Transmission pattern generation method and system for high sparsity multi-user transmission

Abstract

This invention discloses a method and system for generating transmission patterns for highly sparse multi-user transmission, primarily addressing the problem of low system error rate performance in existing ODMA random access models when the number of users is large. The implementation includes: dividing the binary information bits transmitted by the user into two information bit sequences; converting the first information bit sequence into a decimal number; encoding and modulating the second information bit sequence to obtain a modulation symbol sequence; generating an all-zero matrix based on the known lengths of the first information bit sequence and the modulation symbol sequence, as well as a pre-determined number of transmission time slots; iteratively updating this matrix by continuously finding the update positions of matrix elements through path tree generation; and finally updating the initial all-zero matrix into a binary sparse matrix, which is the multi-user transmission pattern. This invention reduces the system error rate and improves system performance when handling the access demands of more active users, and can be used for highly sparse multi-user transmission in massive machine-type communication.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, specifically relating to a method and system for generating transmission patterns, which can be used for highly sparsity multi-user transmission in massive machine-type communications. Background Technology

[0002] Massive Machine Type Communication (mMTC) is one of the three major application scenarios of 5G, the fifth-generation mobile communication system defined by 3GPP. In mMTC, to reduce transmission resource consumption and latency, a scheme is often adopted where active devices can transmit data directly without prior request or approval—the unauthorized ground truth (GF) access scheme. Because active devices do not send access requests beforehand, the base station cannot obtain specific information about them. Therefore, the traditional GF access scheme has the following two problems: 1) For a large number of devices, each device needs to be allocated a unique pilot sequence, resulting in increased resource consumption. 2) An activity detection scheme is required at the receiving end to identify the active devices, increasing transmission latency.

[0003] To address the problems of traditional access schemes in mMTC scenarios, in 2017, Polyanskiy proposed a new random access scheme, namely unidentified random access, in "Aperspective on massive random-access". In this scenario, the transmitted data packets are small, and the base station only needs to focus on recovering the set of transmitted messages, without needing to identify the identity of active users, thus saving the overhead of user identification.

[0004] Patent application CN202310812717.5 discloses a large-scale multiple access method based on adaptive matching pursuit. It receives uplink signals from multiple users at the base station and models them as block sparse vectors, then uses these vectors to reconstruct new uplink received signal equations. Next, an adaptive matching pursuit algorithm based on the block sparse model is used to perform active user detection and reconstruct user-transmitted data. While this method can achieve unidentified random access for a large number of terminals while reducing user terminal overhead, its performance is poor when dealing with the need for access from a large number of active users because it is only suitable for sporadic bursts of uplink transmission from a large number of terminals.

[0005] In his paper "ODMA Transmission and Joint Pattern and Data Recovery for Unsourced Multiple Access," JianXiang Yan presents a switch-division multiple access (OAMD) scheme for large-scale machine-type communication. In this scheme, each active user's message is divided into two parts: the first part determines the switching pattern, and the second part is encoded and transmitted in a time-hopping manner based on the switching pattern. Leveraging the sparsity of ODMA, the user's switching pattern is blindly detected from the received signal without the aid of pilots. Switch pattern detection and data decoding are iteratively performed on a sparse graph to improve the overall reliability of the system. While this scheme can handle the access demands of a large number of active users, the system's bit error rate increases sharply as the number of active users continues to rise. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the prior art by proposing a method and system for generating transmission patterns for highly sparse multi-user transmission, thereby optimizing the sparsity of ODMA transmission patterns, reducing the system error rate, and improving system performance when dealing with the access needs of more active users.

[0007] To achieve the above objectives, the technical solution of the present invention is as follows:

[0008] Technical Solution 1:

[0009] A method for generating transmission patterns for highly sparse multi-user transmission, characterized in that it includes:

[0010] The binary information bits of length B sent by user K. k Divided into u ks and u kc Two parts, and the first information bit sequence u ks Convert to a decimal number d k For the second information bit sequence u kc Encode and modulate to obtain a length of n c Modulation symbol sequence w k , where u ks The length is B s u kc The length is B c =BB s ;

[0011] Based on the known parameter B s n c And a binary sparse matrix is ​​generated based on the predetermined number of transmission time slots n. As a multi-user transmission pattern, in which, The column length of the matrix is ​​given by , and the row length of the matrix is ​​the same as the number of transmission time slots, n. The row weight of the matrix is ​​the same as the modulation symbol sequence w. k Length n c Consistent;

[0012] User K selects the dth element in the transmission pattern. k Line, used to transmit the modulation symbol sequence w k .

