JP2009141768A - Reception device and signal selecting method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a reception device which can detect maximum likelihood of low operation quantity. <P>SOLUTION: A reception device separates signals by maximum likelihood detection based on QR decomposition or QL decomposition. The reception device has a likelihood value calculation part for calculating likelihood value to combination of signal points, an order decision part for ordering the likelihood value in ascending order of value and a minimum likelihood value selection part for selecting the minimum likelihood value among the likelihood values. The likelihood value calculation part further calculates a new likelihood value which becomes the next order of the minimum likelihood value based on the order of the likelihood value. The minimum likelihood value selection part determines the minimum likelihood value which is minimum except the minimum likelihood value among the likelihood values including the new likelihood value, and selects combination of signal points corresponding to each of the minimum likelihood values as transmitted symbol candidate at each step. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a receiving apparatus and a signal detection method. In particular, the present invention relates to a receiving apparatus and a signal selection method of a MIMO (Multiple-Input and Multiple-Output) transmission method in a mobile communication system.

  As one technique for increasing the communication speed between wireless devices, a multi-input / multi-output transmission system is known. This method is literally based on signal input / output using a plurality of antennas. The feature of this method is that a plurality of transmission data can be transmitted at the same time and at the same frequency using a plurality of different antennas. Therefore, as the number of channels that can be transmitted simultaneously increases, the amount of information that can be transmitted per unit time can be increased by the increased number of channels. Further, this method has an advantage that the occupied frequency band does not increase when the communication speed is improved.

  However, since a plurality of modulated signals having carrier components of the same frequency are transmitted at the same time, a means for separating the modulated signals (multiplexed signals) that interfere with each other on the receiving side is required. Therefore, on the receiving side, a channel matrix representing the transmission characteristics of the wireless transmission path is estimated, and a transmission signal corresponding to each substream is separated from the received signal based on the channel matrix. The channel matrix is estimated using pilot symbols or the like.

  However, a special contrivance is required to sufficiently remove the influence of noise added in the transmission path and interference generated between substreams and accurately reproduce the transmission signal for each substream. In this regard, in recent years, various techniques related to MIMO signal detection have been developed. For example, a MMSE (Minimum Mean Squared Error) detection method and an MLD (Maximum Likelihood Detection) detection method capable of improving transmission characteristics as compared with this MMSE detection method are known. There is also known an improved method called QR decomposition MLD detection method (hereinafter referred to as QRD-MLD method) that can reduce the amount of calculation required for signal detection for each substream as compared with the MLD detection method (the following patents). Reference 1).

JP 2005-354678 A

  However, the QRD-MLD scheme has a problem that the number of computations of the Euclidean distance used for maximum likelihood detection is still large, and the amount of computation required for signal detection is large. With respect to this problem, Japanese Patent Application Laid-Open No. 2004-228561 discloses a technique for partitioning signal points on a signal point arrangement map for each quadrant and reducing the number of transmission symbol candidates to be referred to when detecting maximum likelihood. However, since this technique limits the selection range of transmission symbol candidates, there is a problem that the estimation accuracy of transmission symbols is lower than that of the original QRD-MLD scheme. Moreover, even if the technique is applied, the number of times of calculation of the Euclidean distance is still large, and in view of the embodiment, a device for further reducing the amount of calculation is required.

  Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to provide a new and improved receiving apparatus capable of further reducing the calculation load required for maximum likelihood detection, And providing a signal selection method.

  In order to solve the above-described problems, according to an aspect of the present invention, there is provided a receiving apparatus that performs signal separation by maximum likelihood detection based on QR decomposition or QL decomposition. The receiving apparatus includes a likelihood value calculating unit that calculates likelihood values for a combination of signal points, a rank determining unit that ranks the likelihood values in ascending order, and a minimum likelihood from the likelihood values. A minimum likelihood value selection unit that selects a value, and the likelihood value calculation unit further calculates a new likelihood value that is the next rank of the minimum likelihood value based on the rank of the likelihood value. The minimum likelihood value selection unit determines a minimum likelihood value that is minimum by removing the minimum likelihood value from the likelihood values including the new likelihood value, and each minimum likelihood value is determined. The combination of signal points corresponding to is selected as a transmission symbol candidate at each stage.

