WO2014090121A1

WO2014090121A1 - Method and apparatus for detecting signal

Info

Publication number: WO2014090121A1
Application number: PCT/CN2013/088818
Authority: WO
Inventors: 吴凯
Original assignee: 电信科学技术研究院
Priority date: 2012-12-10
Filing date: 2013-12-06
Publication date: 2014-06-19
Also published as: CN103873203A; CN103873203B

Abstract

Embodiments of the present application provide a method and an apparatus for detecting a signal, and relate to wireless communications technologies. In a process of detecting a signal received by an N-layer space multiplexing transmission system, when a (N-1)^th layer to a first layer is detected, judgment signals are first determined after removing interference of last determined signal vectors from symbols received by the channel space and removing effects of equivalent channel gain of the channel space, and then to-be selected symbols most similar to the judgment symbols are determined. Then, signal vectors most similar to signal vectors obtained by removing effects of equivalent channel gain of the corresponding channel space from signal vectors received by a current channel space are determined from the signal vectors determined according to the to-be-selected symbols, and reserved signal vectors on the layer are determined. Therefore, it is only necessary to filter the to-be-selected symbols determined according to the judgment symbols, and it is unnecessary to filter all symbols on the layer, there reducing the complexity of signal detection.

Description

The present invention claims the priority of the Chinese patent application filed on Dec. 10, 2012, the Chinese Patent Application No. 201210529634.7, entitled "A Method and Apparatus for Signal Detection", The entire contents are incorporated herein by reference.

Technical field

The present invention relates to wireless communication technologies, and in particular, to a signal detection method and apparatus.

Background technique

In the multi-antenna system of spatial multiplexing transmission, a signal detection method is a QR maximum-like (QRD-M) detection algorithm that retains M branches based on QR decomposition, and the detection algorithm is a kind of processing. Maximum likelihood detection algorithm, which uses a tree search method to select, for each layer of channel space, a plurality of constellation symbols that are most likely to transmit signals to the transmitting end of the layer in the tree search process, and The combination of the constellation symbols most likely to be the signal vector transmitted by the transmitting end is selected. The combination of the constellation symbols is a vector, which is called a branch in the detection algorithm, and the combination of the constellation symbols is used as the detection result of the output.

Specifically, it is assumed that the transmission model of the Multiple Input Multiple Output (MIMO) system for spatial multiplexing transmission is as follows:

r = Hx + n

Where r is the received signal vector of W _R xl, H is the channel matrix of N _R xN, X represents the transmitted signal vector of ^ xl, and n represents the noise vector of W _R xl.

The QRD-M algorithm is usually implemented by tree search. The tree search can be summarized as follows: Since the search of the signal transmitted by the channel space is not affected by other channel space, the search is performed from the layer, and the layer is searched. Calculating branch metrics of all constellation points in the layer of the tree structure, and the branch metric is specifically a symbol for calculating the influence of each constellation point and the symbol received by the channel space of the layer to remove the equivalent channel gain of the channel space of the layer The Euclidean distance is squared. The search for other layers is based on the reserved branch of the previously searched symbol. The traversal of all constellation points is used to calculate the cumulative branch metric, and the M branches with the smallest cumulative score metric are retained in all constellation points. The reserved branch of the next layer of search.

The algorithm execution steps are as follows:

The first step: First, the channel matrix needs to be layered, QR decomposition of the channel matrix H, QR decomposition is the triangular matrix decomposition on the unitary matrix, which is to decompose the matrix into a normalized orthogonal matrix, ie, the matrix Q and The triangle matrix R, so called the QR decomposition method. A QR decomposition output Q matrix and an R matrix are performed on the channel matrix H. The Q matrix is a unitary matrix, and the R matrix is an upper triangular equivalent channel gain matrix, as shown in the following equation.

The initialization Χ '^ is an all-zero matrix of N _L xM, and bm is a vector of 1 xM.

Layer the received vector and multiply the received signal vector r by Q ^ff to obtain the equivalent received vector y = [; " v _z f . Step 2: As shown in Figure 1, proceed from layer 2 to layer 1. Search layer by layer to detect the signal vector sent by the sender.

Suppose that the number of constellation points corresponding to the modulation method used by the search layer is Q, and all Q constellation points are respectively brought into d =\y _N[ -r _N[tN[ x _q I ² calculates the branch metric, where y _Ni is the Nth row element in the equivalent receiving vector, and is the element of the Nth row and the Nth column in the upper triangular equivalent channel gain matrix. For the constellation point, the M metrics are the smallest among the Q branch metrics. Save bm and store the corresponding constellation point in the first line of Χ'^.

For the search of the symbols of the Λ^- + 1 layer, it is necessary to calculate the cumulative calculation of all the constellation points under each reserved branch.

Where is the number of layers in the channel space of the layer, where is the element of the i-th row in the equivalent received vector, r is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and _riJ is the upper triangular equivalent channel The element of the i-th row and the i-th column of the gain matrix, representing the first row and the first column element of X ^fei . Indicates the constellation point of the current layer, from the constellation set. The cumulative branch metric is added by the increment of the branch metric and the corresponding element of the saved cumulative branch metric bm, BM (( _m -\)xQ _{+ q} ) = bm(m) + d _im — _{l> Q+q} , where q is between 1 and Q, then a total of QxM branch metrics need to be calculated.

Each layer keeps the smallest M BM corresponding branches in M乂Q branches, stores the corresponding accumulated branch metrics into bm, and writes the corresponding searched layer reserved branches and constellation points reserved by the current layer to X' ^.

Finally, the bm output after all symbol searches are completed is the branch metric corresponding to the remaining M smallest branches, and the reserved branch Χ ^1φ contains the M branches corresponding to bm.

Step 3: In the uncoded system, the receiver converts the vector corresponding to the BM minimum to the bit output as the detection result. However, in a coded system, it is necessary to generate a log likelihood ratio (also called soft bit) of each bit input to the decoder. The process of generating a log-likelihood ratio is to use the difference between the minimum cumulative branch metric of the bit in the reserved branch and the minimum cumulative branch metric of the bit in the reserved branch as the output of the soft bit. As shown in the following formula:

b ^{n,) (n _h )^ min (BM) - min (BM) In the above formula, ^τ. The first layer represents a branch reserved "branch of the collection _¾ bits ο, x""indicates the first branch retention layer" _¾ bits set to 1 of the branch, ", ε {1, ··· , Λ ^}, ¾e {l,---,log ₂ o indicates the layer of transmission The number, 1 0 _§ 2 0 represents the modulation order.

The QRD-M detection algorithm has two shortcomings: First, in order to ensure the performance of the detection algorithm, the preset M is relatively large, and under each reserved branch, the branch metric under all constellation point hypotheses needs to be calculated. Resulting in a large number of branch metrics in the tree search; Second, in order to select the smallest M branch metrics and corresponding branch reservations, after each layer completes the calculation of the score metrics, it needs to be multiple The branch metrics are sorted. If the modulation order is high and M is large, the number of branch metrics is large, so the delay of sorting is relatively large.

