Ranking qr decomposition method for mimo system, fsd-based signal detection method and device

Abstract

The application provides a sorting QR decomposition method for a MIMO system, comprising: determining an initial matrix to be sorted according to a channel matrix of a MIMO system to be detected; determining column vectors to be sorted of the matrix and a column number thereof, denoted as a target number; if the target number is greater than a preset value, performing a first sorting strategy: in a first round of iteration, exchanging column orders of a column vector to be sorted with a smallest norm value and a first column vector to be sorted, and in each subsequent round of iteration, exchanging column orders of a column vector to be sorted with a second smallest norm value and a column vector to be sorted in a current round, to obtain a first target matrix; and a last column vector of the first target matrix corresponds to a highest layer traversal constellation point of an FSD algorithm. The above scheme can obtain an optimized target matrix. The application also provides an FSD-based signal detection method, which performs FSD signal detection by using the above optimized target matrix, so that the number of highest layer traversal constellation points is reduced, and the detection accuracy and reliability are still ensured.

Description

Technical Field

[0001] This invention relates to the field of communication technology, specifically to a sorting QR decomposition method for MIMO systems, a signal detection method and apparatus based on FSD, a storage medium, and a terminal. Background Technology

[0002] In Multiple-Input Multiple-Output (MIMO) communication systems, improving the reliability of signal detection at the receiver is a key issue in MIMO detection. In recent years, Sphere Decoding (SD) has gained widespread application due to its performance approaching that of Maximum Likelihood (ML) detection while having significantly lower computational complexity. In particular, the Fixed-complexity Sphere Decoding (FSD) algorithm has been extensively studied due to its superior performance and ease of hardware implementation.

[0003] Traditional FSD algorithms traverse all candidate constellation points at the highest level and treat all candidate constellation points as parent nodes, while in the remaining levels, only the constellation points with the closest Euclidean distance are retained as child nodes. This FSD algorithm must be combined with a unique Sorted QR Decomposition (SQRD) method to achieve its high performance gain. Specifically, SQRD is first performed on the channel matrix to be sorted to obtain the sorted decomposed matrix; then, the FSD algorithm is executed based on the sorted decomposed matrix for signal detection. SQRD is a method that embeds the stream sorting process into the QR decomposition (QR decomposition) process.

[0004] However, in existing SQRD methods, the sorting objective is usually to place the column vector with the smallest norm in the matrix to be sorted in the last column (the last column corresponds to the constellation point traversed at the highest level of the FSD algorithm), and the norms of the sorted first and second-to-last column vectors are incremented. Generally speaking, the norm of a column vector can be used to characterize the signal transmission quality of the corresponding channel; the larger the norm of the column vector, the higher the signal transmission quality of the corresponding channel tends to be.

[0005] However, since the traditional FSD algorithm still has high computational complexity in traversing all candidate constellation points at the highest level, it is necessary to reduce the number of constellation points traversed at the highest level to further reduce computational complexity. In this case, since the number of constellation points traversed at the highest level of the FSD algorithm (denoted as M) is less than the number of all candidate constellation points at the highest level (denoted as Q), SIf the existing SQRD method is used directly, the highest constellation point will correspond to the column vector with the smallest norm value (corresponding to the channel with the worst transmission quality). This may result in the inability to achieve the optimal detection performance of the FSD algorithm, thereby reducing the accuracy and reliability of signal detection. Summary of the Invention

[0006] The technical problem solved by the embodiments of the present invention is: how to obtain a more optimized target matrix by improving the sorting QR decomposition operation, so that when performing FSD signal detection based on the optimized target matrix, the detection accuracy and reliability can still be guaranteed even when the number of constellation points traversed at the highest level is reduced and the algorithm complexity is reduced.

[0007] To address the aforementioned technical problems, this invention provides a sorting QR decomposition method for MIMO systems, comprising the following steps: determining an initial sorting matrix based on the channel matrix of the MIMO system to be detected; determining the sequence vectors to be sorted in the initial sorting matrix, and determining the number of columns of the sequence vectors to be sorted, denoted as the target number; if the target number is greater than a preset value, then executing a first sorting strategy based on the initial sorting matrix and the sequence vectors to be sorted; the first sorting strategy includes: in a first iteration, swapping the column order of the sequence vector with the smallest norm in the initial sorting matrix with the first column of sequence vectors to be sorted, and then reordering the swapped columns. The ordered matrix is ​​subjected to QR decomposition to obtain the first decomposition matrix of the first round. In the t-th round of iteration, the second smallest norm value of the unsorted sequence vector from column t to the last column of the unsorted sequence vector in the (t-1)-th round is swapped with the unsorted sequence vector in column t, and the matrix after the column swap is subjected to QR decomposition to obtain the first decomposition matrix of the t-th round. The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the T-th round of iteration is used as the first target matrix after sorting and decomposition. The last column vector of the first target matrix corresponds to the highest level traversal constellation point of the FSD algorithm, 2≤t≤T, T is the number of targets, and t is a positive integer.

[0008] Optionally, determining the sequence vector to be sorted in the initial unsorted matrix includes: if the transmit antenna array of the MIMO system to be detected contains transmit antennas of interfering users and transmit antennas of the target user, then each column vector corresponding to the transmit antenna of the target user is used as the sequence vector to be sorted; if the transmit antenna array of the MIMO system to be detected does not contain transmit antennas of interfering users, then all column vectors in the initial unsorted matrix are used as the sequence vector to be sorted.

[0009] Optionally, the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of target users; after obtaining the first target matrix, the method further includes: performing modulation mode detection on the first target matrix to determine the maximum modulation mode of the interfering flow; if the maximum modulation mode of the interfering flow is less than or equal to the modulation mode of the target flow, then based on all column vectors of the first target matrix, a second sorting strategy is executed to obtain a second target matrix after sorting and decomposition.

[0010] Optionally, the second sorting strategy includes: in the first iteration, swapping the column vector with the smallest norm value in the first target matrix with the column vector of the first column, and performing QR decomposition on the matrix after swapping the column order to obtain the first target decomposition matrix of the first round; in the t-th iteration, swapping the column vector with the second smallest norm value among the column vectors from the t-th to the last column of the first target decomposition matrix of the (t-1)-th round with the column vector of the t-th column, and performing QR decomposition on the matrix after swapping the column order to obtain the first target decomposition matrix of the t-th round; until t=N T Stop iterating when the Nth iteration is complete. T The first objective decomposition matrix obtained in the first round is used as the second objective matrix; where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the first target matrix, where t is a positive integer.

[0011] Optionally, the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of the target user, and the sequence vectors to be sorted in the initial unsorted matrix are the column vectors corresponding to the transmit antennas of the target user; before executing the first sorting strategy, the method further includes: performing a preliminary sorting operation on the initial unsorted matrix; the preliminary sorting operation includes: placing all column vectors corresponding to the transmit antennas of interfering users on the first side of the initial unsorted matrix, and placing all column vectors corresponding to the transmit antennas of the target user on the second side of the initial unsorted matrix; wherein, the last column vector on the second side of the matrix obtained after performing the preliminary sorting operation is the column vector corresponding to the highest-level traversal constellation point of the FSD algorithm.

