JP2006157390A - Device and method for separating signal - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for separating a signal capable of achieving efficiency of an arithmetic operation related with a QRM-MLD system signal separation. <P>SOLUTION: The device for separating the signal is used by an MIMO system radio receiver, and performs signal separation in a QRM-MLD system. The device is provided with a candidate deciding means 318 for determining the candidates of a plurality of signal points on the constellation of a signal by using the matrix elements of a triangular matrix R, a distance calculating means 320 for squaring a distance between a signal point corresponding to the vector components of a reception signal vector after unitary conversion and the candidates of a plurality of signal points, a selecting means 322 for selecting one or more signal points according to a selection reference from among the candidates of the plurality of signal points, and a selection reference setting means 321 for setting the selection reference according to a communication circumstance. The candidates of the plurality of signal points are further determined based on the matrix elements of the triangular matrix and one or more signal points. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention generally relates to the technical field of multi-input multi-output (MIMO) wireless communication, and more particularly to a signal separation device and a signal separation method used in a receiver for the MIMO system.

  In this type of technical field, research and development for realizing high-capacity high-speed information communication at present and the next generation and beyond are underway. For example, in addition to a single input single output (SISO) method, a single input multiple output (SIMO) method, a multiple input single output (MISO) is used from the viewpoint of increasing communication capacity. Research is being conducted on the Output (Multiple Input Multiple Output) system and the like.

FIG. 1 shows an overview of a MIMO communication system including a transmitter 102 and a receiver 104. In the MIMO scheme, different signals are transmitted simultaneously from the plurality of transmitting antennas 106-1 to 106-N at the same frequency. These transmitted signals x ₁ ,. . . , _{X 4} is received by a plurality of receiving antennas 108-1～N. For simplicity, the number of transmission antennas and the number of reception antennas are both N, but different numbers of antennas may be used. Within the receiver 104, the received signals y ₁ ,. . . , Based on the y _4, the process of separating the individual signals a plurality of signals from the transmitter is performed. The separated signal is supplied to a subsequent processing element for further demodulation processing.

There are several methods for signal separation performed in the receiver 104. One is a method called Maximum Likelihood Detection (MLD) method. This calculates the Euclidean distance or its square for all possible combinations of multiple transmitted and received signals transmitted from multiple transmit antennas, and the most probable estimate of the transmitted signal combination that gives the smallest distance. Select as a result. According to this method, it is possible to reliably separate a plurality of transmission signals into individual signals, but due to the calculation of the square Euclidean distance many times, the calculation burden required for signal separation becomes very large. There is. For example, it is assumed that four transmission signals are respectively transmitted from four transmission antennas in a 16QAM modulation scheme. In this case, since one transmission signal is mapped to any of 16 signal points, the total number of combinations of transmission signals included in the reception signal is (number of signal points for one transmission signal) ^{(number of transmission antennas} ) ⁾ = 16 ⁴ = 65536 ways. Calculation of the squared Euclidean distance for the received signal and all these combinations and selecting the most probable combination requires a very large computing power, and particularly hinders downsizing of the mobile terminal. Furthermore, if the calculation burden is large, the power consumption also increases, which is particularly disadvantageous for small mobile terminals.

  As a signal separation method improved from the MLD method, there is a QRM-MLD method. This is an attempt to reduce the number of square Euclidean distance calculations by improving the MLD method using QR decomposition and the M algorithm. According to this method, the number of squared Euclidean distances is calculated as follows: (number of first-stage signal point candidates) + (number of newly added signal point candidates) × (number of signal surviving in the previous stage) (Number of point candidates) × (number of transmission antennas) = 16 + 16 × 16 × 3 = 784 times. Such QRM-MLD method is described in Non-Patent Document 1, for example.

FIG. 2 is a partial block diagram of a receiver that performs signal separation by the conventional QRM-MLD method. For simplicity, it is assumed that the signal x = (x ₁ ... X ₄ ) ^T is transmitted from each of the four transmission antennas in a 16QAM modulation scheme (the superscript symbol T represents transposition). ). The signal x is also referred to as a transmission signal vector and constitutes one symbol. x ₁ , x ₂ , x ₃ , x ₄ are also referred to as transmission signals or vector components. The receiver includes a plurality of receiving antennas 202-1, 202-2, 202-3, 202-4, a channel estimation unit 204, a ranking unit 206, a reordering unit 208, a QR decomposition unit 210, and a signal conversion. Unit 212, maximum likelihood determination unit 214, and likelihood output unit 215. The maximum likelihood determination unit 214 includes four determination units 216-1, 216-2, 216-3, and 216-4. The number of determination units is provided according to the number of transmission signals. Since each determination part has the same processing block, the 4th determination part 216-4 is demonstrated on behalf of them. The determination unit includes a symbol replica generation unit 218-4, a square Euclidean distance calculation unit 220-4, and a surviving symbol candidate selection unit 222-4.

