Decoding method for low-density parity-check code encoding and decoding multiple-input multiple-output system

Abstract

The present invention provides a decoding method for a low-density parity-check (LDPC) coded and decoded multiple-input multiple-output (MIMO) system, including: in an LDPC coded and decoded and multi-carrier quadrature amplitude modulation (MQAM) MIMO system, performing detection based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations according to the sum of the parity-check information output by the decoding module after a first preset number of decoding iterations in the previous outer loop. The decoding module updates and outputs parity-check information after a first preset number of decoding iterations according to the cavity information output by the detection module after a second preset number of detection iterations, and uses it as the input for the first detection iteration of the detection module in the next outer loop. By retaining the parity-check information output after decoding iterations, the present invention ensures the continuity of messages of the entire receiver, and greatly reduces the complexity of the entire system on the premise of improving the system performance to a certain extent.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and particularly to a method for detecting and decoding a low-density parity-check code encoding and decoding multiple-input multiple-output system. Background Art

[0002] Massive Multiple Input Multiple Output (MIMO) can significantly improve the channel capacity and spectral efficiency, but at the cost of computationally intractable complexity in baseband processing. Therefore, it is extremely important to design an MIMO receiver with excellent performance and complexity trade-off.

[0003] The detection and decoding methods of a receiver can be classified into two types: separate detection and decoding, and iterative detection and decoding. Among them, iterative detection and decoding has better ultimate detection performance. The iterative detection and decoding scheme based on the block expectation propagation algorithm fully utilizes the iterative ultimate performance of the detector and the decoder, gives each module sufficient iteration times, and has excellent performance, but the complexity is unbearable and the delay is very high. The iterative detection and decoding scheme based on the double expectation propagation algorithm considers moment matching (MM) based on the posterior distribution before and after the decoder decodes, so as to achieve continuous propagation of expectations from the perspective of message passing, achieving the effect of low complexity and high performance.

[0004] However, in the iterative detection and decoding scheme based on the double expectation propagation algorithm, the decoding module considers sufficient decoding iteration times in each outer loop, resulting in unnecessary complexity redundancy. Summary of the Invention

[0005] In view of the above problems in the related art, an embodiment of the present invention provides a method for detecting and decoding a low-density parity-check code encoding and decoding multiple-input multiple-output system.

[0006] In a first aspect, the present invention provides a method for detecting and decoding a low-density parity-check code encoding and decoding multiple-input multiple-output system, including: in a multiple-input multiple-output system of low-density parity-check code encoding and decoding and multi-carrier quadrature amplitude modulation, performing detection based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after performing a second preset number of detection iterations according to the sum of the parity-check information output after the decoding module performs a first preset number of decoding iterations in the previous outer loop; the decoding module updates and outputs parity-check information after performing the first preset number of decoding iterations according to the cavity information output by the detection module after performing the second preset number of detection iterations, and uses the parity-check information as the input for the first detection iteration of the detection module in the next outer loop.

[0007] In some embodiments, after the first preset number of decoding iterations, updating and outputting check information includes: in the (l + 1)-th decoding iteration, updating and outputting check information according to bit node information and parity check constraints The bit node information is the check information and the input information of the decoding module For the step of updating and outputting check information according to bit node information and parity check constraints Iterate until the value of l + 1 reaches L, and output the check information as the input for the first detection iteration of the detection module in the next outer loop; where the check information is the check information updated in the l-th decoding iteration; the input information of the decoding module In the first decoding iteration, it is the cavity information output by the detection module after the second preset number of detection iterations; in the 2nd to L-th decoding iterations, it is the sum of the cavity information output by the detection module after the second preset number of detection iterations and the check information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; the value of l is an integer from 0 to L - 1, and L is the first preset number; t is the iteration number of the current outer loop.

[0008] In some embodiments, the detection module updates and outputs cavity information after the second preset number of detection iterations according to the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop, including: in the (s + 1)-th detection iteration, obtaining the prior information PrDI [t][s] by moment matching of the extrinsic information EDI and the bit check node information [t][s] ; updating and outputting the cavity information EDI [t][s] by moment matching according to the prior information PrDI [t][s+1] ; for the step of obtaining the prior information PrDI [t][s] by moment matching of the extrinsic information EDI and the bit check node information [t][s] , and, the step of updating and outputting the cavity information EDI [t][s] by moment matching according to the prior information PrDI [t][s+1] Iterate until the value of s + 1 reaches S, and output the cavity information EDI [t][s] , as the input for the first decoding iteration of the decoding module in the current outer loop; where the extrinsic information EDI [t][s] is the cavity information updated in the s-th detection iteration; the bit check node information is the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop; s takes integer values from 0 to S - 1, where S is the second preset number; and t is the iteration number of the current outer loop.