  In addition, for each of the predetermined grids set on the signal point arrangement diagram, the receiving device stores a storage unit in which information on the Euclidean distance to each signal point is recorded, and a received symbol in the predetermined grid. A quantization unit that performs quantization in association with each other may be further included. Further, the likelihood value calculation unit reads the information on the Euclidean distance from the storage unit based on the grid information corresponding to the received symbol, and the likelihood for the combination of the signal points based on the information on the Euclidean distance A value may be calculated.

  In order to solve the above problem, according to another aspect of the present invention, a signal selection method used for maximum likelihood detection based on QR decomposition or QL decomposition is provided. The signal selection method includes a likelihood value calculating step in which likelihood values for combinations of transmission symbol candidates are calculated, a rank determining step in which the likelihood values are ranked in ascending order, and A minimum likelihood value selection step in which the minimum likelihood value is selected from the next likelihood likelihood value, and a new likelihood value that is the next rank of the minimum likelihood value is calculated based on the rank of the likelihood value A calculation step, and a next-order minimum likelihood value selection step in which a minimum likelihood value that is the minimum is determined by removing the minimum likelihood value from the likelihood values including the new likelihood value, Likelihood that the minimum likelihood value obtained by the minimum likelihood value selecting step, the next-rank likelihood value calculating step and the next-rank likelihood value selecting step repeatedly executed corresponds to a combination of transmission symbol candidates. To be selected as a value And butterflies.

  By using the above-described apparatus or method, the number of calculations of the Euclidean distance used for maximum likelihood detection is reduced.

  As described above, according to the present invention, it is possible to further reduce the calculation load required for maximum likelihood detection.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

<First Embodiment>
First, the system configuration of the MIMO system according to the first embodiment of the present invention will be described with reference to FIG. FIG. 1 is an explanatory diagram showing a system configuration of a MIMO system according to the present embodiment. This embodiment is characterized by signal detection means provided on the receiving side.

  As illustrated in FIG. 1, the MIMO system includes a transmission apparatus 100 having a plurality of antennas and a reception apparatus 200 having a plurality of antennas. The transmission apparatus 100 mainly includes an error correction encoding unit 102 and a modulation unit 104. The transmission data is subjected to error correction coding by the error correction coding unit 102, modulated by the modulation unit 104 according to a predetermined modulation scheme, and transmitted.

  The receiving apparatus 200 mainly includes a MIMO signal detection unit 202, a channel estimation unit 204, and an error correction decoding unit 206. The receiving apparatus 200 receives signals spatially multiplexed in the transmission path with a plurality of antennas. This received signal is input to the MIMO signal detection unit 202 and the channel estimation unit 204. Channel estimation section 204 estimates channel matrix H representing the state of the transmission path based on the received pilot symbols and the like. This channel matrix H is input to the MIMO signal detection unit 202.

  The MIMO signal detection unit 202 is means for separating multiple signals received via a plurality of antennas. In particular, MIMO signal detection section 202 separates the multiplexed signal into substreams by maximum likelihood detection based on channel matrix H estimated by channel estimation section 204. The signal separated by MIMO signal detection section 202 is input to error correction decoding section 206, subjected to error correction decoding, and output as reproduction data for each substream.

  The system configuration of the MIMO system according to this embodiment has been briefly described above. The present embodiment is characterized by the functional configuration of the MIMO signal detection unit 202 included in the reception device 200. In particular, it is characterized by means for selecting a transmission symbol candidate at the time of maximum likelihood detection, and means for acquiring a Euclidean distance between each transmission symbol candidate and a reception symbol. Therefore, a more detailed description of these means will be added.

(Transmission symbol candidate selection control)
Next, the functional configuration of the MIMO signal detection unit 202 according to the present embodiment will be described with reference to FIG. In particular, the selection control means for selecting a transmission symbol candidate at the time of maximum likelihood detection will be described in more detail. FIG. 2 is an explanatory diagram illustrating a functional configuration of the MIMO signal detection unit 202 according to the present embodiment.