The channel estimation unit 204 obtains a channel impulse response value (CIR) or a channel estimation value based on a received signal including a known pilot signal on both transmission and reception sides. The matrix H having the channel estimation value h _nm as a matrix element is called a channel matrix. However, h _nm represents a channel estimation value between the m-th transmitting antenna and the n-th receiving antenna, and in the present example, 1 ≦ n and m ≦ 4.

The ranking unit 206 receives a plurality of received signals y ₁ ,. . . And rating or ranking the y ₄ in the order of magnitude of the power.

  Rearrangement section 208 notifies QR decomposition section 210 and signal conversion section 212 of the order in which a plurality of received signals are arranged.

The QR decomposition unit 210 obtains the matrices Q and R so that the channel matrix H obtained by the channel estimation unit 204 is represented by the product of the unitary matrix Q and the upper triangular matrix R (H = QR). The unitary matrix Q in this case satisfies Q ^H Q = QQ ^H = I, may be a square matrix, and may have different numbers of rows and columns. The superscript H represents conjugate transpose, and I represents the unit matrix.

The signal conversion unit 212 performs signal conversion by multiplying the received signal vector y = (y ₁ ... Y ₄ ) ^T by the conjugate transpose matrix Q ^H of the unitary matrix Q. Y = Hx = QRx is established between the transmission signal and the reception signal. When this expression is multiplied by Q ^H from the left, the left side becomes Q ^H y = z and the right side becomes Q ^H QRx = Rx, so the relationship between the transmission and reception signals can be expressed as z = Rx. However, z = (z ₁ ... Z ₄ ) ^T = Q ^H y. z is referred to as a received signal vector after unitary transformation.

上記の関係式は、
ｚ_１＝ｒ_１１ｘ_１＋ｒ_１２ｘ_２＋ｒ_１３ｘ_３＋ｒ_１４ｘ_４
ｚ_２＝ｒ_２２ｘ_２＋ｒ_２３ｘ_３＋ｒ_２４ｘ_４
ｚ_３＝ｒ_３３ｘ_３＋ｒ_３４ｘ_４
ｚ_４＝ｒ_４４ｘ_４
とも書ける。

The above relational expression is
z ₁ = r ₁₁ x ₁ + r ₁₂ x ₂ + r ₁₃ x ₃ + r ₁₄ x ₄
z ₂ = r ₂₂ x ₂ + r ₂₃ x ₃ + r ₂₄ x ₄
z ₃ = r ₃₃ x ₃ + r ₃₄ x ₄
z ₄ = r ₄₄ x ₄
You can also write.

The maximum likelihood determination unit 214 narrows down transmission signal candidates (also referred to as symbol candidates) by the maximum likelihood determination method (MLD method), that is, reduces the number of candidates. Symbol replica generation unit 218-4 of the determining unit 216-4 uses matrix elements of the upper triangular matrix R, generating a candidate of the transmission signal corresponding to the received signal y _4. The number of candidates is fixedly set, for example, C.

Squared Euclidean distance calculation unit 220-4, and the received signal z ₄ after conversion, and calculates the squared Euclidian distance between a candidate of the C signal points. The squared Euclidean distance represents a metric that is the basis for calculating the likelihood. A candidate with a small square Euclidean distance is determined to be close to the transmitted symbol.

The survival symbol candidate selection unit 222-4 outputs S ₁ (≦ C) candidates as survival candidates based on the squared Euclidean distance for each candidate.

  The likelihood output unit 215 calculates the likelihood or likelihood of the candidate output from the last surviving symbol candidate selection unit. More specifically, this likelihood is expressed by a log likelihood ratio (LLR). The output from the likelihood output unit 215 represents a signal separation result and is transmitted to a demodulator (for example, a turbo decoder) at a subsequent stage.