[0009] In some embodiments, the method further includes:

[0010] Initializing the first outer loop; for the current outer loop, while the detection module performs the second preset number of detection iterations, the decoding module performs the first preset number of decoding iterations in parallel; the outer loop other than the first outer loop of the current outer loop; the detection module outputs the cavity information to the decoding module as input, and the decoding module outputs the check information to the detection module as input; when the preset maximum parallel number is reached, based on the bit soft information output by the decoding module, the information bits are determined.

[0011] In a second aspect, an embodiment of the present invention further provides a detection and decoding device for a low - density parity - check code encoding and decoding multiple - input multiple - output system, including:

[0012] A detection module, configured to perform detection based on expectation propagation in a multiple - input multiple - output system of low - density parity - check code encoding and decoding and multi - carrier quadrature amplitude modulation. For the current outer loop, after the second preset number of detection iterations based on the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop, update and output the cavity information;

[0013] A decoding module, configured to update and output the check information after the first preset number of decoding iterations according to the cavity information output by the detection module after the second preset number of detection iterations, as the input for the first detection iteration of the detection module in the next outer loop.

[0014] In a third aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the detection and decoding method of the low - density parity - check code encoding and decoding multiple - input multiple - output system as described in any one of the above.

[0015] In a fourth aspect, the present invention further provides a non - transitory computer - readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the detection and decoding method of the low - density parity - check code encoding and decoding multiple - input multiple - output system as described in any one of the above.

[0016] Fifth aspect, the present invention further provides a computer program product, including a computer program, which when executed by a processor implements the encoding and decoding method of the low-density parity-check code encoding and decoding multi-input multi-output system as described in any one of the above.

[0017] The encoding and decoding method of the low-density parity-check code encoding and decoding multi-input multi-output system provided by the present invention, based on the iterative detection and decoding scheme of the double expected propagation structure, by retaining the parity information output after the decoding module in the current outer loop of the MIMO system with PLDC encoding and decoding after the first preset number of decoding iterations as the input of the detection module in the next outer loop, ensures the continuity of the messages of the entire receiver. On the premise of improving the system performance to a certain extent, the complexity of the entire system is greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0019] Figure 1 It is a flowchart of the encoding and decoding method of the LDPC encoding and decoding MIMO system provided by the embodiment of the present invention;

[0020] Figure 2 It is a general factor graph of the iterative detection and decoding receiver based on the expected propagation detection algorithm provided by the embodiment of the present invention;

[0021] Figure 3 It is a flowchart of the message passing between DEP and PDD-EP provided by the embodiment of the present invention;

[0022] Figure 4 It is a timing analysis diagram in the LDPC encoding MIMO system provided by the embodiment of the present invention;

[0023] Figure 5 It is one of the performance and complexity comparison diagrams of different detection and decoding receivers provided by the embodiment of the present invention;

[0024] Figure 6 It is another performance and complexity comparison diagram of different detection and decoding receivers provided by the embodiment of the present invention;

[0025] Figure 7 It is the third performance and complexity comparison diagram of different detection and decoding receivers provided by the embodiment of the present invention;

[0026] Figure 8It is a schematic diagram for comparing the complexities of different MIMO receivers under a given bit error rate provided by an embodiment of the present invention;

[0027] Figure 9 It is a comparison diagram of delay / complexity of different MIMO receivers under a given bit error rate provided by an embodiment of the present invention;

[0028] Figure 10 It is a schematic structural diagram of an encoding / decoding device for an LDPC-encoded / decoded MIMO system provided by an embodiment of the present invention;

[0029] Figure 11 It is a schematic hardware structure diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0030] In order to better describe the technical solutions in the embodiments of the present invention, the following introduces relevant knowledge.