  As shown in FIG. 2, the MIMO signal detection unit 202 mainly includes a Euclidean distance calculation unit 232, a cumulative likelihood value calculation unit 234, a likelihood rank determination unit 236, a minimum likelihood value calculation unit 238, The signal point candidate setting unit 240, the next order likelihood calculating unit 242, and the next order likelihood position detecting unit 244 are configured.

(Euclidean distance calculation unit 232)
The Euclidean distance calculation unit 232 determines between the estimated transmission symbol estimated from the received signal using the channel matrix H and the transmission symbol candidate corresponding to the signal point on the signal point arrangement diagram (constellation) of a predetermined modulation scheme. This is a means for obtaining the Euclidean distance.

  When a position on the constellation is given, for example, the Euclidean distance calculation unit 232 ranks the transmission symbol candidates in order from the closest Euclidean distance from the position, and calculates the Euclidean distance corresponding to the transmission symbol candidates of a predetermined rank. Can do. Further, when the position and order corresponding to the received symbol are given, the Euclidean distance calculation unit 232 can output the Euclidean distance corresponding to the position and order.

  The Euclidean distance calculation unit 232 records the Euclidean distance between the constellation position (grid point) and each transmission symbol candidate, and the order of the transmission symbol candidate for each grid point, as will be described below. It may be configured to refer to the prepared database and obtain equivalent information without obtaining the above calculation process.

  Here, the functional configuration of the Euclidean distance calculation unit 232 will be described in more detail with reference to FIG. FIG. 3 is an explanatory diagram illustrating a functional configuration of the Euclidean distance calculation unit 232 according to the present embodiment.

  As illustrated in FIG. 3, the Euclidean distance calculation unit 232 includes a QR decomposition unit 252, a signal reproduction unit 254, a quantization unit 256, a signal point candidate selection unit 258, and a distance information storage unit 260. .

The QR decomposition unit 252 is a unit that decomposes the channel matrix H input from the channel estimation unit 204 into a product of the unitary matrix Q and the upper triangular matrix R. Signal reproducing unit 254, on the received signal vector y received via a plurality of antennas, the action of Hermitian conjugate Q ^H of the unitary matrix Q calculated by the QR decomposition unit 252. In addition, the signal reproduction unit 254 calculates a position x on the constellation corresponding to the received signal using each element of the upper triangular matrix R.

Note that the processing of the signal reproduction unit 254 is sequentially executed in units of rows in order from the lowest row of the upper triangular matrix R. Taking the case where the channel matrix H is 4 × 4 MIMO represented by the following formula (1) as an example, the signal reproduction unit 254 has a position x = {Q ^H y} (3 with respect to the bottom row of the upper triangular matrix R. ) / R ₃₃ is executed. However, {Q ^H y} (k) (k = 0 to 3) represents the k-th element of Q ^H y. Similarly, for other rows, the signal reproduction unit 254 can calculate the position x. Information on the position x is input to the quantization unit 256.

  Next, the quantization unit 256 quantizes the position x calculated by the signal reproduction unit 254. This is shown in FIG. FIG. 4 is a plot of the position x (black circle) output by the signal reproduction unit 254 with respect to the grid set on the constellation. In FIG. 4, a dotted line is an auxiliary line for defining a grid, and an intersection of the dotted lines is a grid. In the example of FIG. 4, the I (In-phase) axis and the Q (Quadrature-phase) axis are each divided into 5-bit discrete values. A white circle represents the position X of the signal point (transmission symbol candidate). The quantization unit 256, when given the position x from the signal reproduction unit 254, selects the grid x 'closest to the position x. The selected grid x ′ is input to the signal point candidate selection unit 258.