The operation will be described next. The receiver receives a received signal y ₁ ~y ₄ transmit signals in four receiving antennas. They are given to the channel estimation unit 204 and the signal conversion unit 212. The channel estimation unit 204, the ranking unit 206, and the rearrangement unit 208 determine the order in which a plurality of received signals are arranged. Here, it is assumed that the received signals are arranged in the order of the magnitude of the received power, and for the sake of simplicity, the received power increases in the order of x ₁ , x ₂ , x ₃ , x ₄ . The received signal is unitarily converted by the signal conversion unit 212 as z = (z ₁ ... Z ₄ ) ^T = Q ^H y, and the converted signal is input to the maximum likelihood determination unit 214.

In the first stage (first stage) in the maximum likelihood determination unit 214, processing corresponding to the initial setting is performed by the determination unit 216-4. At this stage, attention is paid to the above-described equation regarding z ₄ . It can be seen that the matrix element r ₄₄ is known and z ₄ does not interfere with other signals and only depends on one transmission signal x ₄ . Thus, for the transmission signal x _4, there are at most of the signal points 16 different candidates. Symbol candidate generating unit 218-4 generates a candidate signal points of the 16 relates _{x 4} (C = 16). In other words, 16 signal points on the signal constellation are selected. Squared Euclidean distance between the received signal z ₄ after these candidates and the fourth transformation, is calculated as the square Euclidean distance calculation unit 220-4, a distance smaller order _one S candidate is selected, they are surviving candidates It becomes.

The second stage (second stage) is performed by the determination unit 216-3. Here, attention is paid to an expression relating to z ₃ . The matrix elements r ₃₃ and r ₃₄ are known, and there are 16 candidates for x _{4, and} there are 16 signal point candidates for x ₃ . 16 signal points as a new signal points for x ₃ is introduced by the symbol generator 218-3. Therefore, there can be 16 × 16 = 256 combinations of signal points (there are 256 candidates). It calculated squared Euclidean distance 256 between these candidates and the third reception signal x ₃ is, by selecting the combination of 16 in ascending order of the value (S 2 ₌ 16), candidates are narrowed down.

In the third stage (third stage), the determination unit 216-2 performs similar processing. In this stage, attention is focused on expression for z _2. The matrix elements r ₂₂ , r ₂₃ , and r ₂₄ are known, and the combinations of the transmission signals x ₃ and x ₄ are narrowed down to 16 candidates in the previous stage, and there are 16 signal point candidates for x _2. To do. Therefore, the symbol candidate generating unit 218-2 generates 16 candidates for _{x 2.} Also in this case, candidates are narrowed down by selecting 16 (S ₃ = 16) candidates having a small square Euclidean distance from among 256 combinations of signal points.

In the fourth stage (fourth stage), the determination unit 216-1 performs similar processing. In this stage, attention is focused on expression for z _1. The matrix elements r ₁₁ , r ₁₂ , r ₁₃ , r ₁₄ are known, and the combinations of the transmission signals x ₂ , x ₃ , x ₄ are narrowed down to 16 candidates in the previous stage, and 16 combinations of x ₁ There are signal point candidates. Therefore, the symbol candidate generating unit 218-1 generates 16 candidates for _{x 1.} Also in this case, candidates are narrowed down by selecting 16 (S ₄ = 16) candidates having a small square Euclidean distance from among 256 combinations of signal points.

In this way, by limiting the number of candidates to a certain number (such as S ₁ ≦ C) or less at each stage, the signal of the transmission signal can be obtained without calculating the square Euclidean distance for all possible combinations of signal points. Point candidates can be narrowed down.
K. J. et al. Kim, et al. "Joint channel estimation and data detection algorithms for MIMO-OFDM systems", Proc. 36th Asimar Conference on Signals, Systems and Computers, Nov. 2002

By the way, although the number of candidates output at each stage is equal to each other, the probability of the surviving candidates output at each stage is not necessarily the same in all stages. In order to explain this, for example, it is assumed that four transmission signals are modulated by the QPSK method and transmitted from four transmission antennas, respectively. In this case, it is assumed that a signal A, B, C, or D transmitted from a certain transmission antenna is received as a signal a, b, c, or d through a certain channel as shown in FIG. Correspondences between signal points A, B, C, and D and signal points a, b, c, and d can be predicted based on channel estimation values. Similarly, it is assumed that a signal E, F, G, or G transmitted from another transmitting antenna is received as a signal e, f, g, or h through another channel as shown in FIG. Correspondences between signal points E, F, G, and H and signal points e, f, g, and h can also be predicted based on channel estimation values. It is assumed that the number C of surviving candidates at each stage is 2 (C = 2). Furthermore, in the first stage of the QRM-MLD system as described above, it is assumed that the vector component z ₄ included in the received signal vector after unitary conversion is positioned at a signal point as shown in FIG. In the first stage, out of the four signal points a, b, c and d, two signal points a and b having a small square Euclidean distance from z ₄ are selected as survival candidates.

