CN101834827B

CN101834827B - Signal detection method and device in multiple-input multiple-output system

Info

Publication number: CN101834827B
Application number: CN2010101354912A
Authority: CN
Inventors: 杨�远; 曹晏波; 乔元新; 王雪; 颜尧平
Original assignee: DATANG LINKTECH INFOSYSTEM Co Ltd
Current assignee: DATANG LINKTECH INFOSYSTEM Co Ltd
Priority date: 2010-03-29
Filing date: 2010-03-29
Publication date: 2012-07-18
Anticipated expiration: 2030-03-29
Also published as: CN101834827A

Abstract

The embodiment of the invention discloses signal detection method and device in a multiple-input multiple-output system. The signal detection method comprises the following steps of: converting a constellation point of a quadrature amplitude modulation (QAM) transmitting signal into a bit vector weight sum form to obtain a bit-level-expressed QAM transmitting signal; converting a channel matrix into a compound channel matrix according to the bit-level-expressed QAM transmitting signal; carrying out QR decomposition on the compound channel matrix to obtain an upper triangular matrix; constructing a bit-by-bit layered tree structure by using the upper triangular matrix; searching the bit-by-bit layered tree structure layer by layer based on a breadth-first M algorithm to obtain a signal candidate set; calculating a measure value of each candidate signal in the signal candidate set; and calculating the posterior information of the transmitting signal by using the measure value. The embodiment of the invention can reduce the realization difficulty of a signal detection process.

Description

Signal detection method and device in multi-input multi-output system

Technical Field

The present application relates to the field of communications and computer technologies, and in particular, to a method and an apparatus for detecting signals in a mimo system.

Background

With the development of wireless communication technology, the MIMO (Multiple-input Multiple-output) technology using Multiple antenna elements fully utilizes space resources and can greatly improve the frequency utilization rate of the system within a limited bandwidth, so that the MIMO technology provides a capacity potential which cannot be provided by a single antenna technology for a user and becomes an effective way for improving the capacity and reliability of the system by using space dimensions. Currently, MIMO technology has become one of hot spot technologies in 3G or 4G.

In the MIMO system, since the performance of the receiving end will ultimately greatly affect the transmission rate, the error rate performance and the system complexity of the whole transmission system, the improvement of the performance of the receiving end becomes a focus of attention of researchers. At present, researchers have proposed applying a sequence decoding method to the signal detection process of the MIMO system. The tree searching method using the breadth-first M algorithm can search to obtain a signal candidate sequence, calculate the metric values of the signal candidate sequence, and calculate the posterior information of the bits corresponding to the transmitted signals by using the metric values to complete the signal detection process. The tree searching method based on the breadth-first M algorithm not only ensures that the complexity of calculation is not changed along with the change of the signal-to-noise ratio and the channel condition, but also ensures that a transmission system has good error rate performance.

However, the inventors found in their studies that: the tree searching method using the breadth-first M algorithm is based on the signal tree of the symbol level for searching, under the condition of higher constellation dimensionality, the calculation complexity of the whole tree searching process is high, and the realization difficulty of the signal detection process is increased finally.

Disclosure of Invention

In order to solve the above technical problem, embodiments of the present application provide a method and an apparatus for signal detection in a mimo system, so as to reduce implementation difficulty in a signal detection process.

The embodiment of the application discloses the following technical scheme:

a method of signal detection in a multiple-input multiple-output system, comprising: converting constellation points of Quadrature Amplitude Modulation (QAM) transmitting signals into a bit vector weighted sum form to obtain QAM transmitting signals represented by a bit level, and converting a channel matrix into a composite channel matrix according to the QAM transmitting signals represented by the bit level; performing QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and constructing a bit-by-bit hierarchical tree structure by using the upper triangular matrix; searching the tree structure which is layered bit by bit on the basis of a breadth-first M algorithm layer by layer to obtain a signal candidate set, and calculating the metric value of each candidate signal in the signal candidate set; and calculating posterior information of the transmitting signal by using the metric value.