[0012] Optionally, the method further includes: if the target number is less than or equal to the preset value, then a third sorting strategy is executed based on the initial unsorted matrix and the unsorted sequence vectors; the third sorting strategy includes: in the first iteration, swapping the column order of the unsorted sequence vector with the second smallest norm value in the initial unsorted matrix with the first column unsorted sequence vector, and performing QR decomposition on the matrix after swapping the column order to obtain the second decomposition matrix of the first round; in the t-th iteration, swapping the column order of the unsorted sequence vector with the second smallest norm value in the unsorted sequence vectors from the t-th column to the last column of the second decomposition matrix of the t-1-th round with the t-th column unsorted sequence vector, and performing QR decomposition on the matrix after swapping the column order to obtain the second decomposition matrix of the t-th round; until the value of t reaches T, the iteration stops, and the second decomposition matrix obtained in the T-th iteration is used as the third target matrix after sorting and decomposition.

[0013] Optionally, the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of target users; after obtaining the third target matrix, the method further includes: performing modulation mode detection on the third target matrix to determine the maximum modulation mode of the interfering flow; if the maximum modulation mode of the interfering flow is less than or equal to the modulation mode of the target flow, then based on all column vectors of the third target matrix, a fourth sorting strategy is executed to obtain the sorted and decomposed fourth target matrix.

[0014] Optionally, the fourth sorting strategy includes: in the first iteration, swapping the column vector with the smallest norm value in the third target matrix with the column vector of the first column, and performing QR decomposition on the matrix after swapping the column order to obtain the second target decomposition matrix of the first round; in the t-th iteration, swapping the column order of the second target decomposition matrix of the second target decomposition matrix of the (t-1)-th round with the column vector of the second target decomposition matrix of the second target decomposition matrix of the t-th round, and performing QR decomposition on the matrix after swapping the column order to obtain the second target decomposition matrix of the t-th round; until t=N T Stop the iteration and set N T The second objective decomposition matrix obtained in the first round is used as the fourth objective matrix; where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the third target matrix, where t is a positive integer.

[0015] Optionally, an initial unsorted matrix is ​​determined based on the channel matrix of the MIMO system to be detected, including: constructing an augmented matrix based on the channel matrix of the MIMO system to be detected, the total number of transmit antennas, the total number of receive antennas, and the channel noise, wherein the augmented matrix serves as the initial unsorted matrix.

[0016] This invention also provides a signal detection method based on FSD, characterized by comprising: determining the product of the sorted decomposition target matrix and the transmitted signal vector, and summing the product with channel noise to obtain the signal to be detected, wherein the sorted decomposition target matrix is ​​determined by the sorted QR decomposition method for MIMO systems described above; and performing hard decision on the signal to be detected using the FSD algorithm to determine the detection result; wherein the highest level of the FSD algorithm traverses M constellation points, including the hard decision point, and the remaining layers retain only the constellation point with the closest Euclidean distance as child nodes; M≤Q S Q S This represents the total number of all candidate constellation points at the highest level.

[0017] This invention also provides a sorting QR decomposition device for a MIMO system, comprising: an initial unsorted matrix determination module, configured to determine an initial unsorted matrix based on the channel matrix of the MIMO system to be detected; an unsorted sequence vector determination module, configured to determine unsorted sequence vectors in the initial unsorted matrix and determine the number of columns of the unsorted sequence vectors, denoted as the target number; and a first sorting strategy execution module, configured to execute a first sorting strategy based on the initial unsorted matrix and the unsorted sequence vectors if the target number is greater than a preset value; the first sorting strategy includes: in a first iteration, selecting the unsorted sequence vector with the smallest norm in the initial unsorted matrix and the first column of unsorted sequence vectors. The columns of the vectors are swapped, and the matrix after the swap is decomposed using QR decomposition to obtain the first decomposition matrix of the first round. In the t-th iteration, the second smallest norm vector among the vectors to be sorted from the t-th column to the last column of the first decomposition matrix of the (t-1)-th round is swapped with the vector to be sorted in the t-th column, and the matrix after the swap is decomposed using QR decomposition to obtain the first decomposition matrix of the t-th round. The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the T-th iteration is used as the first target matrix after sorting and decomposition. The last column vector of the first target matrix corresponds to the highest level traversal constellation point of the FSD algorithm, 2≤t≤T, T is the number of targets, and t is a positive integer.

[0018] This invention also provides a signal detection device based on FSD, comprising: a signal determination module for determining the product of a target matrix and a transmitted signal vector, and summing the product with channel noise to obtain the signal to be detected, wherein the target matrix is ​​determined using the sorted QR decomposition method for MIMO systems as described in any one of claims 1 to 9; and a signal detection module for performing hard decision on the signal to be detected using the FSD algorithm to determine the detection result; wherein the highest level of the FSD algorithm traverses M constellation points, including the hard decision point, and the remaining layers retain only the constellation point with the closest Euclidean distance as child nodes; M ≤ Q S Q S This represents the total number of all candidate constellation points at the highest level.

[0019] This invention also provides a storage medium storing a computer program, which, when executed by a processor, performs the steps of the sorting QR decomposition method for MIMO systems described above, or performs the steps of the signal detection method based on FSD described above.

[0020] This invention also provides a terminal, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it executes the steps of the sorting QR decomposition method for MIMO systems described above, or executes the steps of the signal detection method based on FSD described above.

[0021] Compared with the prior art, the technical solution of the embodiments of the present invention has the following beneficial effects:

[0022] In this embodiment of the invention, when the number of columns T of the sequence vectors to be sorted in the initial matrix to be sorted is greater than the preset value, by executing the first sorting strategy, in T iterations, except for the first iteration which swaps the column order of the sequence vector to be sorted with the first column of sequence vectors, each subsequent iteration swaps the column order of the sequence vector to be sorted with the second smallest norm value. Thus, after completing T iterations, the first target matrix obtained after sorting decomposition has at least the following characteristics: (1) In the first target matrix, among the sorted column vectors corresponding to the sequence vectors to be sorted, the norm value of the last sorted column vector is the second smallest (i.e., the second smallest) among all sorted column vectors; (2) The first sorted column vector is the smallest among all sorted column vectors; (3) The norm values ​​of the sorted first column vector to the second-to-last sorted column vector show a progressively increasing trend.

[0023] Furthermore, in FSD-based signal detection methods, to reduce algorithm complexity, the number of constellation points traversed at the highest level needs to be reduced. For example, instead of traversing all candidate constellation points at the highest level, the traditional FSD algorithm traverses only a portion of the candidate constellation points (i.e., the number of constellation points traversed at the highest level is less than the total number of candidate constellation points at the highest level). In this case, if the existing conventional sorted QR decomposition method is continued, i.e., the norm of the last column vector of the sorted decomposition matrix is ​​the "smallest" norm among all sorted column vectors (corresponding to the channel with the worst transmission reliability), it may lead to a decrease in the performance of subsequent FSD signal detection.