In the second stage, as shown in FIG. 6, four signal points are introduced as new candidates for each of the two survival candidates a and b. The new signal points to be introduced are due to the signals e, f, g and h. The coordinates of the signal point z ₃ estimated by the second stage, when positioned point P in FIG. 6, the distance a small signal point g between the point P, two of h becomes surviving candidates. In such a case, it can be said that the signal points g and h are significantly more likely than the signal points e and f. However, the coordinates of the signal point z ₃ is the situation when positioned point Q of Figure 6 differs. In this case, two signal points e and h having a short distance from the point Q may become survival candidates, but from the viewpoint of the size of the distance, the signal points f and g and the point Q to be discarded The distance between is comparable to the distance between signal points e and h. Therefore, if the signal points f and g are discarded in such a situation, there is a concern that the subsequent signal estimation accuracy may be adversely affected. Saritote, increasing the number of surviving candidates to four, when the signal point z ₃ as positioned point P, will be considered unnecessary candidates such as signal points e, f, the amount of calculation increases in vain There is a concern that

  The present invention has been made to address at least one of the above-described problems, and its problem is to provide a signal separation apparatus and a signal separation capable of improving the efficiency of operations related to QRM-MLD signal separation. Is to provide a method.

  In the present invention, a signal separation device that performs signal separation by the QRM-MLD method, which is used in a MIMO wireless reception device, is used. This apparatus uses a matrix element of a triangular matrix to determine candidate signal means for determining a plurality of signal point candidates on a signal constellation, a signal point corresponding to a vector component of a received signal vector after unitary transformation, A distance calculating means for obtaining a square of a distance between a plurality of signal point candidates; a selecting means for selecting one or more signal points from the plurality of signal point candidates according to a selection criterion; And a selection criterion setting means for setting the selection criterion accordingly. Based on the matrix elements of the triangular matrix and the one or more signal points, a plurality of signal point candidates are further determined.

  According to the present invention, it is possible to improve the efficiency of operations related to QRM-MLD signal separation.

  According to one aspect of the present invention, in a signal separation apparatus that performs signal separation by the QRM-MLD method, a plurality of signal point candidates on a signal constellation are determined using a matrix element of a triangular matrix, and unitary conversion is performed. A square of a distance between a signal point corresponding to a vector component of a later received signal vector and the plurality of signal point candidates is obtained, and according to a selection criterion that can be changed according to a communication situation, One or more signal points are selected from the candidates, and a plurality of signal point candidates are further determined based on the matrix elements of the triangular matrix and the one or more signal points. Since one or more signal points are selected from the plurality of signal point candidates according to a selection criterion that can be changed according to the communication status, an operation is performed on the number of signal points suitable for the communication status. Arithmetic can be executed without excess or deficiency without degrading the calculation accuracy.

  According to an aspect of the present invention, the selection criterion setting unit calculates a ratio of a desired signal and an undesired signal in a received signal and determines the number of signal point candidates to be selected. According to an aspect of the present invention, the selection criterion setting unit determines the number of candidate signal points to be selected according to the minimum square value of a plurality of distances. According to one aspect of the present invention, the selection criterion setting means determines the number of signal point candidates to be selected according to the average value of the squares of a plurality of distances. The quality of the signal quality and the magnitude of the distance reflect the state of the communication environment. Therefore, an appropriate number of surviving candidates is set by determining a selection criterion based on them.

  According to an aspect of the present invention, the distance calculating unit calculates a square of two or more distances related to two or more received signal vectors or two or more subcarrier components of the received signal vector obtained at different times. According to an aspect of the present invention, the selection criterion setting unit determines the number of signal point candidates to be selected, depending on whether the difference between the squares of the two or more distances is included in a predetermined range. To do. According to an aspect of the present invention, the selection criterion setting unit determines the number of signal point candidates to be selected according to the average value of the squares of the two or more distances. By considering two or more received signal vectors, it is possible to set the number of surviving candidates more appropriately reflecting the communication status.

  According to an aspect of the present invention, the total number of signal point candidates used for estimating all vector components included in one received signal vector is set to be equal to or less than a predetermined value. Thereby, the calculation burden required for the entire signal separation can be suppressed to a certain limit while appropriately changing the number of candidates at each stage.

