A low-complexity receiving method for downlink MIMO-SCMA based on packetization

Abstract

The present invention discloses a downlink MIMO-SCMA low-complexity receiving method based on grouping. The method adopts an information transmission algorithm based on a serial strategy. All variable nodes are grouped before iteration, and the increment of codeword likelihood information is calculated during each iteration, and then the variable node set participating in the next iteration is decided; in order to reduce the amount of information transmission, part of the user information is pre-decoded after a certain number of iterations; then, the remaining variable nodes continue to exchange likelihood information; and finally decode and output. The present invention realizes a downlink MIMO-SCMA low-complexity receiving method based on grouping by grouping all variable nodes and reducing redundant information updates, and the receiving method can obtain approximately optimal bit error rate performance.

Description

Technical Field

[0001] The present invention relates to the technical field of MIMO-SCMA wireless communication systems, and in particular to a packet-based downlink MIMO-SCMA low-complexity receiving method. Background Art

[0002] The network protocols and signal processing schemes used in the current communication architecture cannot well meet the requirements of high-speed, low-latency and ultra-large-scale reliable connections in 5G IoT scenarios. As a new generation of technology that can greatly improve spectrum utilization, non-orthogonal multiple access has attracted widespread attention from academia and industry. Unlike the orthogonal multiple access technology used in current mobile communication networks, non-orthogonal multiple access technology allows multiple users to share the same time-frequency resources, which helps to improve system capacity and meet the needs of more users to access. Non-orthogonal multiple access technology can be mainly divided into two categories: power domain and code domain. In power domain non-orthogonal multiple access, users superimpose signals with different powers and send them on the same time-frequency resources. The receiving end uses an interference elimination scheme to decode the user's signal. The commonly used scheme for code domain non-orthogonal multiple access is SCMA, which was first proposed by Huawei. Its principle is summarized as follows: users select corresponding codewords from their own codebooks according to the transmitted bit data and superimpose and send them. After transmission through the wireless channel, the receiving end uses an information transmission algorithm that approximates the maximum likelihood decoding performance to decode the information of each user.

[0003] In addition, by equipping the transmitting and receiving ends with multiple antennas, MIMO can enable the communication system to obtain greater diversity gain and multiplexing gain, thereby achieving higher spectrum efficiency and throughput, thus meeting people's growing data communication needs. Furthermore, applying the MIMO scheme to the SCMA system will have extremely strong practical significance and value. By integrating the fully connected factor graph of MIMO with the sparse factor graph of SCMA, an information transfer algorithm receiver that approximates maximum likelihood probability decoding can be realized. However, the information transfer algorithm has an exponentially high complexity due to the large amount of information exchange in its variable node updates, which limits its practical use value and makes it difficult to meet the requirements of low latency and high throughput in IoT communications.

[0004] The invention combines the factor graphs of MIMO and SCMA and uses the information transfer algorithm for iterative decoding. At the same time, it fully utilizes the sparsity characteristics of SCMA codewords, reduces the number of variable nodes involved in the iteration, and greatly reduces the complexity of the receiver. Summary of the invention

[0005] The object of the present invention is to provide a packet-based downlink MIMO-SCMA low-complexity receiving method to solve the problem that when a traditional receiver performs a decoding operation, the high complexity limits the application of MIMO-SCMA in an actual communication system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A packet-based downlink MIMO-SCMA low-complexity receiving method, the receiving method comprising the following steps:

[0008] Step S1: for a downlink MIMO-SCMA system with multiple antennas at the transmitting and receiving ends, a factor graph integrating MIMO and SCMA is obtained;

[0009] Step S2: constructing an information transmission algorithm based on a serial update strategy, wherein the serial update strategy is: the information transmission process uses the information updated in the current iteration step;

[0010] Step S3, execute the information transmission algorithm based on the serial update strategy constructed in step S2, and perform iterative decoding, wherein, in each iteration of the algorithm, the likelihood information of each codeword is calculated, and the set of variable nodes to be updated is decided by combining the sum of the codeword likelihood information in the previous iteration; at the same time, the variable nodes are grouped in units of transmitting antennas to ensure that each antenna has a corresponding variable node that can participate in each iteration; after several iterations, the pre-decoded set of variable nodes is decided based on the sum of the codeword likelihood probabilities.