In the 64QAM modulation mode, M = 16 is taken as an example. For the first layer search, the QRD-M algorithm needs to calculate 64 branch metrics, and needs 64 sorts and 16 sorts. For non-first layer search, the QRD-M algorithm needs to calculate 64x16 branch metrics. The QRD-M algorithm requires a 64x16-select 16 sort. The calculation and sorting of these branch metrics requires a lot of computing resources and brings a large processing delay.

In summary, the existing maximum likelihood detection algorithm has higher complexity and occupies more resources.

Summary of the invention

Embodiments of the present invention provide a signal detection method and apparatus to reduce the complexity of signal detection.

A signal detection method includes:

The receiving end determines an equivalent receiving vector according to a signal sequence received through the N-layer spatial multiplexing transmission system and a unitary matrix obtained by channel matrix QR decomposition;

The receiving end determines, in the symbol corresponding to the channel space of the Nth layer, the equivalent channel of the corresponding layer in the equivalent receiving vector to remove the channel space of the layer according to the upper triangular equivalent channel gain matrix obtained by the channel matrix QR decomposition. The symbol obtained after the influence of the gain is the closest to the set number of symbols, and each symbol determined is taken as a signal vector;

For each layer of channel space in the N-1th channel space to the layer 1 channel space, the receiving end determines the equivalent receiving vector for each of the most recently determined signal vectors, the upper triangular equivalent channel gain matrix The element of the layer removes the decision symbol of the interference of the signal vector and the equivalent channel gain of the channel space of the layer, and determines the set number of each decision symbol in the symbol corresponding to the channel space of the layer. The closest candidate symbol, and based on all the candidate symbols, the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix, are determined to be composed of elements corresponding to the Nth layer to the current layer in the equivalent receiving vector. The signal vector removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector is the closest to the set number of signal vectors;

The signal detection result is determined based on the signal vector determined in the layer 1 channel space.

A signal detecting device includes:

a first determining unit, configured to determine an equivalent receiving vector according to a signal sequence received through the N-layer spatial multiplexing transmission system and a unitary matrix obtained by QR matrix QR decomposition;

a second determining unit, configured to: in a symbol corresponding to the channel space of the Nth layer, an upper triangular equivalent channel gain matrix obtained by channel matrix QR decomposition, determining to remove the layer channel from an element corresponding to the corresponding receiving vector Space After the influence of the equivalent channel gain, the symbols obtained are the closest to the set number of symbols, and each symbol determined is taken as a signal vector;

a third determining unit, configured to: for each channel space in the Nth layer channel space to the layer 1 channel space, the receiving end pairs the last determined each signal vector, and the triangular equivalent channel gain on the signal vector The matrix determines a decision symbol in the equivalent receiving vector that removes the influence of the signal vector on the interference of the signal vector and the equivalent channel gain of the channel space of the layer, and corresponds to each decision symbol in the channel space of the layer. The symbol is determined to be the closest candidate symbol to be selected, and all the candidate symbols are selected, the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix are determined, and the corresponding correspondence vector is determined. The signal vector composed of the symbols of the Nth layer to the current layer removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector is the closest to the set number of signal vectors;

And a fourth determining unit, configured to determine a signal detection result according to the signal vector determined in the layer 1 channel space. The embodiment of the invention provides a signal detection method and device. When detecting a signal received by an N-layer spatial multiplexing transmission system, when detecting the N-1 layer to the first layer, the channel is first determined. The spatially received symbol removes the decision symbol of the interference of the most recently determined signal vector and the equivalent channel gain of the layer channel space, and further determines the candidate symbols closest to the decision symbols, and then determines according to these Among the signal vectors determined by the candidate symbols, which are the closest to the signal vector obtained by removing the equivalent channel gain of the corresponding channel space in the signal space received by the current channel space, the reserved signal vector of the layer can be determined, Only the candidate symbols determined according to the decision symbols need to be filtered, and it is not necessary to filter all possible symbols of the layer, which reduces the complexity of signal detection.

DRAWINGS

1 is a schematic diagram of a QRD-M algorithm tree search provided by the prior art;

2 is a flowchart of a signal detection method according to an embodiment of the present invention;

3 is a flowchart of a method for detecting an Nth layer signal according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a constellation area division according to an embodiment of the present invention;

FIG. 5 is a flowchart of a method for determining a reserved signal vector according to an embodiment of the present invention;

6 is a second flowchart of a signal detection method according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of a signal detecting apparatus according to an embodiment of the present invention.

Detailed ways

The embodiment of the invention provides a signal detection method and device. When detecting a signal received by an N-layer spatial multiplexing transmission system, when detecting the N-1 layer to the first layer, the channel is first determined. The spatially received symbol removes the decision symbol of the interference of the most recently determined signal vector and the equivalent channel gain of the layer channel space, and further determines the candidate symbols closest to the decision symbols, and then determines according to these Among the signal vectors determined by the candidate symbols, which are the closest to the signal vector obtained by removing the equivalent channel gain of the corresponding channel space in the signal space received by the current channel space, the reserved signal vector of the layer can be determined, , only need to judge The candidate symbols determined by the decision symbol are filtered, and it is not necessary to filter all possible symbols of the layer, which reduces the complexity of signal detection.

As shown in FIG. 2, an embodiment of the present invention provides a signal detection method, including:

5101. The receiving end determines an equivalent receiving vector according to a signal sequence received by the N-layer spatial multiplexing transmission system and a unitary matrix obtained by channel matrix QR decomposition;

5102. The receiving end determines, in the symbol corresponding to the channel space of the Nth layer, the element corresponding to the corresponding layer in the equivalent receiving vector, according to the upper triangular equivalent channel gain matrix obtained by the channel matrix QR decomposition. The symbol obtained by the effect of the channel gain is the closest to the symbol of the set number, and each of the determined symbols is used as a signal vector;

S103. For each layer channel space in the N-1 layer channel space to the layer 1 channel space, the receiving end determines the equivalent receiving vector for each of the most recently determined signal vectors and the upper triangular equivalent channel gain matrix. The decision symbol corresponding to the element of the layer is removed from the interference of the signal vector and the equivalent channel gain of the channel space of the layer, and each decision symbol is determined and set in the symbol corresponding to the channel space of the layer. The number of the candidate symbols closest to the number, and according to all the candidate symbols, determining the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix, determining the corresponding Nth layer to the current layer in the equivalent receiving vector The signal vector composed of elements removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector is the closest to the set number of signal vectors;

S104. Determine a signal detection result according to the signal vector determined in the layer 1 channel space.

Passing the signal space of each layer in the channel space of the N-1th layer to the channel space of the layer 1 channel, the receiving end determines each symbol vector received in the channel space of the layer to remove the signal vector. The decision symbol after the influence of the interference and the equivalent channel gain of the channel space of the layer, and the candidate symbol close to the decision symbol determine the signal vector of the current layer, which reduces the complexity of signal detection.