  Embodiments relating to the present invention will be described below with reference to the drawings. All or some of the elements depicted in the functional block diagram may be realized as hardware, software, or a combination thereof. In the following embodiment, four transmission signals are transmitted from four transmission antennas and received by four reception antennas. Of course, the number of transmission signals, the number of transmission antennas, the number of reception antennas, etc. are other values. There may be.

  FIG. 7 shows a partial block diagram of a signal separation device according to an embodiment of the present invention. The part shown in the drawing is used instead of one of the plurality of determination units shown in FIG. 2, and all or part of the determination units 216-1 to 216-4 may include the functional unit shown in FIG. 3. . In this embodiment, four determination units 316 are prepared, and the maximum likelihood determination unit 214 is configured by them. The determination unit 316 illustrated in FIG. 7 includes a symbol replica generation unit 318, a square Euclidean distance calculation unit 320, a selection criterion determination unit 321, and a survival candidate selection unit 322.

  The symbol replica generation unit 318 uses the matrix elements of the upper triangular matrix R to generate transmission signal candidates corresponding to the received signal, that is, several signal points on the signal constellation are selected.

  The square Euclidean distance calculation unit 320 calculates a square Euclidean distance between a signal point corresponding to the received signal after unitary conversion and a signal point candidate. The square Euclidean distance represents a metric that is the basis for calculating the likelihood. A candidate with a small square Euclidean distance represents that it is close to the transmitted signal.

  The selection criterion determination unit 321 determines the number of candidates selected by the survival candidate selection unit and / or the determination criterion for selecting the survival candidate according to the communication status. More specifically, the communication status varies depending on fluctuations in received power, noise level, fading, and the like, and such communication status is monitored by elements not shown. As criteria for determining the number of surviving candidates, several criteria can be considered, as will be exemplified below.

  (1) If the state of the channel or the propagation path is good, the calculated square Euclidean distance is considered to be probable, so the number of surviving candidates may be reduced, otherwise the number of candidates may be increased. The quality of the channel state may be determined based on, for example, the ratio of the desired wave and the undesired wave in the received signal. Furthermore, the quality of the channel state may be one of two choices of good or not, and the channel state may be expressed in more than two stages.

  (2) If the least squares Euclidean distance is small, the value is probable, so the number of surviving candidates may be reduced. Otherwise, the number of candidates may be increased.

  (3) The number of surviving candidates may be reduced if the average value of the square Euclidean distance is small, and the number of candidates may be increased if the average value is large. This is because if the average value is large, the signal point of the received signal greatly deviates from the candidate signal point, and it is considered that the probability is low.

The survival candidate selection unit 322 outputs S ₁ candidates as survival candidates based on the square Euclidean distance and selection criteria for each candidate.

The operation will be described with reference to FIGS. The receiver receives a received signal y ₁ ~y ₄ transmit signals in four receiving antennas. They are given to the channel estimation unit 204 and the signal conversion unit 212. The channel estimation unit 204, the ranking unit 206, and the rearrangement unit 208 determine the order in which a plurality of received signals are arranged. Here, it is assumed that the received signals are arranged in the order of the magnitude of the received power, and for the sake of simplicity, the received power increases in the order of x ₁ , x ₂ , x ₃ , x ₄ . The received signal is unitarily converted by the signal conversion unit 212 as z = (z ₁ ... Z ₄ ) ^T = Q ^H y, and the converted signal is input to the maximum likelihood determination unit 214.

In the first stage of maximum likelihood determination section 214, processing corresponding to the initial setting is performed by the determination section. At this stage, attention is paid to the above-described equation regarding z ₄ . The matrix element r ₄₄ is known and z ₄ depends only on one transmitted signal x ₄ (does not interfere with other signals). Thus, for the transmission signal x _4, there are at most of the signal points 16 different candidates. Symbol candidate generator 320 generates candidates of the 16 signal points for _{x 4.} In other words, 16 signal points on the signal constellation are selected. The square Euclidean distance between these candidates and the fourth received signal z ₄ after unitary transformation is calculated by the square Euclidean distance calculation unit 320.

  In this embodiment, the selection criterion determination unit 321 determines a criterion for selecting a survival candidate based on the squared Euclidean distance, and notifies the survival candidate selection unit 322 of the determination criterion. Next, the survival candidate selection unit 322 selects one or more candidates from the 16 candidates based on the notified determination criterion and the squared Euclidean distance. For example, if the notified criterion is that the signal-to-interference power ratio (SIR) exceeds a predetermined value, 8 candidates are selected, and if not, 16 candidates are selected. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates.