[0024] Therefore, to ensure optimal performance of the FSD algorithm in signal detection, the sorted QR decomposition method provided in this embodiment of the invention can obtain a more optimized first target matrix. This ensures that the norm of the column vector corresponding to the constellation points traversed at the highest level of the FSD algorithm (i.e., the last column vector of the first target matrix) is not the "smallest" norm among all sorted column vectors of the first target matrix, but rather the "second smallest" norm (corresponding to relatively more reliable channel transmission). Furthermore, compared to existing sorted QR decomposition schemes, the first target matrix obtained using this embodiment, when subsequently applied to FSD detection, can still maintain detection accuracy and reliability while reducing the complexity of the detection algorithm, even if the number of constellation points traversed at the highest level is reduced.

[0025] Furthermore, in this embodiment of the invention, when the maximum modulation scheme of the interfering flow is less than or equal to the modulation scheme of the target flow, it means that the degree of signal interference is within an acceptable range. Therefore, based on the first sorting strategy, which only performs sorting QR decomposition on the "column vectors corresponding to the target flow" to obtain the first target matrix, all column vectors of the first target matrix can be used as the sequence vectors to be sorted, and the second sorting strategy can be executed to perform a second sorting QR decomposition on both the "column vectors corresponding to the target flow" and the "column vectors corresponding to the interfering flow" (i.e., perform a second sorting QR decomposition on both the target flow and the interfering flow) to further optimize the sorting QR decomposition result and obtain a more optimized target matrix (i.e., the second target matrix). Attached Figure Description

[0026] Figure 1 This is a flowchart of the first sorting QR decomposition method for MIMO systems in this embodiment of the invention;

[0027] Figure 2 This is a partial flowchart of the second sorting QR decomposition method for MIMO systems in this embodiment of the invention;

[0028] Figure 3This is a flowchart of the third sorting QR decomposition method for MIMO systems in this embodiment of the invention;

[0029] Figure 4 This is a partial flowchart of the fourth sorting QR decomposition method for MIMO systems in this embodiment of the invention;

[0030] Figure 5 This is a schematic diagram of the tree search structure using the FSD algorithm in an embodiment of the present invention;

[0031] Figure 6 This is a schematic diagram of the structure of a sorting QR decomposition device for a MIMO system according to an embodiment of the present invention;

[0032] Figure 7 This is a schematic diagram of the structure of a signal detection device based on FSD in an embodiment of the present invention. Detailed Implementation

[0033] To make the above-mentioned objectives, features and beneficial effects of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0034] Reference Figure 1 , Figure 1 This is a flowchart of the first sorting QR decomposition method for MIMO systems in this embodiment of the invention. The method may include steps S11 to S13:

[0035] Step S11: Determine the initial sorting matrix based on the channel matrix of the MIMO system to be detected;

[0036] Step S12: Determine the sequence vector to be sorted in the initial matrix to be sorted, and determine the number of columns of the sequence vector to be sorted, denoted as the target number;

[0037] Step S13: If the target quantity is greater than a preset value, then execute the first sorting strategy based on the initial unsorted matrix and the unsorted sequence vector.

[0038] The first sorting strategy includes:

[0039] In the first iteration, the column order of the sequence vector with the smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector to be sorted, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the first round.

[0040] In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the first decomposition matrix from the t-th column to the last column of the sequence vector to be sorted is swapped with the sequence vector to be sorted in the t-th column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the t-th iteration.

[0041] The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the Tth iteration is used as the first target matrix after sorting and decomposition.

[0042] Wherein, the last column vector of the first target matrix corresponds to the highest level of constellation traversal in the FSD algorithm, 2≤t≤T, where T is the number of targets and t is a positive integer.

[0043] In the specific implementation of step S11, a Multiple-In Multiple-Out (MIMO) system is an antenna system that uses multiple antennas at both the transmitting and receiving ends to greatly improve channel capacity, forming multiple channels between transmission and reception. MIMO technology can significantly improve the capacity of communication systems, thereby achieving high data transmission rates and high spectrum utilization.

[0044] Specifically, the initial matrix to be sorted can be determined in the following way: based on the channel matrix of the MIMO system to be detected, the total number of transmit antennas, the total number of receive antennas, and the channel noise, an augmented matrix is ​​constructed, and the augmented matrix is ​​used as the initial matrix to be sorted.

[0045] Specifically, for the currently received signal Y = H × s + v, the initial matrix to be sorted can be determined using the following formula:

[0046]

[0047] Where Y represents the transmitted signal vector, s represents the received signal vector, and H represents the channel matrix of the MIMO system to be detected. N represents the augmented matrix, i.e., the initial matrix to be sorted. T N represents the total number of transmitting antennas. R σ represents the total number of receiving antennas, v represents the channel noise, and σ represents the total number of receiving antennas. v Let I represent the mean squared error of the channel noise, I represent the identity matrix, and C represent the matrix dimension.

[0048] In specific implementations, other appropriate methods can also be used to determine the initial matrix to be sorted. For example, the original channel matrix of the MIMO system to be detected can be directly used as the initial matrix to be sorted.

[0049] In the specific implementation of step S12, the sequence vectors to be sorted in the initial matrix to be sorted can be different in different application scenarios. Specifically, in some application scenarios (e.g., SU-MIMO system), the sequence vectors to be sorted can be all column vectors of the initial matrix to be sorted; while in other application scenarios (e.g., MU-MIMO system), the sequence vectors to be sorted can be some column vectors of the initial matrix to be sorted.

[0050] In a first specific implementation, if the transmit antenna array of the MIMO system to be detected does not contain transmit antennas that interfere with users, then all column vectors in the initial unsorted matrix are used as the unsorted sequence vector.

[0051] It should be noted that in a Single User Multiple Input Multiple Output (SU-MIMO) system, only one user is connected to the base station at any given time. For example, a common uplink scenario is where a user uses two antennas to transmit signals to the base station, and the base station only receives that user's signal at any given time; this is a typical SU-MIMO scenario. Therefore, if the MIMO system to be detected is a SU-MIMO system, that is, the communication system contains only the target user and not interfering users, then in step S12, all column vectors in the initial unsorted matrix can be used as the unsorted sequence vector.

[0052] In a second specific implementation, if the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of the target user, then each column vector corresponding to the transmit antenna of the target user is used as the sequence vector to be sorted.

[0053] It should be noted that in a Multiple User Multiple Input Multiple Output (MU-MIMO) system, the number of users communicating with the base station at the same time can be multiple. For example, multiple users sharing bandwidth resources for signal transmission, and the base station receiving signals from multiple users at the same time, is a typical MU-MIMO. Since MU-MIMO often includes not only target users (corresponding to the target stream) but also interfering users (corresponding to the interfering stream), it is necessary to detect the modulation scheme of the interfering users. Therefore, if the MIMO system to be detected is an MU-MIMO system, in step S12, the column vectors of the transmit antennas corresponding to the target users in the initial unsorted matrix (i.e., the column vectors corresponding to the target stream) can be used as the unsorted sequence vector.