[0031] (1) MIMO system: The large-scale multiple-input multiple-output technology combined with spatial multiplexing, spatial diversity, and channel encoding / decoding is one of the key technologies for the fifth-generation mobile communication system and future communications. In order to greatly improve the channel capacity, the MIMO system uses multiple antennas at both the transmitter and the receiver, forming multiple channels between the transceiver, which significantly improves the channel capacity and spectral efficiency. However, due to the existence of large-scale configuration and high-order modulation, the improvement of channel capacity and spectral efficiency comes at the cost of unbearable computational complexity. Therefore, it is extremely important to design an MIMO receiver with excellent performance and complexity trade-off.

[0032] The MIMO receiver in the uplink is mainly divided into three modules: channel estimation, signal detection, and decoding. Assuming perfect channel state information, signal detection and decoding have a particularly important impact on the performance of the receiver. The role of MIMO detection is to recover multiple transmitted signals from the received signal and noise, and then recover the bit information according to the established symbol-bit mapping relationship. The bit information is used as the input of the decoder, and the likelihood ratio soft information of the information bits is obtained through the decoding algorithm of the decoder and used to judge each bit.

[0033] (2) Detectors and Detection Algorithms: For detectors, detection algorithms can be mainly classified into linear detection algorithms and Bayesian detection algorithms. In linear algorithms, the mainstream ones are the Zero Forcing (ZF) algorithm and the Minimum Mean Square Error (MMSE) algorithm. The complexity of linear detection algorithms is polynomial, but their performance is not optimal. Bayesian algorithms mainly include the low-complexity Message Passing Algorithm (MPA) and Belief Propagation (BP) algorithm, as well as the high-performance and high-complexity Expectation Propagation (EP) algorithm. Among them, the complexity of the Expectation Propagation algorithm is cubic polynomial. In related technologies, a series of approximate algorithms have been proposed to reduce the complexity by approximating its matrix inversion process. The performance loss associated with this approximation gradually decreases as the antenna ratio load (number of transmit antennas / number of receive antennas) decreases.

[0034] (3) Detection and Decoding of MIMO Receivers: For current MIMO receivers, their detection and decoding methods can be classified into separate detection and decoding and iterative detection and decoding.

[0035] In the separate detection and decoding scheme, the detector reaches its optimal detection performance after a sufficient number of iterations, and converts the soft symbol information updated by it into soft bit information as the input of the decoder. The decoder obtains the soft bit information of the bits through a sufficient number of iterative decoding algorithms for making decisions on information bits. This design has a relatively high latency, and since the performance of the detector is limited by the channel noise and antenna configuration, the decoding performance also has its limitations. Although the detector has reached the optimal performance of a single detection module through sufficient iterations, this optimal performance is the limit of a single detection module. If there is extrinsic information available for the detector to iterate, it should have better ultimate detection performance.

[0036] Based on this, the iterative detection and decoding scheme is proposed. The essence of iterative detection and decoding is to realize the corrective effect of the extrinsic information generated under the respective constraints (Constraints) of detection and decoding on the other module.

[0037] (4) Detection Decoding and Expectation Propagation Algorithm: In traditional receivers based on the Expectation Propagation detection algorithm, after the symbol information updated by the detector is converted into bit information according to a determined symbol-bit mapping relationship and decoded, a bit decision is directly made. This algorithm is called the separate detection and decoding algorithm. For further improving the reliability of air interface transmission and reducing the air interface transmission latency, the traditional separate detection and decoding algorithm can no longer meet the actual requirements.

[0038] Therefore, inspired by the design mechanism of Turbo codes: decoders 1 and 2 obtain corresponding extrinsic information according to their own prior constraints based on the input as the input for the other decoder, and regarding the detector as a decoder, an iterative detection and decoding algorithm is designed, and its performance has been greatly improved compared with the traditional separate detection and decoding algorithm.

[0039] However, for the iterative detection and decoding algorithm based on EP (Expectation Propagation) detection, this design still has high complexity and high latency due to its extremely high number of outer loop iterations and inner loop iterations.