[0011] Furthermore, the step S1 specifically includes:

[0012] Step S101: For a downlink MIMO-SCMA system with J users sharing K resource nodes, the number of antennas at the base station end and the number of antennas at the user end of the system are T and R respectively;

[0013] Step S102: Send the transmitted bit information of user j′ on the tth antenna Mapped to SCMA's K-dimensional complex codeword vector Among them, t = 1, 2 ... T, j ' = 1, 2 ... J, the codeword vector is composed of the transmitted bits from each user's unique codebook Select from

[0014] Step S103: Send the codewords of all J users on antenna t The superposition forms the transmission symbol, and the expression is: Correspondingly, at the receiving end, the received signal on the rth antenna of user j is The expression is:

[0015]

[0016] In the formula, r = 1, 2…R, is the channel gain vector between the tth transmitting antenna at the base station and the rth receiving antenna at user j, is a zero-mean covariance matrix Gaussian white noise vector;

[0017] Step S104: stack the signals on all R receiving antennas, and the received signal of user j is expressed as:

[0018]

[0019] In the formula,

[0020] Step S105: construct a binary indicator matrix F to represent the sparse characteristics of the MIMO-SCMA communication system codebook, the u-th row of the matrix corresponds to the u-th virtual resource, the v-th column of the matrix corresponds to the v-th virtual user, u = 1, 2 ... KR, v = 1, 2 ... JT. The binary indicator matrix F is expressed as:

[0021]

[0022] In the formula, the matrix M represents the MIMO channel configuration, F′ represents the indicator matrix of the single-antenna SCMA system, and the symbol represents the Kronecker product. The column weight of the matrix F is w v and line weight c The column weight of matrix F′ is d v and line weight c R times and T times, the expression is:

[0023] w v =Rd v ,w c =Td c

[0024] Step S106: pass all the signal codewords The possibility of constructing the virtual codebook of the MIMO-SCMA communication system is expressed as:

[0025]

[0026] In the formula, v = J (t-1) + j';

[0027] Step S107: rewrite the expression of the signal received by user j as:

[0028]

[0029] In the formula, v=J(t-1)+j′.

[0030] Furthermore, the step S2 specifically includes:

[0031] Step S201: Definition is the codeword of the u-th resource node The prior probability likelihood information of is defined as and are the information passed from resource node u to variable node v and the information passed from variable node v to resource node u in the lth iteration respectively;

[0032] Step S202: Initialize the information transferred from variable node j to resource node k as follows:

[0033] Step S203: In the lth iteration, for u=K(r-1)+k, k=1,2...K, the information transmitted from the resource node u to the variable node v is:

[0034]

[0035] Among them, the conditional probability It is expressed as:

[0036]

[0037] In the formula, and Respectively represent the received signal and channel state information on the u-th resource node, It's a codeword The u-th element of Indicates that all nodes connected to the u-th resource node except the v-th one w c -A set of 1 variable nodes;

[0038] Step S204: In the first iteration, the information transmitted from the variable node v to the resource node u is:

[0039]

[0040] In the formula, Indicates that the nodes connected to the vth variable except the uth variable w v -A collection of 1 resource nodes.