After receiving the signal sequence at the receiving end, the signal sequence needs to be processed to implement the layered detection in the embodiment of the present invention. Therefore, the channel matrix is QR decomposed and decomposed into a unitary matrix and an upper triangular equivalent channel gain. a matrix R, where N is the number of layers in the channel space and the received signal

The sequence is left-multiplied by the transposed matrix of the matrix to achieve layering of the received signal sequence, resulting in an equivalent received vector y = [n ] ^r .

Preferably, in order to facilitate the implementation, the number of signal vectors determined by each layer of the channel space may be set to be the same. Of course, those skilled in the art may set other feasible methods according to actual conditions. In the embodiment of the present invention, the number of signal vectors determined by the channel space of each layer is the same as an example.

Preferably, the degree of similarity between the two signals can be determined by the square of the Euclidean distance between the two signals, The smaller the square of the Euclidean distance between the signals, the greater the degree of similarity between the two signals. Therefore, the effect of the equivalent received vector in S102 on the equivalent channel gain of the layer's channel space is removed in S102. The symbols with the closest number of symbols obtained can be:

And a symbol of a set number of the equivalent reception vector obtained after the symbol of the layer removes the equivalent channel gain of the layer channel space, and the symbol Euclidean distance squared is the smallest;

In S103, each of the decision symbols is determined in the symbol corresponding to the channel space of the layer, and the set number of the candidate symbols closest to the same is determined, which may be:

Determining, in each of the symbols corresponding to the channel space of the layer, a set number of candidate symbols having the smallest square of the Euclidean distance;

The signal vector in S103 determines a signal with the closest set of signal vectors obtained by removing the influence of the equivalent channel gain of the corresponding channel space by the signal vector composed of the symbols of the Nth layer to the current layer of the equivalent received vector. Vector, can be:

Determining, from a candidate signal vector consisting of the candidate symbol and the signal vector determining the candidate symbol, a set number of signal vectors, the set number of signal vectors being transmitted through the channel space of the Nth layer to the current layer The obtained signal vector and the equivalent Euclidean distance squared of the signal vector composed of the elements corresponding to the Nth layer to the current layer in the equivalent receiving vector are the smallest, and the cumulative Euclidean distance square is specifically between each corresponding symbol of the two signal vectors. The sum of the squared Euclidean distances.

Since the signal transmitted in the Nth channel space is not disturbed by the signal transmitted in other channel space, the elements of the Nth row and the Nth column of the upper triangular equivalent channel gain matrix obtained by QR decomposition are not zero. The other elements in the Nth row are all zero, and the signals transmitted in the N-1th to the 1st channel are subject to the interference of the signals transmitted by other signals, and the corresponding N-1th layer in the upper triangular equivalent channel gain matrix. The number of elements that are not zero in each row of the layer 1 channel space is greater than 1. Therefore, the method for detecting the symbols of the channel space of the Nth layer and the method for detecting the symbols of other channel spaces in the embodiment of the present invention Different, and need to start detection from the Nth layer channel space.

Preferably, when detecting the Nth layer, determining the influence of the symbol and the equivalent vector of the signal vector corresponding to the element in the signal vector to remove the equivalent channel gain of the channel space of the layer, the square of the Euclidean distance obtained is d. _Lq =\ y _NN x _q \ where ^ is the Nth row element in the equivalent received vector, r _w ^ is the element of the Nth row and the Nth column in the upper triangular equivalent channel gain matrix, X is the symbol, specifically, As shown in FIG. 3, in S102, the receiving end is in the symbol corresponding to the Nth channel space, and the upper triangular equivalent channel gain matrix obtained by channel matrix decomposition is used to determine the element removal of the corresponding layer in the equivalent vector of the signal vector. The symbols of the most similarly set symbols obtained after the influence of the equivalent channel gain of the layer channel space specifically include:

S1021: The receiving end determines, according to each symbol in the symbol corresponding to the layer channel space, the Euclidean symbol obtained by the symbol and the equivalent receiving vector after the symbol of the layer removes the equivalent channel gain of the layer channel space. Distance square

^ q =1 yN ~ ^r N, N ^X q I ;

S1022: Determine a symbol of a set number of the smallest Euclidean distance squared. Calculating the square of the Euclidean distance between the symbols obtained by subtracting the influence of the equivalent channel gain on the symbol of the layer corresponding to the equivalent received signal by the formula =\ y _N - _N x _q I ² , and then By selecting the minimum number of set symbols, you can intuitively determine the symbols to be retained in the Nth layer. Certainly, those skilled in the art may use other feasible methods to determine the number of symbols that are closest to the symbol obtained by removing the equivalent channel gain of the channel space of the layer channel space. It is no longer described one by one.

In the embodiment of the present invention, the signal corresponding to the signal received by the receiving end is determined according to the signal received by the receiving end and the equivalent channel gain, and is determined from the signal that may be sent by the transmitting end. The signal is closer to the signal, thereby reducing the complexity of the detection. Since the signal transmitted by the channel space will interfere with the signal transmitted by the upper layer channel space, the relationship between the signal received by the receiving end of each layer and the signal sent by the transmitting end _Let y _t = r _u Xj. _m + r _i x _m , where is the number of the channel space of the layer, e {1, · · · , Μ} , Μ is the last time determined = '■ +1 ''

The number of signal vectors, r is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and is the element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix, and X _m is the mth The symbol of the jth row in the channel sequence is the element of the i-th row in the signal vector equivalent reception vector.

Therefore, when detecting each layer of channel space, the interference caused by the transmission of the layer signal according to the element corresponding to the layer in the equivalent reception vector and the most recently determined signal vector, and the equivalent channel of the current layer can be obtained. Gain to determine the decision symbol, specifically, the decision symbol is ^^二-

After determining the decision symbol, a predetermined number of candidate symbols are determined for each decision symbol. Since the Euclidean distance of the candidate symbol and the decision symbol is small, the degree of similarity is high, and the signal sent by the transmitting end is also close to each other. Therefore, the signal vector of the current layer can be determined according to the candidate symbol and the most recently determined signal vector, and the complexity of signal detection is reduced.

An embodiment of the present invention provides a method for determining a candidate symbol, which specifically includes:

For each decision symbol, determining a region to which the decision symbol belongs on the constellation diagram, and mapping the region to a symbol corresponding to the channel space of the layer, and determining that the symbol corresponding to the region is a candidate symbol, the region and the layer The mapping relationship of the symbols corresponding to the channel space is specifically: determining a constellation point of the symbol corresponding to the channel space of the layer, and determining a corresponding area of the constellation point for the set number of constellation points whose plane distance is closest to each group And determining a mapping relationship between a symbol corresponding to each set of constellation points and a corresponding region of the set of constellation points, where a sum of plane distances of any constellation points in the corresponding region and the set of constellation points is smaller than that of other group constellation points The sum of the plane distances.

Specifically, for the Long Term Evolution (LTE) mode in which four layers of spatial multiplexing transmission are used, the detection in the 16QAM modulation mode is taken as an example, as shown in FIG. 4, the plane distance between each group is two. The nearest four constellation points determine the corresponding regions of the set of constellation points, and the mapping relationship between the corresponding symbols of each set of constellation points and the corresponding regions of the set of constellation points is as shown in Table 1. Stored in Table 1 is the number of the constellation points in Figure 4.