  Refer to FIG. 3 again. The signal point candidate selection unit 258 refers to the database stored in the distance information storage unit 260 and selects transmission symbol candidates of a predetermined order corresponding to the grid x ′ input from the quantization unit 256. Signal point candidate selection section 258 also outputs the Euclidean distance for the transmission symbol candidate. However, the distance information storage unit 260 includes, for example, a database in which the Euclidean distance between each grid and each transmission symbol candidate on the constellation and the rank of the transmission symbol candidate for each grid are recorded in association with each other. Stored. Therefore, the signal point candidate selection unit 258 can output the transmission symbol candidate and the Euclidean distance corresponding to the transmission symbol candidate without performing the calculation process when the grid information and the rank information are given.

(Cumulative likelihood value calculation unit 234)
Refer to FIG. 2 again. The cumulative likelihood value calculation unit 234 is a means for calculating the cumulative likelihood value F based on the Euclidean distance value e output from the Euclidean distance calculation unit 232. The cumulative likelihood value F calculated by the cumulative likelihood value calculation unit 234 is input to the likelihood rank determination unit 236. The cumulative likelihood value F is a value defined by the following equation (2). Here, expressions used for the following explanation are collectively defined.

E (n, k): corresponding to the kth rank, nth smallest Euclidean distance E (m, k): mth stage of processing, candidate likelihood value corresponding to the kth rank (M, n, k) = E (m−1, k) + e (n, k) (2)
However, m ≧ 2.

  Note that the k-th rank is the k-th smallest cumulative likelihood in the (m−1) -th processing stage in the m-th processing stage for calculating the cumulative likelihood value F (m, n, k). The order corresponding to the value. Therefore, the ranking value k included in the argument of the cumulative likelihood value F (m, n, k) corresponds to the combination of transmission symbol candidates selected in the (m−1) th processing stage. Further, the m-th stage process means a process corresponding to the m-th line in the transmission symbol candidate selection process executed in order from the lowest line of the upper triangular matrix R. Therefore, it is noted that in the m-th processing stage for calculating the cumulative likelihood value F (m, n, k), combinations of transmission symbol candidates for (m−1) substreams are ranked. I want.

(Likelihood rank determination unit 236)
The likelihood ranking determination unit 236 is a unit that ranks combinations of transmission symbol candidates based on the cumulative likelihood value F calculated by the cumulative likelihood value calculation unit 234. First, the likelihood rank determination unit 236, for example, with respect to the kth rank, the cumulative likelihood value F (m, n, k) (n = 1 to N; N; The number of signal points is ranked and arranged in ascending order. Then, likelihood ranking determination section 236 includes combinations of transmission symbol candidates corresponding to ranked cumulative likelihood values F (m, n, k) including substreams corresponding to Euclidean distance e (n, k). Determine the ranking for. In the following description, “ranking” may be referred to as “ranking”.

(Minimum likelihood value calculation unit 238)
The minimum likelihood value calculation unit 238 is a unit that refers to the cumulative likelihood value F ranked by the likelihood rank determination unit 236 and selects the smallest cumulative likelihood value F. Note that the cumulative likelihood values F are already arranged in the decreasing order in the likelihood rank determining unit 236. Therefore, the minimum likelihood value calculation unit 238 can obtain the minimum likelihood value by extracting the cumulative likelihood value F corresponding to the first rank. The minimum likelihood value selected by the minimum likelihood value calculation unit 238 is input to the signal point candidate setting unit 240 and the next rank likelihood calculation unit 242.

(Signal point candidate setting unit 240)
The signal point candidate setting unit 240 sets the minimum likelihood value F selected by the minimum likelihood value calculation unit 238 as the candidate likelihood value E. In addition, the signal point candidate setting unit 240 sets a combination of transmission symbol candidates corresponding to the minimum likelihood value F to a combination of transmission symbol candidates {x ′} corresponding to the candidate likelihood value E. For example, the signal point candidate setting unit 240 sets the minimum value F (m, n1, k) of the cumulative likelihood values F (m, n, k) calculated in the m-th processing stage for the k-th rank. Is set to the candidate likelihood value E (m, i) = F (m, n1, k). The candidate likelihood value E (m, i) is the i-th rank candidate likelihood value in the m-th processing stage. A method for determining i, k, etc. will be described later with a specific example.