[0041] Furthermore, the step S3 specifically includes:

[0042] Step S301: During the first iteration, the information transmission algorithm based on the serial update strategy constructed in step S2 is adopted;

[0043] Step S302, when performing the first iteration, specifically includes:

[0044] Step S3021: Calculate the difference between the likelihood probability information and the likelihood probability information after the current and previous iterations of each variable node, expressed as:

[0045]

[0046] Step S3022: It is stipulated that in the l+1th iteration, the variable node corresponding to the minimum value of the difference will not participate in the information transmission update. Therefore, the set of variable nodes participating in the l+1th iteration process and connected to the uth resource node is expressed as:

[0047]

[0048] in,

[0049]

[0050] Step S303, in the Lth 0 -1 and L 0 In the iteration process, the code word likelihood probability corresponding to each variable node is first calculated, and the expression is:

[0051]

[0052] Then, calculate the Lth 0 -1 and L 0 The difference between the sum of the codeword likelihood probability information after each variable node iteration is expressed as:

[0053]

[0054] Finally, decode the smallest J in each group 0 Δ v Codeword information corresponding to the variable node;

[0055] Step S304: In the last LL 0 During the iteration, the variable nodes that have decoded the transmitted codeword will not participate in the information transmission process.

[0056] Further, in step S3, the variable nodes are divided into T groups according to the transmit antenna index number, and in L 0 After the iteration update, J 0 The codeword information of T variable nodes is decoded, thereby reducing the number of codeword combinations in subsequent iterations.

[0057] Furthermore, in step S302, the variable node corresponding to the target user is removed so that it cannot participate in the calculation of the increment of the codeword likelihood information and . Therefore, the expression of the set of variable nodes to be updated connected to the resource node u is:

[0058]

[0059] in,

[0060]

[0061] Further, in step S303, in the L 0 -1 and L 0 During the Lth iteration, the variable node corresponding to the target user is removed so that it cannot calculate the increment of the codeword likelihood probability and the codeword probability, so as to ensure that the variable node corresponding to the target user is always involved in the information transmission process. The corresponding likelihood probability is expressed as:

[0062]

[0063] The codeword corresponding to the maximum likelihood probability is the final decoding output.

[0064] The beneficial effects of the present invention are:

[0065] The present invention proposes a low-complexity receiving method based on grouping MIMO-SCMA, which reduces the complexity of the receiver and ensures the receiving performance of the communication system by reducing redundant iterations in the information transmission process. Simulation results show that the proposed grouping based message passing algorithm (GMPA) receiver greatly reduces the receiving complexity and can achieve a bit error rate performance close to that of the original information passing algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a schematic diagram of the structure of the MIMO-SCMA system provided in Embodiment 1 of the present invention. In which, Figure a represents the transmitting end of MIMO-SCMA, and Figure b represents the receiving end of MIMO-SCMA;

[0067] Figure 2 The relationship curve diagram of the signal to interference noise ratio and the uncoded bit error rate when 2×2 MIMO is configured in Embodiment 1 of the present invention;

[0068] Figure 3 The relationship curve diagram of the signal to interference noise ratio and the uncoded bit error rate when 4×4 MIMO is configured in Embodiment 1 of the present invention;

[0069] Figure 4 This is a diagram showing the comparison results of the number of additions and multiplications of various information transmission algorithms when 2×2 MIMO is configured in Embodiment 1 of the present invention;

[0070] Figure 5 This is a diagram showing the comparison results of the number of additions and multiplications of various information transmission algorithms when 4×4 MIMO is configured in Example 1 of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] Example 1

[0073] See also Figure 1-Figure 5 This embodiment provides a packet-based downlink MIMO-SCMA low-complexity receiving method. Figure 1 In the MIMO-SCMA system, at the transmitting end, multiple users select codewords from their own independent codebooks and send them from multiple antennas. At the receiving end, the MPA receiver can be realized by fusing the fully connected factor graph of MIMO with the sparse factor graph of SCMA. Although MPA can achieve high performance close to the maximum likelihood algorithm, its computational complexity is too high, which limits its practical application.