A signal detection apparatus in a multiple-input multiple-output system, comprising: the conversion unit is used for converting the constellation points of the Quadrature Amplitude Modulation (QAM) sending signals into a bit vector weighted sum form to obtain QAM sending signals represented by a bit level, and obtaining a composite channel matrix according to the QAM sending signals represented by the bit level; the decomposition unit is used for carrying out QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and a bit-by-bit hierarchical tree structure is constructed by utilizing the upper triangular matrix; the searching unit is used for searching the bit-by-bit hierarchical tree structure layer by layer based on a breadth-first M algorithm to obtain a signal candidate set and calculating the metric value of each candidate signal in the signal candidate set; and the detection unit calculates the posterior information of the transmitting signal by using the metric value.

It can be seen from the above embodiments that the bit-level tree search detection method in the present application can reduce the branch paths that need to be searched for each time, and only 2M node metrics need to be calculated after each layer of nodes are expanded correspondingly. Compared with the existing tree searching method based on the symbol, in each stage of the tree searching, each retention path needs to be expanded into a range of constellation set according to the size of the constellation set

The branch paths need to be calculated after the corresponding nodes at each layer are expanded

Measurement of child nodes, and thus complexity and modulation order M of the entire tree search algorithm_rIn an exponential relationshipBy the implementation scheme, the complexity of the whole algorithm is reduced, and therefore the implementation difficulty of the signal detection process is reduced.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.

FIG. 1 is a flow chart of an embodiment of a method for signal detection in a multiple input multiple output system of the present application;

fig. 2 is a schematic diagram of gray mapping according to the present application;

FIG. 3 is a schematic illustration of a natural mapping according to the present application;

fig. 4 is an expanded search graph of the bit-level tree search algorithm when the retention path of the 16QAM constellation is 4 according to the present application;

FIG. 5 is a block diagram of an embodiment of a signal detection apparatus in a MIMO system according to the present application;

fig. 6 is a block diagram of another embodiment of a signal detection apparatus in a mimo system according to the present application.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present application more comprehensible, embodiments accompanying the present application are described in detail below.

Example one

Please refer to fig. 1, which is a flowchart illustrating a signal detection method in a mimo system according to an embodiment of the present invention, the method includes the following steps:

step 101: converting constellation points of Quadrature Amplitude Modulation (QAM) transmitting signals into a bit vector weighted sum form to obtain QAM transmitting signals represented by a bit level, and converting a channel matrix into a composite channel matrix according to the QAM transmitting signals represented by the bit level;

step 102: performing QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and constructing a bit-by-bit hierarchical tree structure by using the upper triangular matrix;

step 103: and searching the bit-by-bit hierarchical tree structure layer by applying a breadth-first M algorithm to obtain a signal candidate set, and calculating the metric value of each candidate signal in the signal candidate set.

Step 104: and calculating posterior information of the transmitting signal by using the metric value.

The above signal detection process is explained in detail below. When the constellation points of the QAM transmission signals are converted into a form represented by bit vectors to obtain QAM transmission signals represented by the bit vectors, and the channel matrix is converted into a composite channel matrix according to the QAM transmission signals represented by the bit vectors, the following procedure may be performed.

Suppose that one uses N_tRoot transmitting antenna and N_rIn the MIMO system model of the root receiving antenna, the sending symbol is quadrature amplitude modulation of a square constellation, and the size of the constellation set isWherein M is_cFor the number of bits of each transmission symbol in the complex field, the discrete-time system model of the MIMO system in the complex field is:

y^c＝H^cs^c+n^c (1)

the system model in the complex domain of equation (1) is equivalent to a system model in the real domain as follows:

y＝Hs+n (2)

it can be concluded that, in the real number domain, the vector dimension of the received signal y is N_RX 1, vector dimension of the transmitted signal s is N_TX 1, the vector dimension of the channel matrix H is N_T×N_TThe vector dimension of the real Gaussian noise N is N_RX 1, mean value ofThe variance matrix is