[0054] Furthermore, based on the first target matrix obtained through a sorted QR decomposition, subsequent steps can be based on the detection results of the modulation mode of the first target matrix to select whether to perform a second sorted QR decomposition on all column vectors of the first target matrix (i.e., perform a second sorted QR decomposition on the target flow and interference flow together).

[0055] In the specific implementation of step S13, the target quantity is compared with a preset value. If the former is greater than the latter, the first sorting strategy is executed based on the initial unsorted matrix and the unsorted sequence vector.

[0056] The preset values ​​can be appropriately set according to the needs of the actual application scenario. For example, they can be determined by factors such as the number of transmitting antennas of the MIMO system to be detected, the number of columns of the sequence vector to be sorted, and the accuracy requirements of the subsequent FSD algorithm detection.

[0057] In one specific embodiment, the preset value can be 2. That is, when the preset value is 2, the first sorting strategy is executed on the premise that the number of columns of the sequence vector to be sorted in the initial matrix to be sorted (i.e., the target number) is at least 3.

[0058] In the first sorting strategy, T rounds of iterative operations are performed. In the first round, the column vectors with the "smallest" norm value are swapped in column order. In subsequent rounds, the column vectors with the "second smallest" norm value are swapped in column order. The total number of iterations T is equal to the number of columns of the sequence vectors to be sorted in the initial matrix to be sorted (i.e., the target number).

[0059] Without limitation, in each iteration, the norm value of the column vector can be the result of squaring the L2 norm of the column vector. Alternatively, other norm operations can be used, as long as they can characterize the signal transmission quality (or signal transmission reliability) of the channel corresponding to the column vector.

[0060] In each iteration, the method for performing QR decomposition on the matrix after swapping column order can be any existing conventional QR decomposition method, such as, but not limited to, Schmidt orthogonalization, Householder transformation, and Givens transformation.

[0061] It should be noted that the positional relationship between the first column vector and the last column vector of each matrix in this application is usually from left to right, that is, the first column vector is located on the leftmost side of the matrix and the last column vector is located on the rightmost side of the matrix. However, it is not limited to this. In specific implementations, it can be transformed according to actual needs. For example, the positional relationship between the first column vector and the last column vector can be from right to left.

[0062] However, it is necessary to ensure that the last column vector of each target matrix obtained after sorting and decomposition corresponds to the highest-level constellation point traversed by the FSD algorithm. More specifically, taking the first target matrix as an example, the last column vector to the first column vector correspond one-to-one with the highest-level constellation point to the lowest-level constellation point traversed by the FSD algorithm; that is, each column vector corresponds to a single-level constellation point traversed. The last column vector corresponds to the highest-level constellation point traversed by the FSD algorithm, and the first column vector corresponds to the lowest-level constellation point traversed.

[0063] In this embodiment of the invention, when the number of columns T of the sequence vectors to be sorted in the initial matrix to be sorted is greater than the preset value, by executing the first sorting strategy, in T iterations, except for the first iteration which swaps the column order of the sequence vector to be sorted with the first column of sequence vectors, each subsequent iteration swaps the column order of the sequence vector to be sorted with the second smallest norm value. Thus, after completing T iterations, the first target matrix obtained after sorting decomposition has at least the following characteristics: (1) In the first target matrix, among the sorted column vectors corresponding to the sequence vectors to be sorted, the norm value of the last sorted column vector is the second smallest norm value among all sorted column vectors; (2) The sorted first column vector is the smallest norm value among all sorted column vectors; (3) The norm values ​​of the sorted first column vector to the second-to-last sorted column vector show a progressively increasing trend.

[0064] Furthermore, in FSD-based signal detection methods, to reduce algorithm complexity, the number of constellation points traversed at the highest level needs to be reduced. For example, instead of traversing all candidate constellation points at the highest level, the traditional FSD algorithm traverses only a portion of the candidate constellation points (i.e., the number of constellation points traversed at the highest level is less than the total number of candidate constellation points at the highest level). In this case, if the existing conventional sorted QR decomposition method is continued, i.e., the norm of the last column vector of the sorted decomposition matrix is ​​the "smallest" norm among all sorted column vectors (corresponding to the channel with the worst transmission reliability), it may lead to a decrease in the performance of subsequent FSD signal detection.

[0065] Therefore, to ensure optimal performance of the FSD algorithm in signal detection, the sorted QR decomposition method provided in this embodiment of the invention can obtain a more optimized first target matrix. This ensures that the norm of the column vector corresponding to the constellation points traversed at the highest level of the FSD algorithm (i.e., the last column vector of the first target matrix) is not the "smallest" norm among all sorted column vectors of the first target matrix, but rather the "second smallest" norm (corresponding to relatively more reliable channel transmission). Compared to existing sorted QR decomposition schemes, the first target matrix obtained using this embodiment, even with a reduced number of constellation points traversed at the highest level, can still maintain detection accuracy and reliability while reducing the complexity of the detection algorithm during subsequent application in FSD detection.

[0066] Furthermore, in a second specific embodiment of step S12 above (for example, the MIMO system to be detected is a MU-MIMO system), the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of the target user, and the sequence vectors to be sorted in the initial unsorted matrix are the column vectors corresponding to the transmit antennas of the target user. In this case, before executing the first sorting strategy in step S13, a preliminary sorting operation may also be performed on the initial unsorted matrix.

[0067] Specifically, the preliminary sorting operation includes: placing all column vectors corresponding to the transmitting antennas of the interfering user on the first side of the initial matrix to be sorted, and placing all column vectors corresponding to the transmitting antennas of the target user on the second side of the initial matrix to be sorted; wherein, the last column vector on the second side of the matrix obtained after performing the preliminary sorting operation is the column vector corresponding to the highest-level traversal constellation point of the FSD algorithm.

[0068] In this context, the first side of the initial matrix to be sorted typically refers to the left side of the matrix, and the second side refers to the right side. Alternatively, the order can be reversed depending on the specific scenario, for example, the first side refers to the right side and the second side to the left side. However, it is essential to ensure that the last column vector of the second side of the matrix obtained after the initial sorting operation corresponds to the column vector of the highest-level constellation point traversed by the FSD algorithm. More specifically, the last column of the second side to the first column of the first side of the matrix obtained by performing the above initial sorting operation correspond one-to-one with the highest-level constellation point to the lowest-level constellation point of the FSD algorithm.

[0069] As a non-limiting embodiment, it can be adopted This represents the matrix obtained after performing a preliminary sorting operation on the initial matrix to be sorted. Here, the first column vector μ1′ to the nth column vector μ... nLocated on the first side (i.e., the left side) of the matrix, each corresponding to a transmitting antenna of the interfering user; the (n+1)th column vector μ n+1 ′ to the Nth T column vector Located on the second side (i.e., the right side) of the matrix, each corresponds to a transmit antenna of the target user.