[0074] This embodiment uses the sparsity of SCMA codewords to propose a group-based downlink MIMO-SCMA low-complexity receiver design. First, the degree of codeword likelihood information update during the iteration process is used to determine the set of variable nodes that participate in information transmission in the next iteration; then, after several iterations, the transmission information of some users is determined; finally, only the translated variable nodes participate in the iteration process until the iteration terminates. The specific steps are as follows:

[0075] Step 1: for a downlink MIMO-SCMA communication system with multiple antennas at the transmitting and receiving ends, construct a MIMO-SCMA expansion factor graph suitable for decoding by an information transmission algorithm;

[0076] Specifically, in this embodiment, step 1 specifically includes:

[0077] For a downlink MIMO-SCMA system in which J users share K resource nodes, the number of antennas at the base station and the user end are T and R respectively.

[0078] First, the transmitted bit information of user j′ on the tth antenna is Mapped to SCMA's K-dimensional complex codeword vector Where t = 1, 2…T, j′ = 1, 2…J. The codeword vector is composed of the transmitted bits from each user’s unique codebook. Select from.

[0079] Then, the codewords of all J users on the transmitting antenna t are The symbols are superimposed to form the transmission symbols, namely:

[0080]

[0081] At the receiving end, the received signal on user j’s rth antenna is It is expressed as:

[0082]

[0083] Where r = 1, 2…R, is the channel gain vector between the tth transmitting antenna at the base station and the rth receiving antenna at user j, is a zero-mean covariance matrix Gaussian white noise vector.

[0084] By stacking the signals from all R receiving antennas, the received signal of user j can be expressed as:

[0085]

[0086] in,

[0087]

[0088] The sparse characteristics of the downlink MIMO-SCMA system codebook can be represented by a KR×JT binary indicator matrix F, where the element of the u-th row (corresponding to the u-th virtual resource) and the v-th column (corresponding to the v-th virtual user) of the matrix is ​​represented as F u,v , where u=1,2…KR, v=1,2…JT.

[0089] In fact, the indicator matrix F of the MIMO-SCMA system can be extended from the indicator matrix F′ of the single-antenna SCMA system. The specific expression is:

[0090]

[0091] The symbol represents the Kronecker product, and the matrix M represents the MIMO channel configuration. It can be seen that the column weight of the matrix F is w v and line weight c The column weight of matrix F′ is d v and line weight c R times and T times, the expression is:

[0092] w v =Rd v ,w c =Td c

[0093] All sent signal codewords The possibility of constitutes the virtual codebook of the MIMO-SCMA system, which is expressed as:

[0094]

[0095] Where v = J(t-1) + j'.

[0096] Therefore, rewriting the expression of the signal received by user j is:

[0097]

[0098] in

[0099] The above process integrates MIMO detection and SCMA decoding into a factor graph so that the MIMO-SCMA system can be demodulated using the information passing algorithm at the receiving end.

[0100] definition is the codeword of the u-th resource node The prior probability likelihood information of . At the same time, define and They are the information passed from resource node u to variable node v and the information passed from variable node v to resource node u in the lth iteration respectively.

[0101] Then, the three update processes of the information transmission algorithm are specifically described as:

[0102] (1) Initialize the information transferred from variable node j to resource node k as follows:

[0103]

[0104] (2) In the lth iteration, for u = K(r-1) + k, k = 1, 2…K, the information transmitted from the resource node u to the variable node v is:

[0105]

[0106] The conditional probability It is expressed as:

[0107]

[0108] In the formula and Respectively represent the received signal and channel state information on the u-th resource node, It's a codeword The u-th element of . Indicates that all nodes connected to the u-th resource node except the v-th one w c -A collection of 1 variable nodes.

[0109] (3) In the lth iteration, the information transferred from the variable node v to the resource node u is:

[0110]

[0111] in, Indicates that the nodes connected to the vth variable except the uth variable w v -A collection of 1 resource nodes.