Regional candidate list 1, 13, 14, 15, 16

2 9,11,13,15

3 9,10,11,12

4 5,6,13,14

5 1,5,9,13

6 1,2,9,10

7 5,6,7,8

8 1,3,5,7

9 1,2,3,4

Table 116QAM area lookup table

When the candidate symbol is determined, if the decision symbol is located in area 1, the candidate symbols are 13, 14, 15, and 16. If the decision symbol is located in area 3, the candidate symbols are 9, 10, 11, and 12. The complexity of the detection.

After the candidate symbol is determined, the signal to be selected and the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix are determined to determine the signal vector to be reserved in the channel space of the layer. In the embodiment of the present invention, first, determining an Euclidean distance squared increment of each candidate symbol and a signal vector for determining the signal vector of the candidate symbol, and then increasing the Euclidean distance squared increment and determining the candidate symbol. The cumulative Euclidean distance square of the signal vector is added to obtain the cumulative Euclidean distance square of the candidate signal vector. The smaller the cumulative Euclidean distance square is, the corresponding Nth layer to the current layer in the candidate signal vector and the equivalent receiving vector. The higher the degree of closeness of the signal vector composed of the elements, the more the candidate signal vector with the smallest cumulative Euclidean distance squared is set as the signal vector to be reserved. In practical applications, as shown in FIG. 5, all the candidate symbols are selected in S103, the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix are determined, and the corresponding Nth layer to the current layer in the equivalent receiving vector is determined. The signal vector composed of the elements removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector has the closest set number of signal vectors, including:

S501. Determine, for each candidate symbol, a candidate signal vector that is formed by the candidate symbol and a signal vector that determines the candidate symbol.

S502: Determine a symbol of a signal that should be selected in the signal vector after the candidate signal vector passes through the Nth channel space to the current channel space.

S503. Determine an Euclidean distance squared increment of the candidate signal vector ^)^^^-

Where _m,. For the a-th candidate symbol determined by the mth decision symbol in the i-th channel space, the i-th channel space is the current channel space, ∞εα···, Μ}, and M is the channel space of the previous layer. The number of signal vectors, ae {1,···,^} , A is the number of candidate symbols determined according to each decision symbol set by the current layer channel space, X _m To determine the signal vector of the candidate symbol, which is the element of the ith row in the equivalent received vector, r is the element of the i-th row and the j-th column in the equivalent channel matrix, which is the i-th in the upper triangular equivalent channel gain matrix. The elements of row i;

5504. For each candidate symbol, determine a cumulative Euclidean distance square BM(( -\)xA + a) = bm(m) + d _im _ _1)xa+a , where bm( ) is at the i+ The cumulative Euclidean distance squared of the mth signal vector determined by the layer 1 channel space, BM(( -l)x + is the cumulative (-l)x + "accumulated channel vector of the i-th channel space" Square of distance;

5505. Determine, from the candidate signal vector, a signal vector with a minimum number of accumulated Euclidean distance squares.

Preferably, in the uncoded system, the receiving end converts the set number of signal vectors determined by the first layer of channel space into bit outputs, that is, the detection result. However, in the coded system, step S104 can be implemented by soft bit value calculation. Of course, those skilled in the art can use other feasible methods to set the detection result, which is not described here.

Preferably, the signal detection method provided by the embodiment of the present invention can be implemented in the form of a tree search. At this time, the square of the Euclidean distance determined when detecting the Nth layer is the branch metric value in the tree search. The cumulative Euclidean distance square determined by the N-1th to the first layer is the cumulative branch metric, and the Euclidean distance squared increment is the cumulative branch metric increment. Specifically, as shown in FIG. The embodiment of the invention provides a specific method for signal detection, including:

S601. Determine, by performing QR decomposition on the channel matrix, a unit matrix of the channel matrix and an upper triangular equivalent channel gain matrix R, where N is a layer number of the channel space;

5602. Determine the equivalent receiving vector y = [; ·· ] ^Τ by multiplying the received signal vector by the transposed matrix of the 酉 matrix, and set the output matrix Χ ^3⁄4 Λ the output matrix is Ν*Μ matrix, at initial setting For an empty matrix, set a vector with bm of 1*Μ;

5603. Searching for the Nth layer symbol, the receiving end determines a symbol set of the symbol that the sending end can send at the Nth layer, and determining, for each symbol in the set, a branch metric value of the symbol=|^_/^ _3⁄4 | ² , where ^ is the Nth row element in the equivalent receiving vector, and _Ν is the element of the Nth row and the Nth column in the upper triangular equivalent channel gain matrix, where X is the symbol, and the branch metric is used as the cumulative branch metric Saved in bm;

S604. Determine M symbols with the smallest branch metric as the M reserved symbols of the symbol sent by the sending end at the Nth layer, save the reserved symbols in the output matrix, and keep the branch metric values corresponding to the reserved symbols in bm;

S605. Search for the N-1th layer to the 1st layer symbol, reserve M branches in the output matrix, and determine M decision symbols, _m -

, where is the number of layers in the channel space of the layer, r is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and _riJ is the element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix, The element of the mth column in the mth row of the output matrix, mG {1,···, } , is the element of the i-th row in the equivalent receiving vector Prime

S606, determining a set of all symbols that the sender can send at the layer;

5607. For each of the decision symbols, the mapping relationship between the area and the symbol determines the A symbols in the set that are the smallest squared with the Euclidean distance of the decision, and are the A candidate symbols of the symbols sent by the transmitting end at the layer;

S608. Determine, according to the M*A candidate symbols and the reserved branch corresponding to each candidate symbol, an M*A branch metric increment of _{1 Μ+} . =^- , where is the number of layers in the channel space of this layer, which is the equivalent receiving direction

The element of the i-th row in the quantity, r is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and _riJ is the element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix, which is in the output matrix The element of the mth column of the i-th row, me {l,---, }, _m ,. The a-th candidate to be selected according to the mth decision amount in the i-th channel, ae {1,····,^};

S609. Determine, according to the M*A branch metric increments, M*A cumulative branch metrics BM((m - ΐ) χΑ + α) = bm(m) + d _m _ _l)xA+a , where, bm ( ) is the cumulative branch metric of the mth branch determined on the i+1th channel, BM((m - 1)χ^ + β) is the (-1)X + in the i-th channel space The cumulative Euclidean distance squared of the candidate channel vectors;

S610. Determine M minimum cumulative branch metric values, and determine corresponding branches, and use the corresponding branch as a reserved branch to be stored in the output matrix, and store the cumulative branch metric corresponding to the reserved branch into bm;

S611. Determine a signal detection result according to the M reserved branches determined in the layer 1 channel space.