[0070] In this embodiment of the invention, when the initial matrix to be sorted contains both column vectors corresponding to the transmitting antennas of the interfering user (or "column vectors corresponding to the target flow") and column vectors corresponding to the transmitting antennas of the interfering user (or "column vectors corresponding to the interference flow"), since the first sorting strategy only performs sorting QR decomposition on the "column vectors corresponding to the target flow", before executing the first sorting strategy, the "column vectors corresponding to the target flow" and the "column vectors corresponding to the interference flow" can be concentrated on both sides of the matrix through the above preliminary sorting operation.

[0071] Therefore, on the one hand, the complexity of subsequent sorting operations can be reduced, and the probability of errors can be decreased. On the other hand, since the second side is the side where the column vectors corresponding to the highest-level traversal constellation points of the FSD algorithm are located, compared with performing the above preliminary sorting operation, by placing all the "column vectors corresponding to the target flow" on the second side, the problem of increasing the signal detection error based on FSD due to mistakenly placing the "column vectors corresponding to the interference flow" in the last column during subsequent iterative sorting operations can be avoided.

[0072] Reference Figure 2 , Figure 2 This is a partial flowchart of the second sorting QR decomposition method for MIMO systems in this embodiment of the invention. In this embodiment, the transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users (corresponding to the interfering flow) and transmit antennas of the target user (corresponding to the target flow). As mentioned above, this situation typically occurs when the MIMO system to be detected is a MU-MIMO system.

[0073] Specifically, since MU-MIMO systems often include not only the target user but also interfering users, Figure 1 In the illustrated embodiment, for a MIMO system containing a target user and interfering users, only the column vectors of the transmit antennas corresponding to the target user are sorted; the column vectors of the transmit antennas corresponding to the interfering users are not sorted. Therefore, it is possible to base it on... Figure 1 The first target matrix obtained in the illustrated embodiment is further subjected to modulation mode detection, and based on the result of modulation mode detection, it is determined whether it is necessary to re-sort the target flow and interference flow together to obtain a more optimized target matrix.

[0074] Specifically, the second sorting QR decomposition method for MIMO systems may include Figure 1 Steps S11 to S13 may further include steps S21 to S22, wherein step S21 is performed after obtaining the first target matrix through step S13. The following is for... Figure 1 The different parts of the text will be explained.

[0075] In step S21, the modulation mode of the first target matrix is ​​detected to determine the maximum modulation mode of the interference flow.

[0076] In practice, existing conventional methods can be used for modulation mode detection, which will not be elaborated here.

[0077] The modulation schemes of the interfering stream and the target stream can be identified or indicated using preset parameter values, index numbers / serial numbers, or other methods. If the maximum modulation scheme of the interfering stream is less than or equal to the modulation scheme of the target stream, it means that the parameter value identifying the maximum modulation scheme of the interfering stream is less than or equal to the parameter value identifying the modulation scheme of the target stream; or, the index number indicating the maximum modulation scheme of the interfering stream is located before the index number indicating the modulation scheme of the target stream.

[0078] In step S22, if the maximum modulation mode of the interference stream is less than or equal to the modulation mode of the target stream, then based on all column vectors of the first target matrix, a second sorting strategy is executed to obtain the second target matrix after sorting and decomposition.

[0079] In this embodiment of the invention, when the maximum modulation scheme of the interfering flow is less than or equal to the modulation scheme of the target flow, it means that the degree of signal interference is within an acceptable range. Therefore, based on the first sorting strategy, which only performs sorting QR decomposition on the "column vectors corresponding to the target flow" to obtain the first target matrix, all column vectors of the first target matrix can be used as the sequence vectors to be sorted, and a second sorting strategy can be executed to perform a second sorting QR decomposition on both the "column vectors corresponding to the target flow" and the "column vectors corresponding to the interfering flow" (i.e., perform a second sorting QR decomposition on both the target flow and the interfering flow) to further optimize the sorting QR decomposition result and obtain a more optimized target matrix (i.e., the second target matrix).

[0080] Further, the second sorting strategy may include: in the first iteration, swapping the column vector with the smallest norm value in the first target matrix with the column vector of the first column, and performing QR decomposition on the matrix after swapping the column order to obtain the first target decomposition matrix of the first round; in the t-th iteration, swapping the column vector with the second smallest norm value among the column vectors from the t-th to the last column of the first target decomposition matrix of the (t-1)-th round with the column vector of the t-th column, and performing QR decomposition on the matrix after swapping the column order to obtain the first target decomposition matrix of the t-th round; until t=N T Stop iterating when the Nth iteration is complete. T The first objective decomposition matrix obtained in the first round is used as the second objective matrix; where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the first target matrix, where t is a positive integer.

[0081] It should be noted that the second sorting strategy described above is similar in steps and principle to the first sorting strategy. Specifically, except that in the first iteration, the column vector with the "smallest" norm value is swapped with the first column vector, subsequent iterations swap the column vector with the "second smallest" norm value with the current round's sequence vector (i.e., a sorting QR decomposition strategy of "smallest in the first round + second smallest in subsequent rounds"). Therefore, after executing N... T After rounds of iteration, the second target matrix obtained after sorting and decomposition has at least the following characteristics: (1) In the second target matrix, the norm value of the last column vector is the second smallest among all column vectors; (2) The first column vector has the smallest norm value among all column vectors; (3) The norm values ​​of the first column vector to the second-to-last column vector show a progressively increasing trend.

[0082] In this embodiment of the invention, by using the second sorting strategy described above to obtain the second target matrix, in the subsequent FSD signal detection process, the norm value of the column vector corresponding to the constellation point traversed at the highest level (i.e., the last column vector of the second target matrix) is not the "smallest" norm value among all column vectors of the second target matrix, but the "second smallest" norm value (corresponding to relatively reliable channel transmission). Based on this, even if the number of constellation points traversed at the highest level of FSD is reduced, it is still possible to ensure detection accuracy and reliability while reducing the complexity of the detection algorithm.

[0083] It should be noted that, as a preferred embodiment, the second sorting strategy described above can also be implemented in practice by exchanging the column order of the column vector with the second smallest norm value with the current round's sequence vector in all iterations from the first to the last round.

[0084] Furthermore, if the maximum modulation scheme of the interference stream is greater than that of the target stream, the first target matrix can be directly used as the final target matrix after sorting and decomposition. That is, the first target matrix can be directly applied to the subsequent FSD-based signal detection process.

[0085] In this embodiment of the invention, if the maximum modulation scheme of the interfering flow is greater than that of the target flow, it means that the signal is heavily interfered with. Therefore, based on the first sorting strategy, which only sorts the column vectors corresponding to the target flow to obtain the first target matrix, it is not advisable to execute the second sorting strategy. Instead, the first target matrix can be directly applied to perform subsequent FSD signal detection. In other words, if the maximum modulation scheme of the interfering flow is greater than that of the target flow, and the second sorting strategy is still executed (i.e., sorting the target flow and the interfering flow together again), the column vectors corresponding to the interfering flow may end up in the last column of the resulting matrix. This will increase the error propagation and accumulation in subsequent FSD-based signal detection operations, reducing the accuracy of signal detection.