[0112] The computational complexity of the information exchange process from resource node u to variable node v increases with w. v The growth of increases exponentially, that is, for each resource node, All need to be calculated An accumulated item.

[0113] Step 2: When decoding the received signal, the variable nodes and check nodes involved in the algorithm are iteratively updated simultaneously using a serial update strategy.

[0114] Specifically, in this embodiment, step 2 specifically includes:

[0115] In order to speed up the convergence, the serial update strategy is used in this iteration, that is, the information transmission process uses the information updated in the current iteration step. At this time, in the lth iteration, the information transmitted from the resource node u to the variable node v is rewritten as

[0116]

[0117] Step 3: When the information transmission algorithm updates each variable node in each iteration, if the difference between the codeword likelihood information after the current and previous iterations is small, it will not participate in the next iteration. Therefore, the general information transmission algorithm cannot guarantee that each transmitting antenna participates in each iteration. To this end, the variable nodes are divided into T groups according to the transmitting antenna index number. To further reduce the decoding complexity, L 0 After the iteration update, J 0 The codeword information of T variable nodes is decoded, thereby reducing the number of codeword combinations in the subsequent iteration process. For the lth iteration, the proposed group information transmission algorithm can be summarized into the following four steps:

[0118] (1) The first iteration process adopts an information transmission scheme based on a serial strategy;

[0119] (2) For the lth iteration, calculate the difference between the likelihood probability information and the previous iteration of each variable node, that is,

[0120] Then, in the l+1th iteration, the variable node corresponding to the minimum value of the difference will not participate in the information transmission update. Therefore, the set of variable nodes participating in the l+1th iteration process and connected to the uth resource node can be expressed as

[0121]

[0122] in,

[0123]

[0124] (3) In the L 0 -1 and L 0 In the iteration process, the code word likelihood probability corresponding to each variable node is calculated respectively, which is expressed as:

[0125]

[0126] Based on the above results, calculate the L 0 -1 and L 0 The difference between the sum of the codeword likelihood probability information after each variable node iteration, that is,

[0127]

[0128] Next, decode the smallest J in each group 0 Δ v Codeword information corresponding to the variable node.

[0129] (4) The final LL 0 In the iteration process, the variable nodes corresponding to the decoded transmission codeword will not participate in the information transmission process. The receiver using the group information transmission algorithm can be applied to both uplink and downlink MIMO-SCMA systems.

[0130] More specifically, since the downlink MIMO-SCMA system user only needs to decode its own codeword information, in order to further improve the bit error rate performance of the packet information transmission algorithm in step 3, the following two improvements are provided in this embodiment:

[0131] Improvement 1: Since the variable node corresponding to the target user will always participate in the iteration process, in step (2) of step 3, that is, before l<L 0 In the iteration process, the variable nodes corresponding to the target user do not participate in the calculation of the codeword likelihood information and the increment of . Therefore, the expression of the set of variable nodes to be updated connected to the resource node u is

[0132]

[0133] in

[0134]

[0135] Improvement 2: In step (3) of step 3, in the L 0 -1 and L 0 In the iteration process, the increment of the codeword likelihood probability and the variable node corresponding to the target user will not be calculated to ensure that the variable node corresponding to the target user is always involved in the information transmission process.

[0136] After L iterations, the codeword can be calculated The corresponding likelihood probability is

[0137]

[0138] The codeword corresponding to the maximum likelihood probability is the final decoding output.

[0139] In order to demonstrate the superiority of the method of this embodiment, and because the initialization process of all MPA algorithms is the same, this embodiment only analyzes the number of multiplications and additions of the two algorithms during the iteration process.

[0140] Assume the total number of iterations of the original MPA algorithm is L o , then its multiplication and addition complexities are expressed as:

[0141] multiplication:

[0142] addition:

[0143] The MPA algorithm based on the serial strategy can speed up the convergence of the original algorithm, so its number of multiplications and additions only needs to be reduced by L in the above two equations. o Replaced by L, where L o <L.