[0040] (5) Iterative detection and decoding and block expectation propagation algorithm: In related technologies, there is already an iterative detection and decoding scheme based on the block expectation propagation (BEP) algorithm. By converting the extrinsic information generated by the parity check constraint of the decoder into symbol prior information, the symbol prior information is approximated as a Gaussian distribution and represented by the mean and variance. The symbol prior information is used as the prior probability in the joint posterior model of the expectation propagation algorithm, enabling the detector to have better initialization and thus better performance. Then, the extrinsic information generated by the detector through the detection algorithm with a sufficient number of iterations (10 times is optimal) under its channel constraint is converted into bit prior information and fed back to the decoder, enabling the decoder to have better performance. This scheme fully utilizes the iterative limit performance of the detector and the iterative limit performance of the decoder, giving each module a sufficient number of iterations and having excellent performance. However, its complexity is unbearable and it has a very high latency.

[0041] (6) Low Density Parity Check Code (LDPC): The binary LDPC code composed of k information bits and n - k parity check bits is a type of (n, k) linear block code, which maps the information sequence to the transmission sequence through a generator matrix G. For the generator matrix, there exists a parity check matrix H that is completely equivalent. All the codeword sequences (information bits) form the null space of the parity check matrix H, and its generator matrix G can be obtained according to the encoding method of general linear block codes. Linear block codes, that is, linear block encoding, include information bits for transmitting information and parity check bits for verification, and the relationship between information code elements and parity check code elements is a linear relationship.

[0042] Low-Density Parity-Check (LDPC) codes are widely adopted in communication systems because of their relatively low decoding complexity (compared with Turbo codes) and their structure being suitable for partial parallel or full parallel decoding, which is conducive to achieving high-throughput decoding. The performance of LDPC decoding mainly depends on its coding construction and decoding algorithm. For coding, there are mainly two types of methods: random construction and algebraic construction. Random construction obtains the parity-check matrix by computer search under certain limited conditions. Algebraic construction mainly designs excellent parity-check matrices by using methods such as combinatorial construction methods and graph theory methods. For decoding, it is divided into the sum-product algorithm and the bit-flipping algorithm. In wireless communication systems, the sum-product algorithm is more widely applied.

[0043] The coding methods of LDPC codes are mainly divided into two categories: one is coding based on the generator matrix G of LDPC codes, and the other is coding based on the parity-check matrix H of LDPC codes. For the iterative detection and decoding receiver, a simple introduction to the first type of coding scheme is as follows:

[0044] For a given (n - k)×n parity-check matrix H, first convert it into an (n - k)×k P matrix and an (n - k)×(n - k) identity matrix I through the Gauss-Jordan elimination method n-k , satisfying the calculation formula: H = [P I n-k .

[0045] Furthermore, according to the relationship between the generator matrix G and the parity-check matrix H in linear block codes, it can be obtained that: G = [I k P T . Among them, P T is the transpose matrix of the P matrix. Then, according to the definition of the generator matrix G, the calculation method of the codeword c can be obtained as: c = uG = [u uP T . Among them, u is the information bit row vector.

[0046] The decoding algorithms of LDPC codes are mainly divided into two categories: one is the sum-product algorithm (also known as soft decision), and the other is the bit-flipping algorithm (also known as hard decision). In wireless communication systems, the soft decision algorithm and its simplified algorithm are mainly adopted.

[0047] The sum-product algorithm is a message-passing algorithm based on factor graphs, which is applicable to all systems described by factor graphs. Its performance mainly depends on the structure of the factor graph of the system. If the factor graph is acyclic, the sum-product algorithm is exact; if the factor graph is cyclic, the sum-product algorithm is approximate, but still has excellent performance. Therefore, the sum-product algorithm and its simplified algorithm are the main decoding algorithms of LDPC. In the present invention, the logarithm domain minimum sum algorithm (Min-Sum Algorithm, MSA) with extremely low complexity is used, and its main process can be described as:

[0048] ① The check node j updates the information of the bit node i ′ to satisfy the following calculation formula:

[0049]

[0050] where, is the bit node adjacent to j other than i′, and α i′j is the information of the bit node i′ about the check node j symbol, and β i′j is the magnitude of the information of the bit node i′ about the check node j The superscript (l) represents the iteration number l, and the superscript (l + 1) represents the iteration number l + 1.