In the determination unit that performs the second stage, attention is paid to an expression relating to z ₃ . Matrix elements r ₃₃ and r ₃₄ are known, x ₄ has been estimated, and 16 signal point candidates exist for x ₃ . Sixteen signal points are introduced by the symbol generator 318 as new signal points for x ₃ . Square Euclidean distances between these candidates and the third received signal x ₃ are calculated, and are supplied to the selection criterion determination unit 321 and the surviving candidate selection unit 322. Also in this case, the selection criterion determination unit 321 determines the determination criterion and notifies the survival candidate selection unit 322. The survival candidate selection unit 322 selects one or more candidates from the 16 candidates based on the notified determination criterion and the squared Euclidean distance. For example, as described above, the judgment criterion is notified so that eight candidates are selected if the signal-to-interference power ratio (SIR) exceeds a predetermined value, and 16 candidates are selected otherwise. . A judgment criterion different from the first stage may be adopted. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates. In this case, the number of candidates S ₁ to survive in the first stage, the number of candidates S ₂ survive in the second stage requires noted that it is not necessarily the same. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates.

Similar processing is performed in the determination unit in the third stage. In this stage, attention is focused on expression for z _2. The matrix elements r ₂₂ , r ₂₃ , and r ₂₄ are known, the transmission signals x ₃ and x ₄ have been estimated in the previous stage, and there are 16 signal point candidates for x ₂ . Therefore, the symbol candidate generating unit 318 generates sixteen candidates for _{x 2.} Square Euclidean distances between these candidates and the second received signal x ₂ are calculated, and are supplied to the selection criterion determining unit 321 and the surviving candidate selecting unit 322. Also in this case, the selection criterion determination unit 321 determines the determination criterion and notifies the survival candidate selection unit 322. The survival candidate selection unit 322 selects one or more candidates from the 16 candidates based on the notified determination criterion and the squared Euclidean distance. For example, as described above, the judgment criterion is notified so that eight candidates are selected if the signal-to-interference power ratio (SIR) exceeds a predetermined value, and 16 candidates are selected otherwise. . A judgment criterion different from the first and second stages may be adopted. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates. In this case, it should be noted that the number of candidates S ₁ and S ₂ that survive in the _first and _second stages is not necessarily the same as the number of candidates S ₃ that survive in the third stage. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates.

Similar processing is performed in the determination unit in the fourth stage. In this stage, attention is focused on expression for z _1. The matrix elements r ₁₁ , r ₁₂ , r ₁₃ , r ₁₄ are known, the transmission signals x ₂ , x ₃ , x ₄ have been estimated, and there are 16 signal point candidates for x ₁ . Therefore, the symbol candidate generating unit 318-1 generates 16 candidates for _{x 1.} Square Euclidean distances between these candidates and the first received signal x ₁ are calculated and provided to the selection criterion determination unit 321 and the surviving candidate selection unit 322. Also in this case, the selection criterion determination unit 321 determines the determination criterion and notifies the survival candidate selection unit 322. The survival candidate selection unit 322 selects one or more candidates from the 16 candidates based on the notified determination criterion (for example, the same criterion as described above) and the square Euclidean distance. Different criteria from the first, second, and third stages may be adopted. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates. In this case, it should be noted that the number of candidates S ₁ , S ₂ , S ₃ that survive in the _first , _second , and _third stages is not necessarily the same as the number of candidates S ₄ that survive in the fourth stage. The survival candidate selection unit 322 selects 8 or 16 candidates in ascending order of the square Euclidean distance based on the SIR value, and outputs them as survival candidates.

  This output is input to the likelihood output unit 215, and likelihood or probability is calculated. More specifically, this likelihood is expressed by a log likelihood ratio (LLR). The output from the likelihood output unit 215 represents a signal separation result and is transmitted to a demodulator (for example, a turbo decoder) at a subsequent stage.

  FIG. 8 shows a partial block diagram of a signal separation device according to an embodiment of the present invention. In the first embodiment, four components included in one symbol are estimated in order. In the second embodiment, a plurality of symbols are processed in parallel. The determination unit 416 shown in FIG. 8 can be used for one of the determination units 216-1 to 216-4 shown in FIG. 2, and in this embodiment, four determination units 416 are prepared. A maximum likelihood determination unit 214 is configured. The determination unit 416 includes M (M ≧ 2) symbol replica generation units 418-1 to 418 -M, M square Euclidean distance calculation units 420-1 to M, one selection criterion determination unit 421, and M pieces. The surviving candidate selection unit 422-1.