[0086] Reference Figure 3 , Figure 3 This is a flowchart of a third sorting QR decomposition method for MIMO systems in this embodiment of the invention. The third sorting QR decomposition method for MIMO systems may include steps S31 to S33.

[0087] Among them, steps S31 to S32 and Figure 1 The embodiment shown is the same as, and Figure 1 The difference in the illustrated embodiment is that, Figure 3 In the illustrated embodiment, the number of columns (i.e., the target number) of the sequence vector to be sorted is less than or equal to the target value, and therefore the sorting strategy (third sorting strategy) executed is different.

[0088] In step S31, the initial sorting matrix is ​​determined based on the channel matrix of the MIMO system to be detected.

[0089] In step S32, the sequence vector to be sorted in the initial matrix to be sorted is determined, and the number of columns of the sequence vector to be sorted is determined, denoted as the target number.

[0090] In step S33, if the target quantity is less than or equal to the preset value, a third sorting strategy is executed based on the initial unsorted matrix and the unsorted sequence vector.

[0091] The third sorting strategy includes:

[0092] In the first iteration, the column order of the sequence vector with the second smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the second decomposition matrix of the first round.

[0093] In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the t-th to last column of the second decomposition matrix of the (t-1)-th iteration is swapped with the t-th column sequence vector, and the matrix after the swap is decomposed into QR decomposition to obtain the second decomposition matrix of the t-th iteration.

[0094] The iteration stops when the value of t reaches T, and the second decomposition matrix obtained in the Tth iteration is used as the third target matrix after sorting and decomposition.

[0095] As mentioned above, in a non-limiting embodiment, the preset value can be 2. That is, when the preset value is 2, the third sorting strategy is executed on the premise that the number of columns of the sequence vector to be sorted in the initial matrix to be sorted is equal to 2. Cases with fewer than 2 columns do not require sorting and can therefore be disregarded.

[0096] The specific difference between the third sorting strategy and the first sorting strategy lies in the following: the first sorting strategy uses the column vector with the "smallest" norm value to exchange column order with the first round of the sequence vector to be sorted in the first iteration, and uses the column vector with the "second smallest" norm value to exchange column order with the current round of the sequence vector to be sorted in each subsequent iteration (i.e., a sorting QR decomposition strategy of "smallest in the first round + second smallest in each subsequent round"); while the third sorting strategy uses the column vector with the "second smallest" norm value to exchange column order with the current round of the sequence vector to be sorted in each iteration (i.e., a sorting QR decomposition strategy of "second smallest" in each round).

[0097] In this embodiment of the invention, when the number of columns T of the sequence vectors to be sorted in the initial matrix to be sorted is less than or equal to the preset value, by executing the above-mentioned sorting QR decomposition strategy of "second smallest" in each round, the obtained third target matrix has at least the following characteristics: (1) In the third target matrix, among the sorted column vectors corresponding to the sequence vectors to be sorted, the norm value of the last sorted column vector is the "smallest" norm value among all sorted column vectors; (2) The sorted first column vector is the "second smallest" norm value among all sorted column vectors; (3) The norm values ​​of the sorted first column vector to the second-to-last sorted column vector show a column-by-column increasing trend.

[0098] Understandably, the FSD algorithm traverses at least a portion of candidate constellation points at the highest level, even if it doesn't traverse all of them, typically involving multiple constellation points. In contrast, other levels only retain the constellation point with the closest Euclidean distance as its child nodes, usually involving only a single constellation point. In the FSD algorithm, the fewer constellation points traversed at a given level, the larger the norm of the corresponding column vector should be. Therefore, if T is less than or equal to the preset value, it means the FSD algorithm tends to traverse fewer constellation point levels (e.g., when T=2, there are only two levels of constellation points). In scenarios where the total number of constellation point levels traversed is small, having the highest-level constellation point correspond to the column vector with the "smallest" norm, and the remaining levels corresponding to column vectors with relatively larger norms, helps achieve better detection performance.

[0099] about Figure 3 For more details on the embodiments shown, please refer to the foregoing and Figure 1 The relevant content in the document will be executed, and will not be elaborated here.

[0100] Reference Figure 4 , Figure 4 This is a partial flowchart of the fourth sorting QR decomposition method for MIMO systems in this embodiment of the invention. The fourth sorting QR decomposition method for MIMO systems may include... Figure 3 Steps S31 to S33 may further include steps S41 to S42, wherein step S41 is executed after obtaining the third target matrix by executing step S33.

[0101] The following are related to Figure 3 The different parts of the text will be explained.

[0102] In step S41, the modulation mode of the third target matrix is ​​detected to determine the maximum modulation mode of the interference flow.

[0103] In step S42, if the maximum modulation scheme of the interference stream is less than or equal to the modulation scheme of the target stream, then based on all column vectors of the third target matrix, the fourth sorting strategy is executed to obtain the sorted and decomposed fourth target matrix.

[0104] The fourth sorting strategy may include: in the first iteration, swapping the column vector with the smallest norm in the third target matrix with the column vector of the first column, and performing QR decomposition on the matrix after the swap to obtain the second target decomposition matrix of the first round; in the t-th iteration, swapping the column vector with the second smallest norm in the column t to the last column of the second target decomposition matrix of the (t-1)-th round with the column vector of the t-th column, and performing QR decomposition on the matrix after the swap to obtain the second target decomposition matrix of the t-th round; until t = NT Stop the iteration and set N T The second objective decomposition matrix obtained in the first round is used as the fourth objective matrix; where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the third target matrix, where t is a positive integer.

[0105] Furthermore, if the maximum modulation scheme of the interference stream is greater than that of the target stream, the third target matrix can be directly used as the final target matrix after sorting and decomposition. That is, the third target matrix can be directly applied to the subsequent FSD-based signal detection process.

[0106] The specific execution process, principle, and technical effects of the fourth sorting strategy are the same as those of the second sorting strategy mentioned above. Therefore, you can refer to the relevant content of the second sorting strategy, and it will not be repeated here.

[0107] This invention also provides a signal detection method based on FSD, comprising: determining the product of a sorted and decomposed target matrix and a transmitted signal vector, and summing the product with channel noise to obtain the signal to be detected, wherein the sorted and decomposed target matrix is ​​obtained by using... Figures 1 to 4 The sorting QR decomposition method for MIMO systems described in any embodiment is used; the FSD algorithm is employed to perform hard decision on the signal to be detected to determine the detection result (e.g., log-likelihood ratio (LLR)); wherein, the highest level of the FSD algorithm traverses M constellation points, including the hard decision point, and the remaining levels only retain the constellation point with the closest Euclidean distance as child nodes; M≤Q S Q S This represents the total number of all candidate constellation points at the highest level.