Wherein N is_RIs the equivalent number of receiving antennas in real number domain, N_TIs the equivalent number of transmitting antennas in the real number domain, and N_R＝2N_r，N_T＝2N_t。

When the MIMO channel is a flat fading Rayleigh channel, N frames per frame_TM_rThe long bit symbols are:

wherein M is_rNumber of bits in real number domain for each transmitted symbol, and M_r＝M_c/2. After serial-to-parallel conversion, the symbols are mapped into a transmission symbol s.

When the system uses gray mapping as shown in fig. 2, then the transmitted symbol s on the kth transmit antenna_kAnd its corresponding mapping bit symbol x_k，j，j＝1，…，M_rThe relationship between can be represented by:

an example of an 8-PAM gray mapped constellation is given in fig. 2, where an intermediate parameter u is introduced_k，i，u_k，iE { +1, -1}, let,

when using u_kTo represent u_k，i，i＝1，…，M_rThe composed vector, denoted by w, corresponds to u_k，iBy a weight factor w_i，i＝1，…，M_rWhen the vector is formed, then,

thus, s_kThe vector of (d) is represented as:

s_k＝w^Tu_k (10)

in the case of M-QAM, the,

the weight factor is:

wherein j represents an imaginary symbol, i.e.

M is the number of constellation points. When using a 16QAM constellation, for example, M-16,

from u_k，iCan be found in_k，iAnd u_k，iAnd u_k，i-1The following relations exist between the following components:

x_k，i＝u_k，iu_k，i-1，i＝2，…，M_r (12)

when i is 1, x_k，i＝u_k，i. By substituting equation (12) into equation (7), the relationship between vectors s and u can be obtained:

<math> <mrow> <mi>s</mi> <mo>=</mo> <mrow> <mo>(</mo> <msub> <mi>I</mi> <msub> <mi>N</mi> <mi>T</mi> </msub> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> <mo>)</mo> </mrow> <mi>u</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>13</mn> <mo>)</mo> </mrow> </mrow> </math>

wherein,

representing the Kronecker direct product.

When the formula (13) is brought into the formula (2), there are

Converting constellation points of Quadrature Amplitude Modulation (QAM) transmitting signals into a bit vector weighted sum form to obtain QAM transmitting signals u represented by bit levels, and defining a composite channel matrix according to the relation between the QAM transmitting signals s and the bit vectors u:

<math> <mrow> <mi>A</mi> <mo>=</mo> <mi>H</mi> <mrow> <mo>(</mo> <msub> <mi>I</mi> <msub> <mi>N</mi> <mi>T</mi> </msub> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> <mo>)</mo> </mrow> </mrow> </math>

<math> <mrow> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>h</mi> <mn>1</mn> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> <mtd> <msub> <mi>h</mi> <mn>2</mn> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> <mtd> <mo>·</mo> <mo>·</mo> <mo>·</mo> </mtd> <mtd> <msub> <mi>h</mi> <msub> <mi>N</mi> <mi>T</mi> </msub> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>14</mn> <mo>)</mo> </mrow> </mrow> </math>

wherein h is_iRepresenting the ith column of the channel matrix H. Definition of N_c＝N_TM_rDimension of the composite channel matrix A is N_R×N_c. The equivalent channel model can be written as:

y＝Au+n (15)

in addition, when the system uses natural mapping as shown in fig. 3, then the symbol s on the kth transmit antenna_kAnd its corresponding mapping bit x_k，j，j＝1，…，M_rThe relationship between can be represented by:

s_kthe vector of (d) is represented as: s_k＝w^Tx_k (17)

Wherein,

when a square QAM constellation is used, the weighting factor w in equation (17) is calculated as equation (11). Then, similar to gray mapping, the composite channel matrix is defined as:

<math> <mrow> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>h</mi> <mn>1</mn> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> <mtd> <msub> <mi>h</mi> <mn>2</mn> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> <mtd> <mo>·</mo> <mo>·</mo> <mo>·</mo> </mtd> <mtd> <msub> <mi>h</mi> <msub> <mi>N</mi> <mi>T</mi> </msub> </msub> <mo>&CircleTimes;</mo> <msup> <mi>w</mi> <mi>T</mi> </msup> </mtd> </mtr> </mtable> </mfenced> </mrow> </math>

the dimension of the composite channel matrix A is N_R×N_c. The equivalent channel model can be written as:

y＝Ax+n (18)

wherein,

when QR decomposition is performed on the composite channel matrix by using MMSE criterion to obtain an upper triangular matrix, and a bit-by-bit hierarchical tree structure is constructed by using the upper triangular matrix, the method may be performed as follows.

First, in order to obtain an equivalent full rank matrix, based on the MMSE criterion, an extended composite channel matrix is defined as:

<math> <mrow> <munder> <mi>A</mi> <mo>&OverBar;</mo> </munder> <mover> <mo>=</mo> <mi>Δ</mi> </mover> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <mi>A</mi> </mtd> </mtr> <mtr> <mtd> <msub> <mi>σ</mi> <mi>n</mi> </msub> <msub> <mi>I</mi> <msub> <mi>N</mi> <mi>c</mi> </msub> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>19</mn> <mo>)</mo> </mrow> </mrow> </math>

secondly, for the expanded composite channel matrixAQR decomposition is carried out to obtainAQR, hereThe column vectors thereof are orthogonal to each other, and

is an upper triangular matrix.

It should be noted here that, in the signal detection process in the mimo system, the metric is calculated according to the following formula:

wherein L is_A(x) Representing a priori information corresponding to the bit vector x.

Finally, to get better detection, we apply matrices before QR decompositionAThe columns of (a) are reordered.

For Gray mapping, | | y-Hs | | sweet calculation in formula (20)²And

and then the measured values are correlated to obtain the gray mapping,

{| | y - Hs | |}^{2} = {| | y - Au | |}^{2}

= y^{H} y - y^{H} Au - u^{H} A^{H} y + u^{H} A^{H} Au

wherein,

order to

Since the symbols in the intermediate vector u are all bit symbols, | u is satisfied_k，i|²At 1, it is clear that the size of C has no relation to the choice of the intermediate variable u at this time, and therefore it can be omitted from the metric calculation. Then the metric can be re-expressed as:

further, for natural mapping, | | y-Hs | | survival in equation (20)²And

and then the measured values are correlated to obtain the natural mapping,

wherein,

the above formula (23) and formula (24) are mathematical expressions of signal metric values under gray mapping and natural mapping, respectively, that is, a tree structure layered bit by bit is searched layer by layer based on a breadth-first M algorithm, a signal candidate set can be obtained, and a mathematical model of a metric value process of each candidate signal in the signal candidate set is calculated.

The method comprises the following steps of searching a tree structure which is layered bit by bit on the basis of a breadth-first M algorithm layer by layer to obtain a signal candidate set, and calculating the metric value of the signal candidate set comprises the following steps: calculating the metric values of 2M nodes on 2M branch paths expanded by M retention nodes of the previous layer in the current layer of the bit-by-bit hierarchical tree structure; searching M nodes with the maximum metric value from the 2M nodes, and taking the M nodes with the maximum metric value as retention nodes of the current layer; and sequentially searching layer by layer, selecting M retention nodes from the 2M nodes of the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

For example, as shown in fig. 4, at the k-th level of the bit-by-bit hierarchical tree structure, metric values of 8 nodes on 8 branch paths expanded by 4 retention nodes of the k-1 level are calculated, metric values of the 8 nodes are compared, and 4 nodes with the largest metric values are selected as retention nodes of the k-th level. And searching layer by layer in sequence, when the last layer is calculated, selecting 4 retention nodes from 8 nodes of the last layer, taking the 4 retention nodes as a signal candidate set, and calculating the metric values of the 4 retention nodes in the signal candidate set.