[0013] Furthermore, the second information bit sequence u kc Encoding and modulation include:

[0014] By u kc Perform channel coding to generate codewords Where v k,j ∈{0,1} represents the j-th symbol of the codeword obtained after channel coding of the information bits of the k-th user, where 1≤j≤n. d The code rate of this encoding is B. c / n d ;

[0015] For the code word v k QPSK modulation is performed to obtain the modulation symbol sequence: Where w k,l Let l be the l-th modulation symbol in the modulation symbol sequence after the information bits of the k-th user are encoded and modulated, where 1 ≤ l ≤ n. c ;

[0016] Furthermore, generating a binary sparse matrix A includes: given a target matrix A with row and column lengths n and n respectively. and row weight n c Initially generate a matrix of all zeros for this matrix;

[0017] Update the all-zero matrix row by row until the first row is updated. The row update yields a binary sparse matrix A, where

[0018]

[0019] Technical Solution 2:

[0020] A transmission pattern generation system for highly sparse multi-user transmission includes:

[0021] The encoding and modulation module is used to acquire relevant parameters for generating multi-user transmission patterns;

[0022] The path tree generation module is used to generate a path tree with the current row node as the root node and pass the generated path tree to the column node selection module.

[0023] The column node selection module is used to select appropriate column nodes in the path tree and pass the selected column nodes to the transmission pattern generation module.

[0024] The matrix generation module generates a multi-user transmission pattern based on the results from the column node selection module, and sends the updating transmission pattern back to the path tree generation module for continued updating until the transmission pattern generation is complete.

[0025] Technical Solution 3:

[0026] A computer program product includes a computer program, characterized in that, when the computer program is executed by a processor, it implements the multi-user transmission pattern generation method described in either technical solution 1 or technical solution 2.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] This invention generates a multi-user transmission pattern, which reduces the collision of information between users during multi-user transmission, enables more sparse multi-user transmission, and improves system performance when dealing with the access needs of more active users.

[0029] In the process of generating transmission patterns, this invention utilizes the current matrix Construct a path tree with the i-th row as the root node, and update the all-zero matrix row by row to obtain the update state of the transmission pattern at each step, thus reducing the computational complexity.

[0030] This invention is due to the fact that in the second information bit sequence u kc During the encoding and modulation process, by controlling the codeword v k QPSK modulation produces shorter modulation symbol sequences, resulting in lower line weights in the generated transmission pattern. This improves sparsity in multi-user information transmission and reduces the system bit error rate. Attached Figure Description

[0031] Figure 1 This is a flowchart of an embodiment 1 of the transmission pattern generation method for highly sparse multi-user transmission of the present invention.

[0032] Figure 2 for Figure 1 Flowchart of the coding modulation sub-process;

[0033] Figure 3 This is a schematic diagram of the path tree generation in Embodiment 1 of the present invention.

[0034] Figure 4 This is a block diagram of the transmission pattern generation system for highly sparse multi-user transmission according to the present invention;

[0035] Figure 5 The figures show the simulation results of signal-to-noise ratio and bit error rate in an ODMA system using the method of this invention and the random generation of transmission patterns, respectively. Detailed Implementation

[0036] The embodiments and effects of the present invention will now be described in detail with reference to the accompanying drawings.

[0037] Example 1: A method for generating transmission patterns for highly sparse multi-user transmission.

[0038] This example demonstrates the generation of a transport pattern in an ODMA system. In this system, multiple users need to send binary information bits of the same length. The message is encoded and modulated using a defined scheme to match the corresponding transport pattern. Finally, the transport pattern selects the appropriate time slot for transmission. Generating the transport pattern requires updating the corresponding all-zero matrix multiple times. Each update requires generating the corresponding path tree of the current matrix to determine the position of the updated element.

[0039] Reference Figure 1 The implementation steps of this example include the following:

[0040] Step 1: The user transmitter determines the coding and modulation scheme, determines the transmission pattern parameters, and obtains the modulation symbol sequence.