[0051] ② The bit node i updates the information about the check node j′ to satisfy the following calculation formula:

[0052]

[0053] where, is the soft decision information input to the decoder, represents the check nodes adjacent to i other than j′, the superscript (l + 1) represents the iteration number l + 1, and the superscript (l - 1) represents the iteration number l - 1. The soft information for decoder decision is:

[0054]

[0055] (7) Expectation Propagation Algorithm in MIMO Detection: For a flat MIMO channel with N t transmit antennas and N r receive antennas, the MIMO channel model can be expressed as:

[0056]

[0057] where, is the received signal, is the Gaussian channel matrix, is the transmitted signal, is the Gaussian white noise, is the set of complex numbers. obeys a complex Gaussian distribution with a mean of 0 and a variance of . The element in the i-th row and j-th column of the channel matrix represents the complex channel gain from the j-th transmit antenna to the i-th receive antenna.

[0058] Based on Bayesian theory, the joint probability density function of the transmitted signal in this channel model can be expressed as:

[0059]

[0060] where p r (u i ) is the true symbol prior distribution. Based on the expectation propagation approximation algorithm, this probability model can be expressed as:

[0061]

[0062] where the variance and mean of the approximated joint Gaussian posterior distribution can be expressed as:

[0063]

[0064]

[0065] where Λ and γ are respectively the reciprocal vector and mean vector of the variance of the fitted marginal prior distribution of the symbols. Since the approximated joint posterior probability density follows a Gaussian distribution, the mean μ i and variance of the marginal posterior distribution of each symbol can be directly obtained. According to the principle of message passing, removing the prior distribution of the current symbol from the marginal posterior distribution can obtain the variance and mean of the cavity distribution for the next round of moment matching iteration, as follows:

[0066]

[0067]

[0068] Based on this cavity distribution, it can be assumed that the prior distribution is a uniform distribution, and then the mean and variance of the updated marginal prior probability can be obtained through moment matching for the next round of obtaining the joint posterior distribution, as follows:

[0069]

[0070]

[0071] And due to the need for data fitting stability, a damping factor is usually used to optimize the iterated data, as follows:

[0072]

[0073]

[0074] (8) LDCP and Expectation Propagation Algorithm: As a highly robust soft information iterative detection algorithm, the output information of the Expectation Propagation detection algorithm is not only superior among all non-linear detection algorithms from the perspective of hard decision, but also superior among all detection algorithms from the perspective of soft decision because the approximate information itself is a probability domain model with clear physical meaning, which makes it possible to combine the Expectation Propagation detection algorithm with LDPC coding and decoding based on soft input and soft output.

[0075] BEP regards the detector and the decoder as two independent modules. After each module has undergone sufficient iterations, it feeds back the optimal soft information to the other module. It is worth mentioning that BEP feeds back information from the perspective of the Turbo principle. The essence of this principle is that the information received by a node from another node should not include itself, so as to maximize the role of the designed prior constraint.

[0076] BEP confirms that the extrinsic information of the detector relative to the decoder is the cavity distribution updated by the detector itself, and the extrinsic information of the decoder relative to the detector is the sum of the information of the check nodes of the LDPC decoding algorithm itself.

[0077] Regarding the use of extrinsic information, the extrinsic information of the detector is directly fed back to the decoder as part of the channel input, while the extrinsic information of the decoder acts on the detector in two cases: First, for multiple iterations of the detector itself (excluding the first detection in each outer loop), BEP combines the information of the decoder with the cavity distribution of the detector itself as prior information, and then fits the prior information with a Gaussian distribution through moment matching. Second, for the first detection in each outer loop, BEP directly approximates the information of the decoder in the form of mean and variance into a Gaussian probability model as the input for the first iteration of the detector. This work has excellent performance, but in terms of both the detector and the decoder, its complexity and latency are unbearable.

[0078] The Double Expectation Propagation (DEP) algorithm can achieve the detection performance of BEP with extremely low detection complexity. Its improvement lies in optimizing the use of the extrinsic information of the decoder. Whether it is the first detection or subsequent iterative processes, DEP takes the extrinsic information of the decoder as part of the moment matching to participate in the fitting of the posterior information of the detector itself, achieving the performance of BEP with extremely low complexity, and at the same time achieving better performance than BEP under any antenna ratio and modulation order.

[0079] In DEP, moment matching based on the posterior distribution is considered both before and after the decoder decodes, thus achieving the desired continuous propagation from the perspective of message passing, and realizing the low complexity and high performance of iterative detection decoding. However, there is still room for optimization in its decoding module and the overall architecture.