In gray mapping, the updated calculation of the metric value can be expressed by the following mathematical expression:

the process of metric hierarchical computation is seen from equation (25), and after the computation of the metric of each layer is completed, the metric is sorted, and M nodes with the maximum metric are reserved as the reserved nodes for the next expansion.

In the natural mapping, the updated calculation of the metric value can be expressed by the following mathematical expression:

in a preferred embodiment, performing a layer-by-layer search on a tree structure layered bit by bit based on a breadth-first M algorithm to obtain a signal candidate set, and calculating a metric value of the signal candidate set includes: in the tree structure of the bit-by-bit layering, acquiring prior information of corresponding bits of all layers from a decoder; the layer corresponding to the node with the prior information larger than the first threshold value is not searched and expanded, the preserved path of the preserved node on the upper layer is expanded by using the symbol of the prior information of the current layer, the node with the prior information not larger than the first threshold value is expanded, and M nodes with the maximum metric value are searched as the preserved nodes of the current layer according to the magnitude of the metric value; and sequentially searching layer by layer, selecting M retention nodes from the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

For example, in a tree structure of bit-by-bit layering, prior information of corresponding bits of all layers is obtained from a decoder, the prior information of the corresponding bits of all layers is compared with a first threshold, for example, if the prior information of the corresponding bits of the K-th layer is greater than the first threshold, the surviving paths of 4 surviving nodes of the K-1-th layer are extended by using symbols of the prior information of the K-th layer, and if the prior information of the corresponding bits of the K + 1-th layer is not greater than the first threshold, 4 surviving nodes of the K-1-th layer are expanded, metric values of 8 nodes on 8 branch paths expanded by the 4 nodes are calculated, the metric values of the 8 nodes are compared, and the largest 4 nodes of the metric values are selected as the surviving nodes of the K + 1-th layer. And sequentially searching layer by layer, selecting 4 retention nodes from the last layer when the last layer is searched, taking the 4 retention nodes as a signal candidate set, and calculating the metric value of the signal candidate set.

Another preferred embodiment is that, performing layer-by-layer search on a tree structure layered bit by bit based on a breadth-first M algorithm to obtain a signal candidate set, and calculating a metric value of the signal candidate set includes: in the tree structure of the bit-by-bit layering, acquiring prior information of corresponding bits of all layers from a decoder; taking the prior information weighted value of the node of which the prior information is greater than the second threshold value as a metric value, searching M nodes with the maximum metric value from 2M nodes spread by M retention nodes on the upper layer, and taking the M nodes with the maximum metric value as the retention nodes on the current layer; and sequentially searching layer by layer, selecting M retention nodes from the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

For example, in a tree structure with bit-by-bit hierarchy, the prior information of the corresponding bits of all layers is obtained from the decoder, the prior information of the corresponding bits of all layers is compared with the second threshold respectively, and the weighted value of the prior information of the node with the prior information greater than the second threshold is used as the metric value, e.g., for gray mapping, the weighted value of the prior information in the formula (19)

I.e. a priori information weighted value, for the natural mapping, in equation (20)

I.e. the prior information weighted value. And if the prior information of the corresponding bit of the k layer is larger than the second threshold, selecting the prior information weighted value of 8 nodes expanded by the 4 retention nodes in the k-1 layer as the metric value, and selecting the 4 nodes with the maximum metric value as the retention nodes of the k layer. . Searching layer by layer in sequence, when the last layer is searched, selecting 4 retention nodes from the last layer, and searching the 4 retention nodesThe retention node serves as a signal candidate set, and a metric value of the signal candidate set is calculated.