As shown in FIG. 7, an embodiment of the present invention provides a signal detecting apparatus, including:

a first determining unit 701, configured to determine an equivalent receiving vector according to a signal sequence received by the N-layer spatial multiplexing transmission system and a unitary matrix obtained by channel matrix QR decomposition;

The second determining unit 702 is configured to: in the symbol corresponding to the Nth channel space, the upper triangular equivalent channel gain matrix obtained by the channel matrix QR decomposition, and determine the element corresponding to the corresponding layer in the equivalent receiving vector to remove the layer The symbol obtained by the equivalent channel gain of the channel space is the closest to the symbol, and the determined symbol is used as a signal vector;

The third determining unit 703 is configured to determine, for each of the most recently determined signal vectors, the upper triangular equivalent channel gain matrix, etc., for each of the Nth layer channel space to the layer 1 channel space. In the effect receiving vector, the element corresponding to the layer removes the influence symbol of the signal vector and the equivalent channel gain of the channel space of the layer, and each decision symbol is in the symbol corresponding to the channel space of the layer. Determining the number of selected symbols to be closest to each other, and occupying all the candidate symbols, determining the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix, and determining the corresponding Nth layer in the equivalent receiving vector The signal vector composed of the elements of the current layer removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector is the closest to the set number of signal vectors;

The fourth determining unit 704 is configured to determine a signal detection result according to the signal vector determined in the layer 1 channel space. By using the channel space of the N1 layer to the channel space of the layer 1 channel space, the receiving end determines the symbol received by the layer channel space for each signal vector that is determined last time, and removes the interference of the signal vector. And the decision symbol after the influence of the equivalent channel gain of the layer channel space, and the candidate symbol close to the decision symbol determines the signal vector of the current layer, which reduces the complexity of signal detection.

After receiving the signal sequence on the receiving end, the signal sequence needs to be processed to implement the layered detection in the embodiment of the present invention. Therefore, the first determining unit 701 first needs to perform QR decomposition on the channel matrix, and decompose it into a matrix and on the matrix. Triangular equivalent channel gain matrix R: where N is the number of layers in the channel space and will be received

The signal sequence is left-multiplied by the transposed matrix of the matrix to achieve layering of the received signal sequence, resulting in an equivalent received vector y = [n ] ^r .

Preferably, the degree of closeness between the two signals can be determined by the square of the Euclidean distance between the two signals. The smaller the square of the Euclidean distance between the two signals, the greater the degree of similarity between the two signals. Therefore, the symbol of the set number that is closest to the symbol obtained by the equivalent received vector after the symbol of the layer is removed from the equivalent channel gain of the layer channel space may be:

For each decision symbol, the set number of the candidate symbol closest to the set is determined in the symbol corresponding to the channel space of the layer, which may be:

Determining a signal vector that is the closest to the signal vector obtained by removing the influence of the equivalent channel gain of the corresponding channel space by the signal vector composed of the symbol of the Nth layer to the current layer of the equivalent reception vector, which may be :

Since the signal transmitted in the Nth channel space is not disturbed by the signal transmitted in other channel space, the elements of the Nth row and the Nth column of the upper triangular equivalent channel gain matrix obtained by QR decomposition are not zero. , No. N The other elements of the row are all zero, and the signals transmitted by the N-th layer to the layer 1 channel space are interfered by the signals transmitted by other signals, and the corresponding N-th layer to the first in the upper triangular equivalent channel gain matrix The number of elements that are not zero in each row of the layer channel space is greater than one. Therefore, the method for detecting the symbol of the channel space of the Nth layer in the embodiment of the present invention is different from the method for detecting the symbol of other channel spaces, and It is necessary to start detection from the Nth layer channel space.

Preferably, when detecting the Nth layer, determining the influence of the symbol and the equivalent vector of the signal vector corresponding to the element in the signal vector to remove the equivalent channel gain of the channel space of the layer, the square of the Euclidean distance obtained is d. _Lq =\ y _N - r _N , _N x _q \ where ^ is the element of the Nth row in the equivalent received vector, r _w ^ is the element of the Nth row and the Nth column of the upper triangular equivalent channel gain matrix, X is the The symbol, specifically, in the symbol corresponding to the Nth channel space in the second determining unit 702, the upper triangular equivalent channel gain matrix obtained by channel matrix decomposition is determined, and the corresponding layer in the equivalent received vector of the signal vector is determined. After the element removes the influence of the equivalent channel gain of the channel space of the layer, the symbols with the closest set number of symbols are specifically included:

Determining the Euclidean distance d _dq of the symbol obtained by the symbol and the equivalent received vector after the symbol of the layer removes the equivalent channel gain of the layer channel space for each symbol in the symbol corresponding to the layer channel space. =\ y _N - r _N , _N x _q \ where ^ is the element of the Nth row in the equivalent received vector, r _w ^ is the element of the Nth row and the Nth column of the triangular equivalent channel gain matrix on the signal vector, X is The symbol;

Determine the sign of the set number of the smallest Euclidean distance squared.

Calculate the Euclidean distance between the symbols obtained by subtracting the equivalent channel gain of the symbol of the layer from the equivalent received signal of the symbol and the signal vector by the formula ^^ ^ - ^, ^^^ | ² Squared, and then select the minimum number of set symbols, you can intuitively determine the symbols to be retained by the third layer. Certainly, those skilled in the art may use other feasible methods to determine the number of symbols that are closest to the symbol obtained by removing the equivalent channel gain of the channel space of the layer channel space. It is no longer described one by one.

In the embodiment of the present invention, the signal corresponding to the signal received by the receiving end is determined according to the signal received by the receiving end and the equivalent channel gain, and is determined from the signal that may be sent by the transmitting end. The signal is closer to the signal, thereby reducing the complexity of the detection. Since the signal transmitted by the channel space will interfere with the signal transmitted by the upper layer channel space, the relationship between the signal received by the receiving end of each layer and the signal sent by the transmitting end For = ∑ ^rx , TM where is the number of the channel space for this layer, ∞ ε {1, · · · , Μ} , Μ is the last determined = '■ +1 ''

The number of signal vectors is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and is the element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix, which is the m-th channel sequence. The symbol of the jth line is the element of the i-th row in the signal vector equivalent reception vector.

Therefore, when detecting each layer of channel space, the interference caused by the transmission of the layer signal according to the element corresponding to the layer in the equivalent reception vector and the most recently determined signal vector, and the equivalent channel of the current layer can be obtained. Gain to be true Determine the decision symbol, specifically, determine the decision symbol as ^^, after determining the decision symbol, for each

The decision symbol determines the set number of candidate symbols, because the Euclidean distance of the candidate symbol and the decision symbol is small, the degree of similarity is high, and the signal sent by the transmitting end is also close, so that the candidate symbol can be selected. The signal vector of the current layer is determined with the last determined signal vector to reduce the complexity of signal detection.