(Next order likelihood calculation unit 242)
The next-rank likelihood calculating unit 242 calculates a cumulative likelihood value that becomes a candidate of the next smallest cumulative likelihood value after the cumulative likelihood value F set to the candidate likelihood value E by the signal point candidate setting unit 240. It is. First, the next-rank likelihood calculation unit 242 notifies the Euclidean distance calculation unit 232 of the rank of the cumulative likelihood value F set to the candidate likelihood value E (k-th rank in the above example), and the candidate likelihood value The Euclidean distance next to the Euclidean distance e included in the cumulative likelihood value F set in E is acquired. For example, when the next-rank likelihood calculation unit 242 calculates the next smallest cumulative likelihood value after the candidate likelihood value E (m, 1) = F (m, n1, k), the information of the rank k is used as the Euclidean distance The calculation unit 232 is notified, and the Euclidean distance e (n2, k) next to the Euclidean distance e (n1, k) (hereinafter, the Euclidean distance of the next rank) is acquired.

  Next, the next order likelihood calculating unit 242 calculates the cumulative likelihood value F using the next order Euclidean distance e. In the case of the above example, the next-rank likelihood calculating unit 242 uses the (m−1) -th candidate likelihood value E (m, k), and the cumulative likelihood value F (m, n2, k) = E. Calculate (m, k) + e (n2, k). The cumulative likelihood value F calculated by the next rank likelihood calculation unit 242 is input to the next rank likelihood position detection unit 244.

(Next-rank likelihood position detection unit 244)
The next-rank likelihood position detection unit 244 has the new cumulative likelihood value F calculated by the next-rank likelihood calculation unit 242 among the cumulative likelihood values F ranked by the likelihood rank determination unit 236. This is a means for detecting the order of the order. Of course, the next-rank likelihood position detection unit 244 may add a new cumulative likelihood value F to all the cumulative likelihood values F calculated by the cumulative likelihood value calculation unit 234 and perform a new ranking. . However, since the cumulative likelihood value F has already been ranked in ascending order by the likelihood rank determination unit 236, it is only necessary to detect the position of a new cumulative likelihood value. This method reduces the amount of calculation required for ranking.

  The position information of the new cumulative likelihood value F detected by the next rank likelihood position detection unit 244 is input to the minimum likelihood value calculation unit 238. Then, the minimum likelihood value calculation unit 238 adds a new cumulative likelihood value F and removes the cumulative likelihood value F already selected as the candidate likelihood value E from the set of cumulative likelihood values F. A new candidate likelihood value E is calculated. The above processing is repeated each time the candidate likelihood value E is set by the signal point candidate setting unit 240 until a predetermined number of candidate likelihood values E are determined.

(Selection method of transmission symbol candidate)
Here, with reference to FIG. 5, a flow of a method for selecting a transmission symbol candidate according to the present embodiment will be briefly described. FIG. 5 is an explanatory diagram showing a method for selecting transmission symbol candidates according to the present embodiment.

  As shown in FIG. 5, the MIMO signal detection unit 202 first performs initial setting (S100). First, cumulative likelihood values F (m, 1, 1) to F (m, N, 1) (N is the number of signal points) are calculated for ranking 1 (S102). Next, the cumulative likelihood values F (m, 1, 1) to F (m, N, 1) are arranged and ranked in ascending order (S104). Next, the cumulative likelihood value F (m, n, 1) having the smallest value is selected and set to the candidate likelihood value E (m, 1) (S106). The initial setting is completed by the above processing.

  Next, processing for the j-th ranking in will be described. First, cumulative likelihood values F (m, 1, in) to F (m, N, in) are calculated (S108). Next, the cumulative likelihood values F (m, 1, 1) to F (m, N, 1) are arranged and ranked in ascending order (S110). Next, the cumulative likelihood value F (m, n, in) having the smallest value is selected and set to the candidate likelihood value E (m, j) (S112). Next, 1 is added to the ranking in, and 1 is added to j, the process proceeds to the next candidate likelihood value selection process (S114). At this time, it is determined whether or not j <N (S116). If j <N, the process proceeds to step S108, and if j ≧ N, the process ends.