  Similar to the first embodiment, the symbol replica generation units 418-1 to 418-1 M each determine a plurality of signal point candidates based on matrix elements given in accordance with each stage (stage) of estimating a transmission signal. . The square Euclidean distance calculation units 420-1 to 420-1 M calculate the square Euclidean distance based on the signal point candidates and the received signal after the unitary transformation, and the calculation results are used as the selection criterion determination unit 421 and the survival candidate selection unit 422-1. Give to ~ M respectively. The selection criterion determination unit 421 determines a criterion for selecting a survival candidate for each symbol based on the square Euclidean distance for a plurality of symbols, and notifies the survival candidate selection units 422-1 to M to M, respectively. The plurality of surviving candidate selection units 422-1 to 422-1 to M output surviving candidates in accordance with the instructed criterion.

In the present embodiment, the received signal after unitary conversion for a plurality of symbols is provided to the determination unit 416. As shown in the figure, the received signals related to the symbols 1 to M are given to the symbol replica generation units 418-1 to 418-1. As described above, one symbol is a concept including a plurality of (four in this embodiment) signals, and one transmission symbol is expressed by x = (x ₁ ,..., X ₄ ). One received symbol is represented by y = (y ₁ ,..., Y ₄ ). The plurality of symbols appearing in FIG. 8 may be symbols received at different timings, for example. Alternatively, these symbols may correspond to different subcarrier components in the same symbol.

  The selection criterion determination unit 421 in the present embodiment determines a criterion for selecting a survival candidate for each symbol based on the calculation result of the square Euclidean distance for each symbol. In addition to (1) to (3) shown in the first embodiment, various determination criteria as exemplified below may be adopted.

  (A) If the difference between two or more least squares Euclidean distances calculated for each of two or more symbols is within a predetermined range, the number of candidates to be selected is set to be small; otherwise, the number of candidates is large. May be selected. The minimum value among a plurality of distances calculated by the square Euclidean distance calculation unit 420-1, the minimum value calculated by 420-2,. . . And the minimum value of M calculated in 420-M is as small as each other (if the difference is within a certain small range), the communication status is good. This is because the probability of the calculation result is considered high.

  (B) An average value of two or more least square Euclidean distances calculated for each of two or more symbols may be obtained, and the magnitude of the square Euclidean distance for each symbol may be determined based on the average value. Or size comparison may be performed based on the quantity proportional to the average value. Based on the minimum value of the squared Euclidean distance calculated for each of the two or more symbols (or an amount proportional thereto), the magnitude of the squared Euclidean distance for each symbol may be compared. Furthermore, not only the square Euclidean distance but also the quality of the signal quality may be determined based on an averaged amount or the best value over two or more symbols.

  (C) Although the number of surviving candidates output at each stage of signal estimation is variable, the number of surviving candidates output through all stages may be limited to a certain value or less. For example, the number of candidate signal points output at each stage has 4, 8, or 16 options, but the number of candidates output throughout the four stages is limited to 8 × 4 = 32. Also good. Thereby, it is possible to suppress the amount of calculation (calculation burden) required for the entire signal estimation while making the number of candidates output at each stage variable.

  The judgment criteria (1) to (3) and (a) to (c) exemplified above may be adopted individually or may be used in appropriate combination. Furthermore, for the magnitude of the distance, the quality of the quality, etc., more than two multiple-choice determinations may be performed in addition to the two-way determination. The present invention is not limited to the above criteria, and any appropriate criteria for evaluating the communication status can be adopted.

FIG. 9 conceptually shows the order of processing in signal separation. FIG. 9A shows the order of processing related to the first embodiment, and FIG. 9B shows the order of processing related to the second embodiment. In each figure, a matrix element of a triangular matrix and first and second symbols are drawn. These symbols may be obtained at different timings in time, or may correspond to different subcarrier components. The first symbol has four signals x _1,1,. . . , X _4,1 . The second symbol has four signals x ₁ , _2,. . . , X _4,2 . As shown in FIG. 9 (A), in the manner of Example 1, of the first symbol, the signal x _{4, 1} which does not interfere with other symbol is first estimated, the signal x ₃ subsequent the same symbol _{, 1} , x _2,1 and x _1,1 are estimated in order, after which the second symbol is estimated. This order is indicated by thick arrows in the figure. With respect to such processing order, the conventional method is the same, but Example 1 is significantly different from the conventional method in that the number of surviving candidates output at each stage is variable.