For each decision symbol, determining a region to which the decision symbol belongs on the constellation map, and mapping the region corresponding to the symbol corresponding to the channel space of the layer determines that the symbol corresponding to the region is a candidate symbol, the region and the channel of the layer The mapping relationship between the symbols corresponding to the space is specifically: determining a constellation point of the symbol corresponding to the channel space of the layer, and determining a corresponding region of the constellation point for the constellation point of the set number of the closest distance between the two groups And determining a mapping relationship between a symbol corresponding to each group of constellation points and a corresponding region of the group of constellation points, where a sum of plane distances of any constellation points in the corresponding region and the group of constellation points is smaller than a plane of the other group of constellation points The sum of the distance.

Specifically, for the Long Term Evolution (LTE) mode in which four layers of spatial multiplexing transmission are used, the detection in the 16QAM modulation mode is taken as an example, as shown in FIG. 4, the plane distance between each group is two. The nearest four constellation points determine the corresponding regions of the set of constellation points, and the mapping relationship between the corresponding symbols of each set of constellation points and the corresponding regions of the set of constellation points is as shown in Table 2. Stored in Table 2 is the number of the constellation points in Figure 4.

Table 2 Area Lookup Table under 16QAM

When the candidate symbol is determined, if the decision symbol is located in the area 1, the candidate symbol is 13 14 15 16, and if the decision symbol is located in the area 3, the candidate symbol is the complexity of the detection of 9 10 11 12 .

After the candidate symbol is determined, the signal to be selected and the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix are determined to determine the signal vector to be reserved in the channel space of the layer. In the embodiment of the present invention, first, each candidate is selected. And a sign of the Euclidean distance squared increment of the candidate signal vector formed by the signal vector of the candidate symbol, and adding the squared increment of the Euclidean distance to the square of the cumulative Euclidean distance of the signal vector determining the candidate symbol , the cumulative Euclidean distance squared of the candidate signal vector is obtained, and the smaller the cumulative Euclidean distance square is, the closer the candidate signal vector is to the signal vector corresponding to the element corresponding to the Nth layer to the current layer in the equivalent receiving vector. It can be determined that the set candidate signal vector having the smallest cumulative Euclidean distance squared is the signal vector to be reserved. In an actual application, the third determining unit 703 is configured to occupy all the candidate symbols, determine the signal vector of the candidate symbol, and the upper triangular equivalent channel gain matrix, and determine the elements corresponding to the corresponding Nth layer to the current layer in the equivalent receiving vector. The signal vector of the composition removes the influence of the equivalent channel gain of the corresponding channel space, and the signal vector obtained by the signal vector is the closest to the set number of signal vectors, which is specifically used for:

Determining, to each candidate symbol, a candidate signal direction composed of the candidate symbol and a signal vector determining the candidate symbol;

Determining a symbol of the current layer in the signal vector of the signal vector candidate signal vector passing through the Nth layer channel space to the current channel space;

Determining the Euclidean distance squared increment of the candidate signal vector „— _{1 +fl} = - £ r^ - rH where

;='■+1

The number of the channel space for this layer, _m , . For the a-th candidate symbol determined by the mth decision symbol in the i-th channel space, we {l,., M}, M is the number of signal vectors determined by the previous layer of channel space, ae {l , —, } , A is the number of candidate symbols determined according to each decision symbol set by the current layer channel space, and is a signal vector for determining the candidate symbol, which is an element of the i-th row in the equivalent receiving vector, r is an element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and is an element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix;

For each candidate symbol, determine the cumulative Euclidean distance squared BM (( - l)

, where bm( ) is the cumulative Euclidean distance squared of the mth signal vector determined in the i+1th channel space, and BM((m - 1) ^ + «) is the first in the layer 1 channel space ( - 1) X + "the cumulative Euclidean distance squared of the candidate channel vectors;

The signal vector with the smallest cumulative Euclidean distance square of the set number is determined from the candidate signal vector, and the cumulative Euclidean distance square corresponding to the vector is stored in bm.

Preferably, in the uncoded system, the receiving end converts the set number of signal vectors determined by the first layer of channel space into bit outputs, that is, the detection result. However, in the coding system, the fourth determining unit 704 is configured to calculate a set number of signal vectors determined by the first layer channel space and convert the bits into soft bit values of the bits, and determine the detection result according to the soft bit values. The technicians in the field can use other feasible methods to set the test results, which are no longer described here.

Preferably, the signal detection method provided by the embodiment of the present invention can be implemented in the form of a tree search. At this time, the square of the Euclidean distance determined when detecting the Nth layer is the branch metric value in the tree search. The cumulative Euclidean distance square determined by the N-1th to the first layer is the cumulative branch metric, and the Euclidean distance squared increment is the cumulative branch metric increment. Specifically, as shown in FIG. Embodiments of the present invention provide a specific method for signal detection, including Includes:

S601. Determine a unit matrix of the channel matrix and an upper triangular equivalent channel gain by performing QR decomposition on the channel matrix.

1,1 1,2 ^r \,N

Matrix R= ? "²; ² , where N is the number of layers of channel space;

0 ... 0 r _NN

5603. Searching for the Nth layer symbol, the receiving end determines a symbol set of the symbol that the sending end can send at the Nth layer, and determining, for each symbol in the set, a branch metric value of the symbol=|^_/^ _3⁄4 | ² , where ^ is the Nth row element in the equivalent receiving vector, _Ν is the element of the Nth row and the Nth column in the upper triangular equivalent channel gain matrix, X is the symbol, and the score metric is used as the cumulative score metric The value is saved in bm;

5604. Determine M symbols with the smallest branch metric as the M reserved symbols of the symbol sent by the sending end at the Nth layer, and save the reserved symbols in the output matrix, and keep the branch metric values corresponding to the reserved symbols in bm;

5605. Search for the N-1th layer to the 1st layer symbol, reserve M branches in the output matrix, and determine M decision symbols, _m -

Where is the number of layers in the channel space of the layer, r is the element of the i-th row and the j-th column in the upper triangular equivalent channel gain matrix, and is the element of the i-th row and the i-th column in the upper triangular equivalent channel gain matrix, The element of the i-th row and the m-th column in the output matrix, e {1,···, } , is the element of the i-th row in the equivalent receiving vector;

S606, determining a set of all symbols that the sender can send at the layer;

The selected symbol and the reserved branch corresponding to each candidate symbol determine M* A branch metric increments ^3⁄4-., ι ² , where, the number of layers of the layer channel space, etc.