[0041] like Figure 2 As shown, the implementation of this step includes the following:

[0042] 1.1) User K sends binary information bits u of length B. k Divide into two parts of different lengths u ks and u kc ,in:

[0043] First information bit sequence u ks The length is B s It is used to determine the second information bit sequence u kc Specifically, in which time slot will the transmission take place;

[0044] Second information bit sequence u kc The length is B c =BB s Used to transmit the first information bit sequence u after encoding and modulation. ks Transmission will take place within the determined transmission time slot;

[0045] Since the binary information bits sent by different users may not be the same, the two parts that each user divides may also not be the same.

[0046] 1.2) For the first information bit sequence u ks The length is selected to reduce the probability of user collisions:

[0047] If the first information bit sequence u in the binary information bits of different users ks If the preamble is identical, different users will ultimately choose the same preamble, resulting in them transmitting information in the exact same time slot, causing a complete collision that cannot be decoded. Therefore, it is advisable to appropriately increase the first information bit sequence u in the binary information bits. ks The length is increased to reduce the probability of user collisions;

[0048] However, the first information bit sequence u in the binary information bits is added. ks Length B s This will also increase the cost of random access resources and simulation experiments. Therefore, in the example, it is necessary to select an appropriate first information bit sequence u based on the actual situation. ks Length B s No specific limitations are specified here;

[0049] 1.3) The first information bit sequence u ks Convert to a decimal number d k It can be used to determine the second information bit sequence u by subsequently selecting a row in the transmission pattern. kc The specific transmission time slot;

[0050] 1.4) Use RA code to pair the second information bit sequence u kc Perform channel coding to generate codeword v k :

[0051]

[0052] Where v k,j ∈{0,1} represents the j-th symbol of the codeword obtained after channel coding of the information bits of the k-th user, where 1≤j≤n. d

[0053] 1.5) For codeword v k QPSK modulation is performed to obtain the modulation symbol sequence w. k :

[0054]

[0055] Where w k,a Let a be the a-th modulation symbol in the modulation symbol sequence after the information bits of the k-th user are encoded and modulated, where 1 ≤ a ≤ n. c .

[0056] Step 2: Based on the parameters obtained in Step 1 and an initial fixed value n, generate an all-zero matrix.

[0057] 2.1) Set an initial fixed value n, and use it as the row length of the all-zero matrix, representing that the fixed channel resource is n time slots. 2.2) Based on u obtained in step 1... ks Length B s Let the column length of the all-zero matrix be... Because the binary information sent by the user consists of only 0s and 1s, its length is B. s The binary information bits can represent A decimal number;

[0058] 2.3) Based on the modulation symbol sequence w obtained in step 1 k Length n c Let the row weight of the all-zero matrix be n. c .

[0059] After the all-zero matrix is ​​updated and the transmission pattern is obtained, each row of the transmission pattern will have n. c One position is 1, and the remaining positions are 0; in the transmission pattern, the position with an element of 1 indicates that the time slot can be used to carry symbols in the modulation symbol sequence sent by user K, while the position with an element of 0 indicates that the time slot is closed and cannot carry symbols.

[0060] Step 3: Generate the path tree.

[0061] Common methods for generating path trees include depth-first search and breadth-first search methods.

[0062] Depth-first search-based generation methods typically use a recursive approach, generating nodes along a path until a leaf node is reached. Then, the process returns to the previously generated node and checks if there are any other paths that can be generated. If so, the process continues along the new path until all paths have been generated. If not, the process returns to the previously generated node and checks if there are any other paths that can be generated. This process is repeated until all leaf nodes have been generated.

[0063] The generation method based on breadth-first search usually generates tree nodes layer by layer. After generating all the nodes in the current layer, it moves on to the next layer until all the leaf nodes are generated.

[0064] Given that the depth-first search algorithm may cause stack overflow when the tree has a large number of levels, and it cannot control the number of levels generated, therefore...

[0065] This example uses a breadth-first search-based generation method to reduce the path tree generation time and computational complexity by setting the maximum number of levels to generate the path tree.

[0066] Reference Figure 3 As shown, the implementation of this step includes the following:

[0067] 3.1) Set a row counter i, indicating that the current update is up to row i, where The initial value of i is 1;

[0068] 3.2) Let column counter k represent the number of elements updated in the i-th row, where 0 ≤ k ≤ n. c The initial value of k is 0;

[0069] 3.3) Define the current matrix for updating the all-zero matrix as:

[0070] 3.4) Through the current matrix Construct a matrix about the current matrix bipartite graph

[0071] 3.5) Based on the obtained bipartite graph With row node r i For path tree R i The root node of R i Each path in the array is a sequence of alternating row and column nodes, with all leaf nodes being column nodes, and each node on each path appears only once.