(Specific example: Signal point candidate selection at the second stage)
Next, the transmission symbol candidate selection method according to the present embodiment will be described with reference to FIG. 6 with more specific examples. FIG. 6 is an explanatory diagram showing an example of a method for selecting transmission symbol candidates according to the present embodiment. However, the example of FIG. 6 relates to processing for selecting signal point candidates in the second stage.

  First, as a premise, Euclidean distances e (1, 1) to e (1, N) for each transmission symbol candidate are calculated in the first stage, and the candidate likelihood values E (1, 1) in ascending order. -E (1, N) is set. As shown in FIG. 6, in the second stage, Euclidean distances e (2,1) to e (()) based on the candidate likelihood values E (1,1) to E (1, N) calculated in the first stage. 2, N) is calculated and added to calculate the cumulative likelihood value F (2, 1, k) = E (1, k) + e (k, 1) (k = 1 to N) (S132). . Further, the cumulative likelihood values F (2, 1, k) (k = 1 to N) are arranged in ascending order, and the cumulative likelihood value F (2, 1, k1) having the smallest value is the candidate likelihood value E. It is set to (2, k1) (S134). However, FIG. 6 shows an example in which k1 = 2.

  Next, for the rank k1 of the cumulative likelihood value F corresponding to the candidate likelihood value E (2, 2), the next rank Euclidean distance e (2, 2) is acquired, and the previous candidate likelihood corresponding to the rank k1 is obtained. The new cumulative likelihood value F (2,2,2) = E (1,2) + e (2,2) is calculated by adding the value E (1,2) (S136). This new cumulative likelihood value F (2,2,2) is then inserted into the above cumulative likelihood value F (2,1, k ′) (k ′ = 1,3-N) It is detected whether it is positioned in order (S138). Then, the cumulative likelihood value {F (2,1, k ′) (k ′ = 1, 3 to N), F (2, 2, N) including the new cumulative likelihood value F (2, 2, 2) is included. 2)}, the cumulative likelihood value F having the smallest value is selected as the next candidate likelihood value E (2, 2). However, FIG. 6 shows an example of E (2,2) = F (2,1,3).

  By repeatedly executing the above processing at each stage, a combination of transmission symbol candidates corresponding to the minimum likelihood is determined. However, the signal point candidate selection method in the final stage is as follows. This is shown in FIG. FIG. 7 is an explanatory diagram showing an example of a method for selecting transmission symbol candidates according to the present embodiment. In particular, it relates to the final stage processing.

  As shown in FIG. 7, since it is not necessary to calculate the candidate likelihood value E used in the next stage in the final stage process, each of the candidate likelihood values E calculated in the previous stage is calculated. The cumulative likelihood value F is calculated by adding the Euclidean distance e, and the one with the smallest value may be selected. As a result, a combination of transmission symbol candidates corresponding to the minimum likelihood is determined.

  The apparatus and method according to this embodiment have been described above. As described above, by selecting the combinations of transmission symbol candidates based on the ranking information, the number of calculation times of the Euclidean distance required for maximum likelihood detection can be reduced. As a result, the amount of calculation required for maximum likelihood detection can be greatly reduced. In addition, the method according to the present embodiment does not limit the number of signal points selected as transmission symbol candidates, and does not cause a decrease in accuracy as in the technique described in Patent Document 1 above. Therefore, even when compared with the primitive QRD-MLD maximum likelihood detection method, the calculation load is successfully reduced without lowering the signal estimation accuracy.

  More specifically, it is as follows. For example, it is assumed that four types of transmission signals are transmitted from four transmission antennas. Also, consider the case where the modulation method is 16QAM. In this case, when the complete MLD method is employed, the number of times of calculating the Euclidean distance is 16 ^ 4 = 65535. Further, in the QRM-MLD system in which the M algorithm is combined with the QRD-MLD system, if the number of signal point candidates Sm = 16 to be inherited when the stage shifts, 16 + 16 × 16 × 3 = 784 times. In the present embodiment, when the calculation is performed with the same accuracy, 16+ (16 + 16) × 2 + 16 = 96 calculations are sufficient. That is, the calculation amount is reduced to about 1/8.