The element of the i-th row in the received vector, r is the element of the i-th row and the j-th column of the upper triangular equivalent channel gain matrix, and _riJ is the element of the i-th row and the i-th column of the upper triangular equivalent channel gain matrix, which is an output The element of the mth column of the i-th row in the matrix, we{l,---, }, _m ,. The a-th candidate to be selected according to the mth decision amount in the i-th channel,

S609. Determine, according to the M*A branch metric increments, M*A cumulative branch metrics BM((m - ΐ) χΑ + α) = bm(m) + d _m _ _l)xA+a , where bm ( ) is the cumulative branch metric of the mth branch determined on the i+1th channel, BM((

+ «) is the ( - 1) X + "selected channel vector in the layer 1 channel space Cumulative Euclidean distance squared;

5610. Determine M minimum cumulative branch metric values, and determine corresponding branches, and use the corresponding branch as a reserved branch to be stored in the output matrix, and store the cumulative branch metric corresponding to the reserved branch into bm;

5611. Determine a signal detection result according to the M reserved branches determined in the layer 1 channel space.

Embodiments of the present invention provide a signal detecting method and apparatus. When detecting a signal received by an N-layer spatial multiplexing transmission system, when detecting the N-1 layer to the first layer, the layer is first determined. The symbol received by the channel space removes the decision symbol of the interference of the most recently determined signal vector and the equivalent channel gain of the channel space of the layer, and further determines the candidate symbol closest to the decision symbols, and then determines the basis Which of the signal vectors determined by the candidate symbols are closest to the signal vector obtained by removing the equivalent channel gain of the corresponding channel space from the signal vector received by the current channel space, and the reserved signal vector of the layer can be determined. Therefore, only the candidate symbols determined according to the decision symbols need to be filtered, and it is not necessary to filter all possible symbols of the layer, which reduces the complexity of signal detection.

Those skilled in the art will appreciate that embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware. Moreover, the present invention can be embodied in the form of a computer program product embodied on one or more computer-usable storage interfaces (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the invention. It will be understood that each process and/or block of the flowchart illustrations and/or FIG. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine for the execution of instructions for execution by a processor of a computer or other programmable data processing device. Means for implementing the functions specified in one or more of the flow or in a block or blocks of the flow chart.

The computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device. The apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.

These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

Although a preferred embodiment of the present invention has been described, one of ordinary skill in the art will be aware of the basic inventive Additional changes and modifications may be made to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and the modifications and

It is apparent that those skilled in the art can make various modifications and variations to the invention without departing from the spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and modifications of the invention

Claims

Rights request

1. A signal detection method, characterized by including:

The receiving end determines the equivalent reception vector based on the signal sequence received through the N-layer spatial multiplexing transmission system and the unitary matrix obtained through the QR decomposition of the channel matrix;

In the symbol corresponding to the Nth layer channel space, the receiving end determines, based on the upper triangular equivalent channel gain matrix obtained through the QR decomposition of the channel matrix, the element corresponding to the layer in the equivalent reception vector, excluding the channel space of the layer. The set number of symbols that are closest to the symbols obtained after the influence of the effective channel gain, and each determined symbol is regarded as a signal vector;

For each layer of channel space from the N-1th layer channel space to the 1st layer channel space, the receiving end determines the equivalent reception based on the upper triangular equivalent channel gain matrix for each signal vector most recently determined. The element in the vector corresponding to the layer is the decision symbol after removing the interference of the signal vector and the equivalent channel gain of the channel space of the layer. For each decision symbol, the setting is determined in the symbol corresponding to the channel space of the layer. A certain number of candidate symbols that are closest to it, and based on all candidate symbols, the signal vector of the candidate symbol and the upper triangular equivalent channel gain matrix, determine the equivalent receive vector The signal vector composed of elements corresponding to the Nth layer to the current layer is the closest set number of signal vectors to the signal vector obtained after removing the influence of the equivalent channel gain of the corresponding channel space;

2. The method according to claim 1, characterized in that the number of signal vectors determined in each layer of channel space is the same.

3. The method according to claim 1, characterized in that, the symbol most similar to the element in the equivalent reception vector corresponding to the layer is obtained by removing the influence of the equivalent channel gain of the channel space of the layer. A certain number of symbols, including:

The set number of symbols with the minimum squared Euclidean distance between the symbols obtained by removing the influence of the equivalent channel gain of the channel space of the layer from the elements corresponding to the layer in the equivalent receive vector;

For each decision symbol, a set number of candidate symbols closest to it are determined among the symbols corresponding to the channel space of the layer, specifically including:

For each decision symbol, a set number of candidate symbols with the smallest squared Euclidean distance between the symbols corresponding to the channel space of the layer are determined;

Determining the set number that is closest to the signal vector obtained by removing the influence of the equivalent channel gain of the corresponding channel space from the signal vector composed of the elements corresponding to the Nth layer to the current layer in the equivalent receive vector. Signal vector, specifically including:

Determine a set number of signal vectors from the candidate signal vectors composed of the candidate symbols and the signal vectors that determine the candidate symbols, and the set number of signal vectors are transmitted through the channel space from the Nth layer to the current layer The squared cumulative Euclidean distance between the signal vector obtained and the signal vector composed of elements corresponding to the Nth layer to the current layer in the equivalent receiving vector is the most Small, the cumulative squared Euclidean distance is specifically the sum of the squared Euclidean distances between each corresponding symbol of the two signal vectors.

4. The method according to claim 1, characterized in that, in the symbols corresponding to the Nth layer channel space, the receiving end determines the equivalent reception vector in the symbol of the layer minus the equivalent value of the layer channel space. The set number of symbols closest to the symbols affected by the channel gain, specifically including:

The receiving end determines, for each symbol among the symbols corresponding to the channel space of the layer, the symbol obtained by removing the influence of the equivalent channel gain of the channel space of the layer from the symbol and the element corresponding to the layer in the equivalent reception vector. The squared Euclidean distance 4, _¥ =1 ^ - ^ 1 ² , where ^ is the Nth row element in the equivalent reception vector, ^ is the Nth row and Nth column in the upper triangular equivalent channel gain matrix Element, x _q is the symbol;

Determine the minimum set number of symbols for the squared Euclidean distance.

5. The method according to claim 1, characterized in that, for each signal vector most recently determined, the receiving end uses the upper triangular equivalent channel gain matrix to determine the corresponding layer in the equivalent reception vector. The elements of the decision symbol after removing the interference of the signal vector and the equivalent channel gain of the layer channel space include:

For each signal vector that was last determined, the decision symbol for determining the signal vector is:

. _m = ( . - r _jm ) lr _i , where is the number of the channel space of this layer, e {l, . ,M}, M is the latest

' =w ' ' '

The number of signal vectors determined at this time, r is the element in the i-th row and j-th column in the upper triangular equivalent channel gain matrix, . is the element in the i-th row and i-th column in the upper triangular equivalent channel gain matrix , X _Lm is the symbol of the j-th row in the m-th channel vector, . is the element of the i-th row in the equivalent reception vector.

6. The method according to claim 1, characterized in that, for each decision symbol, a set number of candidate symbols closest to it are determined in the symbols corresponding to the channel space of the layer, specifically including:

For each decision symbol, the area to which the decision symbol belongs on the constellation diagram is determined, and the mapping relationship between the area and the symbol corresponding to the channel space of the layer is determined, and the symbol corresponding to the area is determined to be the candidate symbol, and the area and the corresponding symbol are determined The mapping relationship of the symbols corresponding to the layer channel space is specifically: determining the constellation points of the symbols corresponding to the layer channel space, and determining the correspondence of the group of constellation points for the set number of constellation points with the shortest plane distance between each group. After the area is determined, the mapping relationship between the symbols corresponding to each group of constellation points and the corresponding area of the group of constellation points is determined. The sum of the plane distances of any constellation point in the corresponding area and the group of constellation points is less than the sum of the plane distances of any constellation point in the corresponding area and the group of constellation points. The sum of the plane distances of points.