[0108] like Figure 5 As shown, Figure 5 This is a schematic diagram of the tree search structure using the FSD algorithm in this embodiment of the invention. During the traversal, the highest level (or top level) traverses M constellation points, including the hard-determined points, while the remaining levels only retain the constellation point with the closest Euclidean distance as its child nodes. Figure 5 The partial hierarchical structure of the partial traversal and the partial constellation points of each level of traversal are shown only schematically.

[0109] Specifically, the signal to be detected can be determined using the following formula: z = R c s+v;

[0110] Where z represents the signal to be detected, R c express Figures 1 to 4The target matrix (i.e., any one of the first to fourth target matrices) determined by the sorted QR decomposition method for MIMO systems described in any embodiment, where s represents the transmitted signal vector and v represents the channel noise.

[0111] It should be noted that, although in the above-mentioned FSD-based signal detection method, the highest layer does not traverse all candidate constellation points, that is, the number of constellation points M traversed by the highest layer is less than or equal to the total number of candidate constellation points Q of all higher layers. S However, this application Figures 1 to 4 The sorted QR decomposition method for MIMO systems provided in any embodiment can also be adapted to it. Specifically, by using the more optimized target matrix obtained by any of the sorted QR decomposition methods for MIMO systems provided in this application, the signal to be detected is determined and FSD signal detection is performed, which can reduce the complexity of the detection algorithm while ensuring detection accuracy and reliability.

[0112] Reference Figure 6 , Figure 6 This is a schematic diagram of a sorting QR decomposition device for a MIMO system according to an embodiment of the present invention. The sorting QR decomposition device for a MIMO system may include:

[0113] The initial unsorted matrix determination module 61 is used to determine the initial unsorted matrix based on the channel matrix of the MIMO system to be detected.

[0114] The unsorted sequence vector determination module 62 is used to determine the unsorted sequence vector in the initial unsorted matrix, and to determine the number of columns of the unsorted sequence vector, denoted as the target number;

[0115] The first sorting strategy execution module 63 is used to execute a first sorting strategy based on the initial unsorted matrix and the unsorted sequence vector if the target number is greater than a preset value.

[0116] The first sorting strategy includes:

[0117] In the first iteration, the column order of the sequence vector with the smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector to be sorted, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the first round.

[0118] In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the first decomposition matrix from the t-th column to the last column of the sequence vector to be sorted is swapped with the sequence vector to be sorted in the t-th column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the t-th iteration.

[0119] The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the Tth iteration is used as the first target matrix after sorting and decomposition.

[0120] Wherein, the last column vector of the first target matrix corresponds to the highest level of constellation traversal in the FSD algorithm, 2≤t≤T, where T is the number of targets and t is a positive integer.

[0121] For the principle, implementation, and beneficial effects of the sorting QR decomposition device used in MIMO systems, please refer to the previous text. Figures 1 to 4 The description of the sorting QR decomposition method for MIMO systems shown is not repeated here.

[0122] Reference Figure 7 , Figure 7 This is a schematic diagram of a signal detection device based on FSD according to an embodiment of the present invention. The FSD-based signal detection device includes:

[0123] The signal to be detected determination module 71 is used to determine the product of the target matrix and the transmitted signal vector, and sum the product with the channel noise to obtain the signal to be detected. The target matrix is ​​obtained using the above-mentioned... Figures 1 to 4 Any of the items shown is used to determine the sorting QR decomposition method for MIMO systems;

[0124] Signal detection module 72 is used to perform hard decision on the signal to be detected using the FSD algorithm to determine the detection result;

[0125] In the FSD algorithm, the highest level traverses M constellation points, including hard-determined points, while the remaining levels only retain the constellation point with the closest Euclidean distance as child nodes.

[0126] M≤Q S Q S This represents the total number of all candidate constellation points at the highest level.

[0127] For the principles, implementation details, and beneficial effects of this FSD-based signal detection device, please refer to the preceding text. Figures 1 to 4 The descriptions of the sorting QR decomposition method for MIMO systems and the aforementioned FSD-based signal detection method are presented here and will not be repeated.

[0128] This invention also provides a storage medium, such as a computer-readable storage medium, on which a computer program is stored, the computer program being executed by a processor. Figures 1 to 4The steps of any of the sorting QR decomposition methods shown for MIMO systems, or the steps of the above-described FSD-based signal detection method, may be performed. The computer-readable storage medium may include non-volatile or non-transitory memory, and may also include optical discs, hard disk drives, solid-state drives, etc.

[0129] Specifically, in this embodiment of the invention, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0130] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0131] This invention also provides a terminal, including a memory and a processor. The memory stores a computer program that can run on the processor, and when the processor runs the computer program, it performs the above-described... Figures 1 to 4 The steps of any of the sorting QR decomposition methods shown for MIMO systems, or the steps of the above-described FSD-based signal detection method, may be performed. The terminal may include, but is not limited to, mobile phones, computers, tablets, and other terminal devices, and may also be servers, cloud platforms, etc.

[0132] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article indicates that the preceding and following related objects have an "or" relationship.

[0133] In the embodiments of this application, "multiple" refers to two or more.

[0134] The descriptions of "first," "second," etc., appearing in the embodiments of this application are for illustrative purposes and to distinguish the objects being described. They have no order and do not indicate any special limitation on the number of devices in the embodiments of this application, nor do they constitute any limitation on the embodiments of this application.

[0135] It should be noted that the sequence number of each step in this embodiment does not represent a limitation on the execution order of each step.

[0136] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A sorting QR decomposition method for MIMO systems, characterized in that, include: Determine the initial sorting matrix based on the channel matrix of the MIMO system to be detected; Determine the sequence vector to be sorted in the initial matrix to be sorted, and determine the number of columns of the sequence vector to be sorted, denoted as the target number; If the target number is greater than a preset value, then the first sorting strategy is executed based on the initial unsorted matrix and the unsorted sequence vector; The first sorting strategy includes: In the first iteration, the column order of the sequence vector with the smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector to be sorted, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the first round. In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the first decomposition matrix from the t-th column to the last column of the sequence vector to be sorted is swapped with the sequence vector to be sorted in the t-th column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the t-th iteration. The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the Tth iteration is used as the first target matrix after sorting and decomposition. Wherein, the last column vector of the first target matrix corresponds to the highest level of constellation traversal in the FSD algorithm, 2≤t≤T, where T is the number of targets and t is a positive integer.

2. The method according to claim 1, characterized in that, Determining the sequence vector to be sorted in the initial matrix to be sorted includes: If the transmit antenna array of the MIMO system to be detected contains transmit antennas of interfering users and transmit antennas of the target user, then each column vector corresponding to the transmit antenna of the target user is used as the sequence vector to be sorted. If the transmit antenna array of the MIMO system to be detected does not contain transmit antennas that interfere with users, then all column vectors in the initial unsorted matrix are used as the unsorted sequence vector.