[0072] 3.6) Set the number of levels of the path tree, d, and each level consists of one row node and one column node, and set the initial value of d to 0;

[0073] 3.7) In a bipartite graph Find the distance row node r i The nearest column node, take it as R i Middle root node r i The child nodes complete the construction of the 0th level of the path tree;

[0074] 3.8) Based on the bipartite graph In tree R i Add all connection trees R i The edges from the 0th-level column nodes to the corresponding 1st-level row nodes are then connected to the path tree R. i The path tree R is obtained by taking each row node in level 1 as its nearest neighbor column node. i The first layer;

[0075] 3.9) And so on, according to the bipartite graph In tree R i Add all connection trees R i The edges from the (d-1)th level column nodes to the corresponding row nodes in the dth level are then connected to the path tree R. i The path tree R is obtained by finding the nearest column node from each row node in the d-th level. i For the d-th layer, increment the layer number d by 1.

[0076] 3.10) Repeat step 3.9) until the path tree R is reached. i The path tree R has been completed because it has reached a certain level and can no longer be expanded. i The construction.

[0077] Step 4, based on the path tree R i For the current matrix renew.

[0078] 4.1) For path tree R i Let the distance d(r) i ,c j ) is the root node r i To column node c j The length of the shortest path, selected path tree R i The Lth layer, let Represents the path tree R i In and root node r i Let the set of column nodes whose distance is less than 2L+1 be... express Let the complement of the set of all column nodes be represented as C = {c1, c2, ..., c3}. j ,…,c n},but Where j∈{1,2,......,n}, and n is the total number of column nodes;

[0079] 4.2) Based on the layer number L selected in step 4.1, from the path tree R... i Select the columns that meet the requirements and satisfy one of the following two conditions:

[0080] The first case: The path tree cannot be expanded after reaching the Lth level, and the candidate set... Size It must be less than n; this indicates that in the current matrix bipartite graph In the middle, row node r i Since not all column nodes are reachable, row node r is selected. i Unreachable column nodes can prevent the addition of extra short cycles in the current matrix. In this case, the current matrix can be used as a reference. exist Select the column node c with the minimum column weight. j ;

[0081] The second scenario: This situation indicates that in the current matrix bipartite graph In the middle, row node r iSince all column nodes are reachable, selecting any column node would create additional cycles. To maximize the sparse transmission of the transmission pattern, in the path tree R... i In the (L+1)th layer, select distance r i The furthest column node c j ;

[0082] 4.3) Based on the column node c obtained in step 4.2), j Given the j-th column of the matrix and the i-th row currently being updated, we obtain a specific position in the matrix, i.e., the i-th row and j-th column. We set the element at that position to 1 and increment the column counter k by 1. This completes one element update.

[0083] Step 5: After completing an element update, it is necessary to determine whether the transmission pattern has been updated.

[0084] 5.1) Determine whether the current row's row weight meets the requirement, i.e., whether the target row weight n has been reached. c ,

[0085] If not, it means that the current row has not been updated yet and you need to return to step 3.3);

[0086] If so, it means that the current row has been updated. At this time, the column counter k needs to be set to zero and step 5.2) should be executed.

[0087] 5.2) Determine if the current row is the last row of the matrix:

[0088] If not, it means that although the current row has been updated, it still needs to be updated to complete the generation of the transmission pattern. In this case, the row counter i needs to be incremented by 1 and the process returns to step 3.3.

[0089] If so, it means that the last row of the current behavior matrix has been generated and the transmission pattern has been completed.

[0090] The generated transmission pattern can be used for highly sparse multi-user transmission, that is, user K can use the decimal number d obtained in step 1.3). k Select the dth element in the transmission pattern. k Line transmission modulation symbol sequence w k .

[0091] Example 2: Transmission pattern generation system for highly sparse multi-user transmission.