[0080] Its detection module has achieved low-complexity design. However, due to considering a sufficient number of decoding iterations in each outer loop, the complexity and latency of the decoding module are still very high. Moreover, in related technologies, the iterative decoding detection scheme based on DEP usually considers the decoder as an independent module. Although a sufficient number of iterations guarantees the optimal performance of the decoder, it ignores the continuity of message passing of the decoder itself due to the belief propagation algorithm, resulting in unnecessary complexity redundancy.

[0081] In addition, the operations of the detector and the decoder are not synchronized. When the detector is working, the decoder is in a stopped state, and when the decoder is working, the detector is in a stopped state. The module idle time is relatively long, causing waste of hardware resources. Therefore, it is of great significance to propose a low-latency receiver for MIMO systems applicable to LDPC encoding and decoding and expectation propagation detection algorithms.

[0082] In view of the above problems existing in related technologies, an embodiment of the present invention provides a method for detecting and decoding an LDPC-encoded MIMO system. Based on the iterative detection and decoding scheme of the double-expectation propagation structure, by retaining the parity-check information output by the decoding module in the current outer loop after the first preset number of decoding iterations in the MIMO system with PLDC encoding and decoding as the input of the detection module in the next outer loop, the continuity of messages of the entire receiver is ensured, and the complexity of the entire system is greatly reduced on the premise of improving the system performance to a certain extent.

[0083] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0084] Figure 1 is a schematic flowchart of the method for detecting and decoding an LDPC-encoded MIMO system provided by an embodiment of the present invention. As Figure 1 shown, the execution subject of this method can be an MIMO receiver encoded by LDPC, and this method at least includes the following steps:

[0085] Step 101: In a multiple-input multiple-output (MIMO) system with low-density parity-check (LDPC) coding and decoding and multiple quadrature amplitude modulation (MQAM), detection is performed based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity-check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

[0086] Step 102: The decoding module updates and outputs parity-check information after the first preset number of decoding iterations based on the cavity information output by the detection module after the second preset number of detection iterations, which is used as the input for the first detection iteration of the detection module in the next outer loop.

[0087] Specifically, in an MIMO system with LDPC coding and decoding and multiple quadrature amplitude modulation (MQAM), detection is performed based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity-check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

[0088] Furthermore, the decoding module updates and outputs parity-check information after the first preset number of decoding iterations based on the cavity information output by the detection module after the second preset number of detection iterations, which is used as the input for the first detection iteration of the detection module in the next outer loop. Here, the detection module can be a detector, and the decoding module can be a decoder. The first preset number is the maximum number of decoding iterations, and the second preset number is the maximum number of detection iterations. Both the detection module and the decoding module achieve better performance through sufficient numbers of iterations, while ensuring the continuity of message passing between the detector and the decoder, achieving the effect of reducing complexity. It should be noted that Step 101 and Step 102 can be executed simultaneously to reduce system latency.

[0089] The method for detecting and decoding an LDPC-coded MIMO system provided by the embodiments of the present invention, based on the iterative detection and decoding scheme of a double-expectation propagation structure, by retaining the parity-check information output by the decoding module after the first preset number of decoding iterations in the current outer loop of the MIMO system with PLDC coding and decoding as the input for the detection module in the next outer loop, ensures the continuity of messages throughout the receiver. On the premise of improving system performance to a certain extent, it greatly reduces the complexity of the entire system.

[0090] In some embodiments, in step 102, the detection module updates and outputs cavity information after the second preset number of detection iterations based on the sum of the parity information output by the decoding module after the first preset number of decoding iterations in the previous outer loop, including: in the (s + 1)-th detection iteration, according to the extrinsic information EDI [t][s] and the bit-check node information to obtain the prior information PrDI through moment matching [t][s] ; update and output the cavity information EDI [t][s] through moment matching based on the prior information PrDI [t][s+1] ; iterate the steps of obtaining the prior information PrDI [t][s] through moment matching according to the extrinsic information EDI and the bit-check node information, and updating and outputting the cavity information EDI [t][s] through moment matching based on the prior information PrDI [t][s] until the value of s + 1 reaches S, and output the cavity information EDI [t][s+1] as the input for the first decoding iteration of the decoding module in the current outer loop. Among them, the extrinsic information EDI [t][S] is the cavity information updated in the s-th detection iteration; the bit-check node information [t][s] is the sum of the parity information output by the decoding module after the first preset number of decoding iterations in the previous outer loop; the value of s is an integer from 0 to S - 1, and S is the second preset number; t is the iteration number of the current outer loop.

