the channel response, and (α) _i is the coordinate on the extended constellation of the candidate constellation point. The method of claim 1, wherein the LLR Demapping method using the following equation.

Where α is one of the coordinates of the constellation symbol on the constellation map,

Is a symbol whose 0 is the k-th bit mapped to the real value of α,

Is a symbol whose k-th bit is mapped to the real value of α. Generating a received signal by performing a modulo operation on the input signal; Generating, by the received signal, candidate function values for the constellation points using a plurality of constellation points representing the same bit string; And Generating a log-likelihood ratio (LLR) using the candidate function values. The method of claim 8, wherein the LLR And generating using the maximum function value which is the maximum value among the candidate function values. A modulo operator for performing a modulo operation on the input signal to generate a received signal; A demapper for generating a log-likelihood ratio (LLR) by selecting one function value among candidate function values indicating a possibility that the received signal corresponds to a candidate constellation point on an extended constellation; And A decoder for decoding a symbol generated by the LLR, The extended constellation includes a basic constellation point of the basic constellation and at least one extended constellation point.

Images