  In the method described in Patent Document 1, more signal point candidate information must be held at each stage in order to obtain the same calculation accuracy as that of the present embodiment. Therefore, a large amount of memory is consumed compared to the present embodiment. In view of these points, by applying the technique according to the present embodiment, it is possible to obtain an effect of reducing the amount of calculation, suppressing the amount of memory, and reducing the circuit scale as a result.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to the example which concerns. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

  For example, although QR decomposition is assumed throughout the description of the above embodiment, QL decomposition that decomposes the channel matrix H into a product of the unitary matrix Q and the lower triangular matrix L may be used. These processes are substantially the same and, of course, are included in the technical scope of the present invention.

It is explanatory drawing which shows the structure of the communication system which concerns on a MIMO transmission system. It is explanatory drawing which shows the structure of the signal separation means which concerns on one Embodiment of this invention. It is explanatory drawing which shows the structure of the Euclidean distance acquisition means which concerns on the same embodiment. It is explanatory drawing which shows the quantization method and ranking method of a received symbol. It is explanatory drawing which shows the flow of the signal detection process which concerns on the same embodiment. It is explanatory drawing which shows the selection method of the signal point candidate which concerns on the same embodiment. It is explanatory drawing which shows the selection method of the signal point candidate which concerns on the same embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Transmission apparatus 102 Error correction encoding part 104 Modulation part 200 Reception apparatus 202 MIMO signal detection part 204 Channel estimation part 206 Error correction decoding part 232 Euclidean distance calculation part 234 Cumulative likelihood value calculation part 236 Likelihood rank determination part 238 Minimum likelihood Frequency value calculation unit 240 Signal point candidate setting unit 242 Order rank likelihood calculation unit 244 Order rank likelihood position detection unit 252 QR decomposition unit 254 Signal reproduction unit 256 Quantization unit 258 Signal point candidate selection unit 260 Distance information storage unit

Claims (3)

A receiver that separates signals by maximum likelihood detection based on QR decomposition or QL decomposition,
A likelihood value calculator for calculating a likelihood value for the combination of signal points;
A rank determining unit that ranks the likelihood values in ascending order of values;
A minimum likelihood value selector for selecting a minimum likelihood value from the likelihood values;
With
The likelihood value calculation unit further calculates a new likelihood value that is the next rank of the minimum likelihood value based on the rank of the likelihood value,
The minimum likelihood value selection unit determines a minimum likelihood value that is minimum by removing the minimum likelihood value from the likelihood values including the new likelihood value,
A receiving apparatus, wherein a combination of signal points corresponding to each minimum likelihood value is selected as a transmission symbol candidate at each stage. For each of the predetermined grids set on the signal point arrangement diagram, a storage unit in which information on the Euclidean distance to each signal point is recorded,
A quantization unit that quantizes received symbols in association with the predetermined grid;
Further comprising
The likelihood value calculation unit reads information on the Euclidean distance from the storage unit based on grid information corresponding to the received symbol, and calculates a likelihood value for the combination of signal points based on the information on the Euclidean distance. The receiving device according to claim 1, wherein the receiving device calculates. A signal selection method used for maximum likelihood detection based on QR decomposition or QL decomposition,
A likelihood value calculating step in which likelihood values for combinations of transmission symbol candidates are calculated;
A rank determining step in which the likelihood values are ranked in ascending order of value;
A minimum likelihood value selecting step in which a minimum likelihood value is selected from the likelihood values;
A next-rank likelihood value calculating step in which a new likelihood value that is the next rank of the minimum likelihood value is calculated based on the rank of the likelihood value;
A next-order minimum likelihood value selecting step in which a minimum likelihood value that is the minimum by removing the minimum likelihood value from the likelihood values including the new likelihood value is determined;
Including
Likelihood that the minimum likelihood value obtained by the minimum likelihood value selecting step, the next-rank likelihood value calculating step and the next-rank likelihood value selecting step repeatedly executed corresponds to a combination of transmission symbol candidates. A signal selection method characterized by being selected as a value.

Images