Of course, a combination of the two preferred modes can also be achieved. For example, after obtaining the prior information of the corresponding bits of all layers from the decoder in the bit-by-bit hierarchical tree structure, the prior information of the corresponding bits of all layers is compared with the first threshold value, the layer corresponding to the node of which the prior information is greater than the first threshold value is not searched and expanded, for example, when the prior information of the corresponding bit of the k-th layer is greater than the first threshold value, the retention paths of 4 retention nodes of the k-1-th layer are extended by using the symbol of the prior information of the k-th layer, and the prior information weighted value of the node of which the prior information is greater than the second threshold value is taken as the metric value from the remaining layers. And if the prior information of the corresponding bit of the k layer is larger than the second threshold, selecting the prior information weighted value of 8 nodes expanded by the 4 retention nodes in the k-1 layer as the metric value, and selecting the 4 nodes with the maximum metric value as the retention nodes of the k layer. And for the nodes of which the prior information is not greater than the second threshold, searching M nodes with the maximum metric value from the expanded 2M nodes according to the magnitude of the metric value, and taking the M nodes with the maximum metric value as retention nodes of the current layer. And sequentially searching layer by layer, selecting 4 retention nodes from the last layer when the last layer is searched, taking the 4 retention nodes as a signal candidate set, and calculating the metric value of the signal candidate set.

It should be noted that when gray mapping is used, it can be obtained from the definition of the metric if

Then we need only compute the metric related to the a priori information. The following relationships between the metrics are:

||y-Hs||²the expectation of + C may be approximated as σ_n ²(N_R+N_c). When | L_A(x_i)|＞＞T₁，T₁≥N_R+N_cIn which T is₁For the second threshold, only the metric associated with the a priori information may be calculated, and the metric associated with the i-th layer signal may be written as:

as the signal-to-noise ratio increases, the absolute value of the a priori information available from the channel decoder increases. The a priori information about these bits can be considered reliable when the absolute value of the a priori information obtained from the decoder exceeds a set threshold (first threshold). Then we do not compute their metrics for these bits any more, and the depth of the search tree can be reduced significantly. The computational complexity of the algorithm can therefore be reduced as the signal-to-noise ratio increases.

The principles described above are equally applicable to natural mapping.

It should be further noted that the value range of the first threshold is 3-5 times of the second threshold.

It can be seen from the above embodiments that the bit-level tree search detection method in the present application can reduce the branch paths that need to be searched for each time, and only 2M node metrics need to be calculated after each layer of nodes are expanded correspondingly. Compared with the existing tree searching method based on the symbol, in each stage of the tree searching, each retention path needs to be expanded into a range of constellation set according to the size of the constellation setThe branch paths need to be calculated after the corresponding nodes at each layer are expanded

Measurement of child nodes, and thus complexity and modulation order M of the entire tree search algorithm_rIn an exponential relation, the implementation scheme of the method reduces the complexity of the whole algorithm, and therefore the implementation difficulty of the signal detection process is reduced.

EXAMPLE III

Corresponding to the signal detection method in the mimo system, the embodiment of the present application further provides a signal detection apparatus in the mimo system. Please refer to fig. 5, which is a block diagram of an embodiment of a signal detection apparatus in a mimo system according to the present application, the apparatus includes a conversion unit 501, a decomposition unit 502, a search unit 503, and a detection unit 504. The internal structure and connection relationship of the device will be further described below in conjunction with the working principle of the device.

A converting unit 501, configured to convert a constellation point of a quadrature amplitude modulation QAM transmission signal into a bit vector weighted sum form to obtain a bit-level-represented QAM transmission signal, and convert a channel matrix into a composite channel matrix according to the bit-level-represented QAM transmission signal;

a decomposition unit 502, configured to perform QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and construct a bit-by-bit hierarchical tree structure using the upper triangular matrix;

the searching unit 503 is configured to perform layer-by-layer search on the bit-by-bit hierarchical tree structure based on a breadth-first M algorithm to obtain a signal candidate set, and calculate a metric value of each candidate signal in the signal candidate set;

a detecting unit 504, configured to calculate a posteriori information of the transmitted signal using the metric value.