[0092] Reference Figure 4 This example includes an encoding and modulation module 1, a matrix generation module 2, a path tree generation module 3, and a column node selection module 4, wherein:

[0093] The encoding and modulation module is used to acquire relevant parameters for generating multi-user transmission patterns;

[0094] The matrix generation module is initially used to generate an all-zero matrix. Subsequently, based on the results from the column node selection module, the all-zero matrix is ​​updated to generate a multi-user transmission pattern, and the updated transmission pattern is passed to the path tree generation module.

[0095] The path tree generation module is used to generate a path tree with the current row node as the root node and pass the generated path tree to the column node selection module.

[0096] The column node selection module is used to select appropriate column nodes in the path tree and pass the selected column nodes to the matrix generation module.

[0097] The system works as follows:

[0098] The encoding and modulation module 1 encodes and modulates the bits sent by the user to obtain parameters for the matrix generation module, including the length B of the first information bit sequence. s Modulation symbol sequence length n c The matrix generation module 2 generates the first information bit sequence length B provided by the encoding and modulation module 1. s Modulation symbol sequence length n c First, an initial all-zero matrix is ​​generated, along with an initial fixed parameter n. Then, based on the output of column node selection module 4, the current all-zero matrix is ​​iteratively updated, and the updated matrix is ​​passed to path tree generation module 3 until the current matrix iteration ends, resulting in a transmission pattern. Path tree generation module 3 constructs a path tree based on the current matrix provided by matrix generation module 2, with the currently updated row node in the current matrix as the root node, and passes the path tree to column node selection module 4. Column node selection module 4 selects appropriate column nodes based on the path tree provided by path tree generation module 3 and passes the selected column nodes to matrix generation module 2.

[0099] Example 3, Computer Program Product

[0100] This invention provides a computer program product, including a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the above-described multi-user transmission pattern generation method. The method includes: determining the encoding and modulation scheme of the binary bit sequence sent by the user; generating an all-zero matrix based on the parameters obtained after encoding and modulation and iteratively updating the matrix; generating a path tree for selecting appropriate column nodes; and selecting column nodes based on the path tree for iterative updating of the matrix.

[0101] The effects of this invention can be further illustrated by the following simulation experiments.

[0102] 1. Simulation data parameters

[0103] Assume that the number of users in the ODMA random access model is 140, the coding scheme is RA coding with a code rate of 1 / 5, and the modulation scheme is QPSK modulation.

[0104] 2. Simulation Content and Result Analysis

[0105] Under the above conditions, transport patterns were generated in an ODMA random access model using both the present invention and existing random transport pattern generation methods, and the per-user error probability was compared under different signal-to-noise ratios. The results are as follows: Figure 5 .

[0106] from Figure 5 As can be seen, when the signal-to-noise ratio (SNR) is less than 1.8 dB, its impact on the error probability per user is the largest, and the advantage of this invention cannot be seen at this time. When the SNR is greater than 1.8 dB, the error probability per user is smaller when using the transport pattern generated by this invention for the ODMA random access model than when using a randomly generated transport pattern for the ODMA random access model. For example, when the SNR is about 2.2 dB, the error probability per user using the randomly generated transport pattern method is about 15%, while the error probability per user using this invention is about 6%, indicating that using this invention can bring a significant performance improvement to the ODMA random access scheme.

[0107] The above descriptions are merely a few specific examples of the present invention and do not constitute any limitation on the present invention. Obviously, those skilled in the art, after understanding the content and principles of the present invention, may make various modifications and changes in form and details without departing from the principles and structure of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

[0108] It should be noted that the step numbers in the specification and claims of this invention are only for the purpose of clearly describing the embodiments of this invention and facilitating understanding, and their order is not limited.