3. The method according to claim 1 or 2, characterized in that, The transmit antenna array of the MIMO system to be detected includes transmit antennas for interfering users and transmit antennas for the target user; After obtaining the first target matrix, the method further includes: Modulation mode detection is performed on the first target matrix to determine the maximum modulation mode of the interference flow; If the maximum modulation scheme of the interference stream is less than or equal to the modulation scheme of the target stream, then based on all column vectors of the first target matrix, a second sorting strategy is executed to obtain the second target matrix after sorting and decomposition.

4. The method according to claim 3, characterized in that, The second sorting strategy includes: In the first iteration, the column vector with the smallest norm in the first target matrix is ​​swapped with the column vector of the first column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first target decomposition matrix of the first round. In the t-th iteration, the column vector with the second smallest norm value among the column vectors from the t-th column to the last column of the first objective decomposition matrix in the (t-1)-th iteration is swapped with the column vector of the t-th column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first objective decomposition matrix in the t-th iteration. Until t=N T Stop iterating when the Nth iteration is complete. T The first objective decomposition matrix obtained in the first round is used as the second objective matrix; Where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the first target matrix, where t is a positive integer.

5. The method according to claim 1 or 2, characterized in that, The transmit antenna array of the MIMO system to be detected includes transmit antennas of interfering users and transmit antennas of the target user, and the sequence vector to be sorted in the initial unsorted matrix is ​​the column vector corresponding to the transmit antenna of the target user; Before executing the first sorting strategy, the method further includes: Perform a preliminary sorting operation on the initial matrix to be sorted; The preliminary sorting operation includes: Place all column vectors corresponding to the transmitting antennas of the interfering user on the first side of the initial unsorted matrix, and place all column vectors corresponding to the transmitting antennas of the target user on the second side of the initial unsorted matrix; The last column vector on the second side of the matrix obtained after performing the preliminary sorting operation is the column vector corresponding to the highest-level constellation traversal point of the FSD algorithm.

6. The method according to claim 1, characterized in that, The method further includes: If the target number is less than or equal to the preset value, then a third sorting strategy is executed based on the initial unsorted matrix and the unsorted sequence vector; The third sorting strategy includes: In the first iteration, the column order of the sequence vector with the second smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the second decomposition matrix of the first round. In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the t-th to last column of the second decomposition matrix of the (t-1)-th iteration is swapped with the t-th column sequence vector, and the matrix after the swap is decomposed into QR decomposition to obtain the second decomposition matrix of the t-th iteration. The iteration stops when the value of t reaches T, and the second decomposition matrix obtained in the Tth iteration is used as the third target matrix after sorting and decomposition.

7. The method according to claim 6, characterized in that, The transmit antenna array of the MIMO system to be detected includes transmit antennas for interfering users and transmit antennas for the target user; After obtaining the third target matrix, the method further includes: Modulation mode detection is performed on the third target matrix to determine the maximum modulation mode of the interference flow; If the maximum modulation scheme of the interference stream is less than or equal to the modulation scheme of the target stream, then based on all column vectors of the third target matrix, the fourth sorting strategy is executed to obtain the fourth target matrix after sorting and decomposition.

8. The method according to claim 7, characterized in that, The fourth sorting strategy includes: In the first iteration, the column vector with the smallest norm in the third objective matrix is ​​swapped with the column vector in the first column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the second objective decomposition matrix of the first round. In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the t-th to last column of the second objective decomposition matrix of the (t-1)-th iteration is swapped with the t-th column sequence vector, and the matrix after the column swap is decomposed by QR to obtain the second objective decomposition matrix of the t-th iteration. Until t=N T Stop the iteration and set N T The second objective decomposition matrix obtained in the round is used as the fourth objective matrix; Where 1≤t≤N T N T Let t be the total number of columns in the column vectors of the third target matrix, where t is a positive integer.

9. The method according to claim 1, characterized in that, Based on the channel matrix of the MIMO system to be detected, determine the initial matrix to be sorted, including: An augmented matrix is ​​constructed based on the channel matrix, total number of transmit antennas, total number of receive antennas, and channel noise of the MIMO system to be detected. This augmented matrix serves as the initial matrix to be sorted.

10. A signal detection method based on FSD, characterized in that, include: The product of the sorted decomposition target matrix and the transmitted signal vector is determined, and the product is summed with the channel noise to obtain the signal to be detected. The sorted decomposition target matrix is ​​determined by the sorted QR decomposition method for MIMO systems as described in any one of claims 1 to 9. The FSD algorithm is used to perform hard decision on the signal to be detected in order to determine the detection result; In the FSD algorithm, the highest level traverses M constellation points, including hard-determined points, while the remaining levels only retain the constellation point with the closest Euclidean distance as child nodes. M≤Q S Q S This represents the total number of all candidate constellation points at the highest level.

11. A sorting QR decomposition device for a MIMO system, characterized in that, include: The initial unsorted matrix determination module is used to determine the initial unsorted matrix based on the channel matrix of the MIMO system to be detected. The module for determining the sequence vector to be sorted is used to determine the sequence vector to be sorted in the initial matrix to be sorted, and to determine the number of columns of the sequence vector to be sorted, denoted as the target number; The first sorting strategy execution module is used to execute a first sorting strategy based on the initial unsorted matrix and the unsorted sequence vector if the target number is greater than a preset value. The first sorting strategy includes: In the first iteration, the column order of the sequence vector with the smallest norm in the initial matrix to be sorted is swapped with the first column sequence vector to be sorted, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the first round. In the t-th iteration, the second smallest norm value of the sequence vector to be sorted in the first decomposition matrix from the t-th column to the last column of the sequence vector to be sorted is swapped with the sequence vector to be sorted in the t-th column, and the matrix after swapping the column order is decomposed into QR decomposition to obtain the first decomposition matrix of the t-th iteration. The iteration stops when the value of t reaches T, and the first decomposition matrix obtained in the Tth iteration is used as the first target matrix after sorting and decomposition. Wherein, the last column vector of the first target matrix corresponds to the highest level of constellation traversal in the FSD algorithm, 2≤t≤T, where T is the number of targets and t is a positive integer.

12. A signal detection device based on FSD, characterized in that, include: The signal to be detected module is used to determine the product of the target matrix and the transmitted signal vector, and sum the product with the channel noise to obtain the signal to be detected, wherein the target matrix is ​​determined by the sorted QR decomposition method for MIMO systems as described in any one of claims 1 to 9; the signal detection module is used to perform hard decision on the signal to be detected using the FSD algorithm to determine the detection result. In the FSD algorithm, the highest level traverses M constellation points, including hard-determined points, while the remaining levels only retain the constellation point with the closest Euclidean distance as child nodes. M≤Q S Q S This represents the total number of all candidate constellation points at the highest level.

13. A storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it executes the steps of the sorting QR decomposition method for MIMO systems as described in any one of claims 1 to 9, or the steps of the signal detection method based on FSD as described in claim 10.

14. A terminal comprising a memory and a processor, wherein the memory stores a computer program capable of running on the processor, characterized in that, When the processor runs the computer program, it performs the steps of the sorting QR decomposition method for MIMO systems as described in any one of claims 1 to 9, or the steps of the signal detection method based on FSD as described in claim 10.

Images