Please refer to fig. 6, which is a block diagram illustrating a signal detection apparatus in a mimo system according to another embodiment of the present application. Wherein, the search unit 503 further includes: a calculation sub-unit 5031, a first search sub-unit 5032 and a selection sub-unit 5033,

a computation subunit 5031, configured to compute, in a current layer of the bit-wise hierarchical tree structure, metric values of 2M nodes on 2M branch paths expanded by M retention nodes of a previous layer;

a first searching subunit 5032, configured to search M nodes with the largest metric value from the 2M nodes, and use the M nodes with the largest metric value as retention nodes of a current layer;

a selecting subunit 5033, configured to sequentially search layer by layer, select M surviving nodes from the 2M nodes in the last layer as a signal candidate set, and calculate a metric value of the signal candidate set.

Preferably, the calculating sub-unit 5031 may be replaced by an obtaining sub-unit, and the first searching sub-unit 5032 may be replaced by a second searching sub-unit, and then the searching unit 503 includes:

an obtaining subunit, configured to obtain, from a decoder, prior information of bits corresponding to all layers in the bit-by-bit hierarchical tree structure;

the second searching subunit is used for not searching and expanding the layer corresponding to the node with the prior information being greater than the first threshold, extending the reserved path of the reserved node in the previous layer by using the symbol of the prior information of the current layer, expanding the node with the prior information being not greater than the first threshold, and searching M nodes with the maximum metric value as the reserved nodes of the current layer according to the magnitude of the metric value;

and the selecting subunit is used for sequentially searching layer by layer, selecting M retention nodes from the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

Preferably, the calculating sub-unit 5031 may be replaced by an obtaining sub-unit, and the first searching sub-unit 5032 may be replaced by a third searching sub-unit, and then the searching unit 503 includes:

a third searching subunit, configured to select, as metric values, M nodes with the largest metric values from the 2M nodes expanded by the M retention nodes on the upper layer, and use the M nodes with the largest metric values as retention nodes on the current layer, where the prior information weighted value of the node whose prior information is greater than the second threshold is used as the metric value;

The mapping mode of the constellation points of the quadrature amplitude modulation QAM transmission signals comprises Gray mapping or natural mapping.

It should be noted that, as will be understood by those skilled in the art, all or part of the processes in the methods of the above embodiments may be implemented by a computer program, which may be stored in a computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

The foregoing detailed description is directed to a method and apparatus for detecting signals in a mimo system, and the principles and embodiments of the present application are explained with reference to specific embodiments, which are merely used to help understand the method and its core ideas of the present application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.

Claims

1. A method for signal detection in a multiple-input multiple-output system, comprising:

converting constellation points of Quadrature Amplitude Modulation (QAM) transmitting signals into a bit vector weighted sum form to obtain QAM transmitting signals represented by a bit level, and converting a channel matrix into a composite channel matrix according to the QAM transmitting signals represented by the bit level;

performing QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and constructing a bit-by-bit hierarchical tree structure by using the upper triangular matrix;

searching the tree structure which is layered bit by bit on the basis of a breadth-first M algorithm layer by layer to obtain a signal candidate set, and calculating the metric value of each candidate signal in the signal candidate set;

and calculating posterior information of the transmitting signal by using the metric value.