Claims

1. A method for generating transmission patterns for highly sparse multi-user transmission, characterized in that, include: users The length sent is binary information bits Divided into and Two parts, and the first information bit sequence Convert to a decimal number For the second information bit sequence Encode and modulate to obtain a length of modulation symbol sequence ,in The length is , ; Based on known parameters , and the number of transmission time slots determined in advance. Generate a binary sparse matrix As a multi-user transmission pattern, in which, The column length of the matrix is ​​given by n, and the row length of the matrix is ​​the same as the number of transmission time slots n. The row weight of the matrix is ​​the same as the modulation symbol sequence. length Consistent; According to known parameters , and the number of transmission time slots determined in advance. Generate a binary sparse matrix Its implementation includes the following: 3a) Given the target matrix The row and column lengths are n and n respectively. and line weight Initially generate a matrix of all zeros for this matrix; 3b) Update the all-zero matrix row by row until the first row is updated. The row update yields a binary sparse matrix A, where The implementation includes the following: 3b1) Set a row counter i, indicating that the current update is up to row i, where The initial value of i is 1; 3b2) Set a column counter k, indicating that k elements have been updated in the i-th row, where The initial value of k is 0; 3b3) Define the current matrix for updating an all-zero matrix as: Using the current matrix Construct a path tree with the i-th row as the root node. Its implementation includes the following: 3b3a) Through the current matrix Construct a matrix about the current matrix bipartite graph ; 3b3b) Based on the obtained bipartite graph row nodes Path tree The root node, Each path in the array is a sequence of alternating row and column nodes, with all leaf nodes being column nodes, and each node on each path appears only once. 3b3c) Set the number of levels of the path tree, d, and each level consists of one row node and one column node, and set the initial value of d to 0; 3b3d) in a bipartite graph Find the distance row node The most recent column node, take it as Middle root node The child nodes complete the construction of the 0th level of the path tree; 3b3e) Based on the bipartite graph In the tree Add all connection trees The edges from the 0th-level column nodes to the corresponding 1st-level row nodes are then connected to the path tree. The path tree is obtained by finding the nearest column node from each row node in level 1. The first layer; 3b3f) and so on, according to the bipartite graph In the tree Add all connection trees The edges from the (d-1)th level column nodes to the corresponding row nodes in the dth level are then connected to the path tree. The path tree is obtained by finding the nearest column node from each row node in the d-th level. For the d-th layer, increment the layer number d by 1. Repeat step 3b3f) until the path tree is complete. Once a certain level is reached and can no longer be expanded, the path tree is complete. 3b4) Each element update requires selecting a column that meets the requirements based on the constructed path tree, updating the element pointed to by the i-th row and the selected column to 1, and incrementing the counter k by 1. This completes one element update. 3b5) Determine if k is equal to : If k is not equal to Then return to step 3b3); If k equals Then proceed to step 3b5); 3b6) Determine if i is equal to : If i is not equal to Then increment i by 1, set k to 0, and return to step 3b3); If i equals If the update ends, the binary sparse matrix A is obtained.

2. The method according to claim 1, characterized in that, The second information bit sequence Encoding and modulation are performed, and their implementation includes the following: 2a) By means of Generate codewords ,in For the first The codeword obtained after channel coding of the information bits of the user is the first... Each code element The code rate of this encoding is ; 2b) For code words QPSK modulation is performed to obtain the modulation symbol sequence. ,in For the first The information bits of the i-th user are encoded and modulated into the modulated symbol sequence of the i-th user. One modulation symbol, .

3. The method according to claim 1, characterized in that, In step 3b4), each time an element is updated, a column that meets the requirements is selected based on the constructed path tree. The implementation includes the following: 3b4a) For path trees Let distance root node To column node The length of the shortest path, selected path tree The Lth layer, let Representing a path tree middle and root node Let the set of column nodes whose distance is less than 2L+1 be... express Let the complement of the set of all column nodes be represented as . ,but ,in n is the total number of column nodes; 3b4b) Based on the level L selected in step 3b4a), from the path tree Select the columns that meet the requirements and satisfy one of the following two conditions: The first case: The path tree cannot be expanded after reaching the Lth level, and the candidate set... Size It must be less than n; at this point, according to the current matrix... ,exist Select the column node with the lowest column weight. ; The second scenario: In , Under the condition of path tree Select distance in layer (L+1) farthest column node .

4. A transmission pattern generation system for implementing the method of claim 1, characterized in that, include: The encoding and modulation module is used to acquire relevant parameters for generating multi-user transmission patterns; The matrix generation module is initially used to generate an all-zero matrix. Subsequently, based on the results from the column node selection module, the all-zero matrix is ​​updated to generate a multi-user transmission pattern, and the updated transmission pattern is passed to the path tree generation module. The path tree generation module is used to generate a path tree with the current row node as the root node and pass the generated path tree to the column node selection module. The column node selection module is used to select appropriate column nodes in the path tree and pass the selected column nodes to the matrix generation module.

5. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the multi-user transmission pattern generation method as described in any one of claims 1 to 2.

Images