7. The method according to claim 3, characterized in that, based on all the candidate symbols, the signal vectors of the candidate symbols and the upper triangular equivalent channel gain matrix, determining and The signal vector consisting of the elements corresponding to the Nth layer to the current layer in the effective reception vector is the signal vector obtained by removing the influence of the equivalent channel gain of the corresponding channel space and the closest set number of signal vectors, specifically including:

For each candidate symbol, determine the candidate signal vector consisting of the candidate symbol and the signal vector that determines the candidate symbol; Determine the symbol corresponding to the current layer in the signal vector after the signal vector to be selected is transmitted through the Nth layer channel space to the current channel space;

Determine the squared Euclidean distance increment of the candidate signal vector^)^. ^- £ r _J X^ _m - _ri x _m ,its

;='■+1

Among them, is the number of the channel space of this layer, _m . is the a-th candidate symbol determined by the m-th decision symbol in the i-th layer channel space, we {l,.,M}, M is the number of signal vectors determined by the previous layer channel space, αε {1 ,... The element in the i-th row in the received vector, r is the element in the i-th row and j-th column in the upper triangular equivalent channel gain matrix, _riJ is the element in the i-th row and i-th column in the upper triangular equivalent channel gain matrix ; For each candidate symbol, determine the cumulative Euclidean distance squared BM(( - l)x +

, where bm( ) is the cumulative Euclidean distance square of the m-th signal vector determined in the i+1th layer channel space, BM((m -1) ^ + «) is the ((m -1) ^ + «) in the first-layer channel space - 1) The cumulative Euclidean distance square of X + " candidate channel vectors;

From the candidate signal vectors, a signal vector with the smallest cumulative Euclidean distance square of a set number is determined.

8. A signal detection device, characterized in that it includes:

The first determination unit is used to determine the equivalent reception vector based on the signal sequence received through the N-layer spatial multiplexing transmission system and the unitary matrix obtained through the channel matrix QR decomposition;

The second determination unit is used to determine, in the symbol corresponding to the Nth layer channel space, the upper triangular equivalent channel gain matrix obtained by decomposing the channel matrix QR with the element corresponding to the layer in the equivalent reception vector, excluding the After the influence of the equivalent channel gain of the layer channel space, the closest set number of symbols is obtained, and each determined symbol is regarded as a signal vector;

The third determination unit is used to determine, for each layer of channel space from the N-1th layer channel space to the first layer channel space, the receiving end determines the upper triangle equivalent channel gain for each signal vector most recently determined. The matrix determines the decision symbol after removing the interference of the signal vector and the equivalent channel gain of the channel space of the layer from the element corresponding to the layer in the equivalent received vector. For each decision symbol, the signal vector is in the channel of the layer. Determine a set number of candidate symbols that are closest to them among the symbols corresponding to the space, and based on all the candidate symbols, the signal vectors of the candidate symbols and the upper triangular equivalent channel gain matrix, determine the equation with The set number of signal vectors closest to the signal vector obtained by removing the influence of the equivalent channel gain of the corresponding channel space from the signal vector composed of elements corresponding to the Nth layer to the current layer in the equivalent reception vector;

The fourth determination unit is used to determine the signal detection result based on the signal vector determined in the first layer channel space.

9. The device according to claim 8, characterized in that the number of signal vectors determined in each layer of channel space is the same.

10. The device according to claim 8, characterized in that: the symbol of the equivalent receive vector in the layer The set number of symbols with the closest symbols obtained after removing the influence of the equivalent channel gain in the channel space of this layer, specifically includes:

Determine the set number of signals that are closest to the signal vector composed of the symbols from the Nth layer to the current layer of the equivalent reception vector, removing the influence of the equivalent channel gain of the corresponding channel space. Vector, specifically including:

Determine a set number of signal vectors from the candidate signal vectors composed of the candidate symbols and the signal vectors that determine the candidate symbols, and the set number of signal vectors are transmitted through the channel space from the Nth layer to the current layer The cumulative Euclidean distance square between the signal vector obtained and the signal vector consisting of elements from the Nth layer to the current layer in the equivalent received vector is the smallest. The cumulative Euclidean distance square is specifically for each of the two signal vectors. The sum of squared Euclidean distances between corresponding symbols.

11. The apparatus according to claim 8, wherein the second determination unit determines, in the symbols corresponding to the Nth layer channel space, the equivalent reception vector in the symbol of the layer minus the symbol of the layer channel space. The set number of symbols closest to the symbols affected by the equivalent channel gain, specifically used for:

Determine the minimum set number of symbols with the squared Euclidean distance.

12. The device according to claim 8, characterized in that, for each signal vector most recently determined, the third determination unit determines, based on the upper triangular equivalent channel gain matrix, the equivalent receive vector at the The symbol of the layer is the decision symbol after removing the influence of the signal vector on it and the equivalent channel gain of the channel space of the layer. It is specifically used to: for each signal vector recently determined, determine the decision symbol of the signal vector. for:

x _im = ( _yi - r _jm ) lr _i , where is the number of the channel space of this layer, e {l, . ,M} , M is the number of the most recently determined signal vectors, r is the upper triangular equivalent The element in the i-th row and j-th column in the channel gain matrix, . is the element in the i-th row and i-th column in the upper triangular equivalent channel gain matrix, Xj, _m is the symbol of the j-th row in the m-th channel sequence , . is the element of the i-th row in the equivalent reception vector.

13. The device according to claim 8, characterized in that, for each decision symbol, the third determination unit determines a set number of candidate symbols that are closest to it among the symbols corresponding to the channel space of the layer, Specifically used for:

14. The apparatus according to claim 10, wherein the third determination unit determines and The signal vector composed of the symbols from the Nth layer to the current layer of the equivalent reception vector removes the influence of the equivalent channel gain of the corresponding channel space and is the closest set number of signal vectors to the signal vector, specifically used for :

For each candidate symbol, determine the candidate signal vector consisting of the candidate symbol and the signal vector that determines the candidate symbol;

Determine the symbol corresponding to the current layer in the signal vector after the signal vector to be selected is transmitted through the Nth layer channel space to the current channel space;

;='■+1

Among them, is the number of the channel space of this layer, _m . is the a-th candidate symbol determined by the m-th decision symbol in the i-th layer channel space, we{l,.,M}, M is the number of signal vectors determined by the previous layer channel space, αε {1 ,... The element in the i-th row in the received vector, r is the element in the i-th row and j-th column in the upper triangular equivalent channel gain matrix, _riJ is the element in the i-th row and i-th column in the upper triangular equivalent channel gain matrix ; For each candidate symbol, determine the cumulative Euclidean distance squared BM(( - l)

, where bm( ) is the cumulative Euclidean distance square of the m-th signal vector determined in the i+1th layer channel space, BM((m -1) ^ + «) is the ((m -1) ^ + «) in the first-layer channel space - 1) The cumulative Euclidean distance square of x + "candidate channel vectors;