[0144] multiplication:

[0145] addition:

[0146] The method of this embodiment has only different user indexes participating in the iteration, so the computational complexity can be considered the same. The number of multiplications and additions can be expressed as

[0147] multiplication:

[0148]

[0149] addition:

[0150]

[0151] It can be seen that the method of this embodiment has a fixed complexity and is beneficial to the design of a practical receiver.

[0152] Simulation Example

[0153] In the embodiment of the present invention, in order to ensure that the method of the present embodiment is fairly compared with the existing serial strategy based MPA receiver and the original MPA receiver, the same system configuration is used for Monte Carlo simulation. The single-antenna SCMA system includes J = 6 users and K = 4 resource blocks. The channel vector between the rth receiving antenna of user j and the tth transmitting antenna of the base station is Each element in is an independent and identically distributed complex Gaussian random variable with mean 0 and variance 1. The indicator matrix of the single-antenna SCMA system used is:

[0154]

[0155] The row weight of this matrix is ​​d c =3, column weight is d v =2, that is, each resource block is shared by 3 users and each user occupies 2 resource blocks, and the overload factor is λ=6 / 4=150%.

[0156] Figure 2 and Figure 3 The corresponding system bit error rate (BER) performance when simulating the original message passing algorithm (MPA), the serial strategy based MPA (SMPA) and the packet-based GMPA receiver proposed in this embodiment under different system signal to noise ratios (SNR) when configuring 2×2 and 4×4 MIMO-SCMA systems is depicted respectively. It can be seen that the GMPA receiver proposed in this embodiment has a BER performance close to that of the original message passing algorithm. Experiments show that the serial strategy based message passing algorithm can achieve a better convergence speed than the original MPA receiver, that is, SMPA and GMPA only need 6 iterations to converge, while MPA needs at least 8 iterations to converge. Figure 4 and Figure 5The decoding complexity of various receivers when configuring 2×2 and 4×4 MIMO-SCMA systems is depicted respectively. It can be seen from the figure that the GMPA receiver proposed in this embodiment can greatly reduce the complexity of the receiver and can obtain a receiving performance comparable to the original information transmission algorithm.