In the method of the second embodiment shown in FIG. 9B, in the first stage, the signals x _4,1 and x _4,2 relating to the first and second symbols are estimated in parallel. Thereafter, similarly, signals x _3,1 and x _3,2 are estimated in parallel, signals x _2,1 and x _2,2 are estimated in parallel, and signals x _1,1 and x _1,2 are parallel. Estimated. This order is indicated by thick arrows in the figure. In this example, since the determination criterion is determined based on the reception status regarding a plurality of symbols, the number of surviving candidates can be set more appropriately according to the communication status.

It is a figure which shows the outline | summary of the communication system of a MIMO system. It is a block diagram of the receiver which performs signal separation. It is a figure which shows the candidate of the signal point in a signal constellation. It is a figure which shows the candidate of the signal point in a signal constellation. It is a figure which shows the signal point corresponding to the candidate of the signal point in a signal constellation, and a received signal. It is a figure which shows the signal point corresponding to the candidate of the signal point in a signal constellation, and a received signal. 1 shows a partial block diagram of a signal separation device according to an embodiment of the present invention. FIG. 1 shows a partial block diagram of a signal separation device according to an embodiment of the present invention. FIG. It is a figure which shows typically the order of the process in Example 1 and 2. FIG.

Explanation of symbols

102 transmitter; 104 receiver; 106-1 to N transmitting antenna; 108-1 to N receiving antenna;
202-1-4 receiving antenna; 204 channel estimation unit; 206 ranking unit; 208 rearrangement unit; 210 QR decomposition unit; 212 signal conversion unit; 214 maximum likelihood determination unit; 215 likelihood output unit; 218-1-4 symbol replica generation unit; 220-1-4 squared Euclidean distance calculation unit; 222-1-4 survival candidate selection unit;
316 determination unit; 318 symbol replica generation unit; 320 squared Euclidean distance calculation unit; 321 selection criterion determination unit; 322 survival candidate selection unit;
416 determination unit; 418-1 to M symbol replica generation unit; 420-1 to M square Euclidean distance calculation unit; 421 selection criterion determination unit; 422-1 to M survival candidate selection unit

Claims (9)

A signal separation device that performs signal separation using a QRM-MLD method, which is used in a MIMO wireless reception device,
Candidate determination means for determining a plurality of signal point candidates on the signal constellation using matrix elements of a triangular matrix;
Distance calculating means for obtaining a square of a distance between a signal point corresponding to a vector component of the received signal vector after unitary conversion and the plurality of signal point candidates;
A selection means for selecting one or more signal points according to a selection criterion from the plurality of signal point candidates;
Selection criteria setting means for setting the selection criteria according to communication conditions;
And a plurality of signal point candidates are further determined based on the matrix elements of the triangular matrix and the one or more signal points. The signal separation device according to claim 1, wherein the selection criterion setting unit calculates a ratio of a desired signal and an undesired signal in a received signal, and determines the number of signal point candidates to be selected. The signal separation device according to claim 1, wherein the selection criterion setting unit determines the number of candidate signal points to be selected according to the minimum value of the square of a plurality of distances. The signal separation device according to claim 1, wherein the selection criterion setting unit determines the number of candidate signal points to be selected according to the average value of the squares of a plurality of distances. The distance calculation unit calculates a square of two or more distances related to two or more received signal vectors obtained at different time points or two or more subcarrier components of the received signal vector. Signal separation device. 6. The selection criterion setting means determines the number of signal point candidates to be selected according to whether or not the difference between the squares of two or more distances is included in a predetermined range. The signal separation device as described. 6. The signal separation device according to claim 5, wherein the selection criterion setting unit determines the number of candidate signal points to be selected according to the average value of the squares of the two or more distances. 6. The total number of signal point candidates used for estimating all vector components included in one received signal vector is set to be equal to or less than a predetermined value. A signal separation device according to claim 1. A signal separation method for performing signal separation by a QRM-MLD method in a MIMO wireless receiver,
Using the matrix elements of the triangular matrix, determine multiple signal point candidates on the signal constellation,
Obtain the square of the distance between the signal point corresponding to the vector component of the received signal vector after unitary transformation and the plurality of signal point candidates,
Selecting one or more signal points from the plurality of signal point candidates according to a selection criterion that can be changed according to a communication situation;
A signal separation method, wherein a plurality of signal point candidates are further determined based on a matrix element of a triangular matrix and the one or more signal points.

Images