2. The method of claim 1, wherein the bit-by-bit hierarchical tree structure is searched layer by layer based on a breadth-first M algorithm to obtain a signal candidate set, and calculating the metric value of the signal candidate set comprises:

calculating the metric values of 2M nodes on 2M branch paths expanded by M retention nodes of the previous layer in the current layer of the bit-by-bit hierarchical tree structure, wherein M is the number of constellation points;

searching M nodes with the maximum metric value from the 2M nodes, and taking the M nodes with the maximum metric value as retention nodes of the current layer;

and sequentially searching layer by layer, selecting M retention nodes from the 2M nodes of the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

3. The method of claim 1, wherein the bit-by-bit hierarchical tree structure is searched layer by layer based on a breadth-first M algorithm to obtain a signal candidate set, and calculating the metric value of the signal candidate set comprises:

in the tree structure of the bit-by-bit layering, acquiring prior information of corresponding bits of all layers from a decoder;

the method comprises the steps that a layer corresponding to a node with prior information larger than a first threshold value is not searched and expanded, a reserved path of a reserved node in the previous layer is expanded by using a symbol of prior information of the current layer, a node with prior information not larger than the first threshold value is expanded, M nodes with the maximum metric value are searched as reserved nodes of the current layer according to the magnitude of the metric value, and M is the number of constellation points;

and sequentially searching layer by layer, selecting M retention nodes from the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

4. The method of claim 1, wherein the bit-by-bit hierarchical tree structure is searched layer by layer based on a breadth-first M algorithm to obtain a signal candidate set, and calculating the metric value of the signal candidate set comprises:

using the prior information weighted value of the bit node with the prior information larger than the second threshold value as a metric value, searching M nodes with the maximum metric value from 2M nodes spread by M retention nodes on the upper layer, using the M nodes with the maximum metric value as the retention nodes on the current layer, wherein M is the number of constellation points;

5. The method according to any of claims 1-4, wherein the mapping manner of the constellation points of the QAM transmission signal comprises Gray mapping or natural mapping.

6. A signal detection apparatus in a multiple-input multiple-output system, comprising:

the conversion unit is used for converting the constellation points of the Quadrature Amplitude Modulation (QAM) sending signals into a bit vector weighted sum form to obtain QAM sending signals represented by a bit level, and obtaining a composite channel matrix according to the QAM sending signals represented by the bit level;

the decomposition unit is used for carrying out QR decomposition on the composite channel matrix to obtain an upper triangular matrix, and a bit-by-bit hierarchical tree structure is constructed by utilizing the upper triangular matrix;

the searching unit is used for searching the bit-by-bit hierarchical tree structure layer by layer based on a breadth-first M algorithm to obtain a signal candidate set and calculating the metric value of each candidate signal in the signal candidate set;

and the detection unit calculates the posterior information of the transmitting signal by using the metric value.

7. The apparatus of claim 6, wherein the search unit comprises:

a calculating subunit, configured to calculate, in a current layer of the bit-by-bit hierarchical tree structure, metric values of 2M nodes on 2M branch paths expanded by M surviving nodes of a previous layer, where M is a number of constellation points;

the first searching subunit is configured to search M nodes with the largest metric value from the 2M nodes, and use the M nodes with the largest metric value as retention nodes of a current layer;

and the selecting subunit is used for sequentially searching layer by layer, selecting M retention nodes from the 2M nodes of the last layer as a signal candidate set, and calculating the metric value of the signal candidate set.

8. The apparatus of claim 6, wherein the search unit comprises:

the second searching subunit is used for not performing expansion searching on a layer corresponding to a node of which the prior information is greater than the first threshold, directly extending a retention path of a retention node of a previous layer by using a symbol of the prior information of the current layer, expanding a node of which the prior information is not greater than the first threshold, and searching M nodes with the maximum metric value as retention nodes of the current layer according to the magnitude of the metric value, wherein M is the number of constellation points;

9. The apparatus of claim 6, wherein the search unit comprises:

a third searching subunit, configured to select, from 2M nodes spread by M surviving nodes on an upper layer, M nodes with a largest metric value as surviving nodes on a current layer, where M is a number of constellation points, by using a prior information weighted value of a node whose prior information is greater than a second threshold as the metric value;

10. The apparatus according to any of claims 6-9, wherein the mapping manner of the constellation points of the QAM transmission signal comprises gray mapping or natural mapping.