[0157] The parts not described in detail in the present invention are all known technologies to those skilled in the art. The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A packet-based downlink MIMO-SCMA low-complexity receiving method, characterized in that: The receiving method comprises the following steps: Step S1: for a downlink MIMO-SCMA system with multiple antennas at the transmitting and receiving ends, a factor graph integrating MIMO and SCMA is obtained; Step S2: constructing an information transmission algorithm based on a serial update strategy, wherein the serial update strategy is: the information transmission process uses the information updated in the current iteration step; Step S3, executing the information transmission algorithm based on the serial update strategy constructed in step S2, and performing iterative decoding, wherein, in each iteration of the algorithm, the likelihood information of each codeword is calculated, and the set of variable nodes to be updated is determined by combining the sum of the codeword likelihood information in the previous iteration; at the same time, the variable nodes are grouped in units of transmitting antennas to ensure that each antenna has a corresponding variable node that can participate in each iteration; after several iterations, the set of variable nodes to be pre-decoded is determined according to the sum of the codeword likelihood probabilities; The step S1 specifically includes: Step S101: For a downlink MIMO-SCMA system with J users sharing K resource nodes, the number of antennas at the base station end and the number of antennas at the user end of the system are T and R respectively; Step S102: Send the transmitted bit information of user j′ on the tth antenna Mapped to SCMA's K-dimensional complex codeword vector Among them, t = 1, 2 ... T, j ' = 1, 2 ... J, the codeword vector is composed of the transmitted bits from each user's unique codebook Select from Step S103: Send the codewords of all J users on antenna t The superposition forms the transmission symbol, and the expression is: Correspondingly, at the receiving end, the received signal on the rth antenna of user j is The expression is: In the formula, r=1,2…R, is the channel gain vector between the tth transmitting antenna at the base station and the rth receiving antenna at user j, is a zero-mean covariance matrix Gaussian white noise vector; Step S104: stack the signals on all R receiving antennas, and the received signal of user j is expressed as: Officially, Step S105, construct a binary indicator matrix F to represent the sparse characteristics of the MIMO-SCMA communication system codebook, the u-th row of the matrix corresponds to the u-th virtual resource, the v-th column of the matrix corresponds to the v-th virtual user, u = 1, 2 ... KR, v = 1, 2 ... JT, wherein the binary indicator matrix F is expressed as: In the formula, the matrix M represents the MIMO channel configuration, F′ represents the indicator matrix of the single-antenna SCMA system, and the symbol represents the Kronecker product, where the column weight of the matrix F is w v and line weight c The column weight of matrix F′ is d v and line weight c R times and T times, the expression is: w v =Rd v ,w c =Td c Step S106: pass all the signal codewords The possibility of constructing the virtual codebook of the MIMO-SCMA communication system is expressed as: In the formula, v = J (t-1) + j'; Step S107: rewrite the expression of the signal received by user j as: Officially, The step S2 specifically includes: Step S201: Definition is the codeword of the u-th resource node The prior probability likelihood information of is defined as and are the information passed from resource node u to variable node v and the information passed from variable node v to resource node u in the lth iteration respectively; Step S202: Initialize the information transferred from variable node j to resource node k as follows: Step S203: In the lth iteration, for u=K(r-1)+k, k=1,2...K, the information transmitted from the resource node u to the variable node v is: Among them, the conditional probability It is expressed as: In the formula, and Respectively represent the received signal and channel state information on the u-th resource node, It's a codeword The u-th element of V u \v indicates all nodes connected to the u-th resource node except the v-th node w c -A set of 1 variable nodes; Step S204: In the first iteration, the information transmitted from the variable node v to the resource node u is: In the formula, Indicates that the nodes connected to the vth variable except the uth variable w v -A collection of 1 resource nodes; The step S3 specifically includes: Step S301: During the first iteration, the information transmission algorithm based on the serial update strategy constructed in step S2 is adopted; Step S302, when performing the first iteration, specifically includes: Step S3021, calculate the difference between the likelihood probability information and the likelihood probability information after the current and previous iterations of each variable node, expression: Step S3022: It is stipulated that in the l+1th iteration, the variable node corresponding to the minimum value of the difference will not participate in the information transmission update. Therefore, the expression of the set of variable nodes participating in the l+1th iteration process and connected to the uth resource node is: in, Step S303: During the L0th and L0-1th iterations, the codeword likelihood probability corresponding to each variable node is first calculated, and the expression is: Then, the difference between the sum of the codeword likelihood probability information after the L0-1th and L0th iterations of each variable node is calculated, and the expression is: Finally, decode the smallest J0 Δ in each group v Codeword information corresponding to the variable node; Step S304: In the last L-L0 iteration process, the variable node corresponding to the decoded transmission codeword will not participate in the information transmission process; In the step S3, the variable nodes are divided into T groups according to the transmit antenna index number, and the codeword information of J0T variable nodes is first decoded after L0 iteration updates, thereby reducing the number of codeword combinations in the subsequent iteration process; In step S302, the variable node corresponding to the target user is removed so that it cannot participate in the calculation of the increment of the codeword likelihood information and . Therefore, the expression of the set of variable nodes to be updated connected to the resource node u is: in, In step S303, during the L0-1 and L0 iterations, the variable nodes corresponding to the target user are removed so that they cannot calculate the increment of the codeword likelihood probability and, thereby ensuring that the variable nodes corresponding to the target user are always involved in the information transmission process. When the codeword is calculated after L iterations, The corresponding likelihood probability is expressed as: The codeword corresponding to the maximum likelihood probability is the final decoding output.

Images