[0091] Specifically, for the detector of the LDPC-coded MIMO receiver, the entire DEP process can be represented by the t-th outer loop, as follows:

[0092] ① The symbol node u i updates the prior information PrDI [t][s] through moment matching according to the extrinsic information EDI from adjacent received nodes and the bit-check node information [t][s] .

[0093] ② The received node y updates the cavity information EDI [t][s+1] through moment matching based on the prior information of the bit nodes.

[0094] ③ If s + 1 is equal to S (S is the preset maximum number of inner loop iterations, i.e., the second preset number), that is, the inner loop iteration reaches S times, then use the cavity information updated in ② as the input for the decoder for a fixed number of decodings to obtain the bit-check node information Let t = t + 1, s = 0, EDI [t][0] = EDI [t-1][S]Jump to ①; otherwise, let s = s + 1 and jump to ①.

[0095] For ease of explanation, superscripts t and s are used to distinguish the transmitted messages, where t represents the (t + 1)-th Turbo iteration (i.e., the outer loop iteration), and s represents the (s + 1)-th detection iteration (i.e., the inner loop iteration). Among them, EDI is the extrinsic distribution information, which is also the cavity information; PrDI is the prior distribution information.

[0096] Figure 2 is the general factor graph of the iterative detection and decoding receiver based on the expectation propagation detection algorithm provided by the embodiments of the present invention, as Figure 2 shown, where the Polar Decoder is the polar code decoder. For the detector, the generation of the cavity information can be regarded as the receiving node y obtaining the extrinsic information (EDI) of each symbol variable node in the form of the sum-product algorithm and then transmitting it to each symbol node connected to it

[0097] For each symbol node u i , the information processing process is to multiply the extrinsic information EDI transmitted by the connected receiving nodes and the feedback prior information (Feedback Prior Distribution Information, FPrDI) (corresponding to the bit check node information) transmitted by the decoder, and obtain the prior information PrDI through moment matching and transmit it to the corresponding receiving node.

[0098] Repeating this process can optimize the extrinsic information in the way of minimizing the KL (Kullback-Leibler) divergence through the expectation propagation algorithm, and then transmit the optimal extrinsic information to the decoder; furthermore, through the sum-product algorithm of the decoder, each bit node c is obtained under the operation of the i,q (i ∈ 1,…, N t ; q ∈ 1,…, Q) (where Q is the capacity of the modulation constellation), and the extrinsic information of each bit node is transmitted to the mapping node M i (i ∈ 1,…, N t ). The mapping node M i integrates the bit extrinsic information into the feedback prior information through a deterministic modulation relationship and transmits it to the symbol variable node. The symbol variable node updates the prior information of each symbol node through moment matching and then transmits it to the receiving node of the detector for the detection inner loop in the next outer loop.

[0099] For DEP, the main innovation lies in "EDI" in ③ [t][0] = EDI [t-1][S] ", and for the sake of facilitating the explanation from the perspective of the factor graph, it is called the non-restart mechanism. For BEP, EDI [t][0] will be reset to 0 again. That is, in the process of each outer loop iteration, the cavity information of the last round of detection iteration in the inner loop of the detector in the previous outer loop is removed.

[0100] The iterative receiver based on message passing iterates based on soft information. Therefore, the factor graph has a universal message passing mode, and from this, it can be analyzed that the message passing on the decoding side can also achieve complexity reduction and retention of effective information through the non-restart mechanism.

[0101] In some embodiments, after the first preset number of decoding iterations in step 102, the parity information is updated and output, including: in the (l + 1)-th decoding iteration, according to the bit node information and the parity constraint, the parity information is updated and output The bit node information is the parity information and the input information of the decoding module For the step of updating and outputting the parity information according to the bit node information and the parity constraint is iterated until the value of l + 1 reaches L, and the parity information is output as the input for the first detection iteration of the detection module in the next outer loop. Among them, the parity information is the parity information updated in the l-th decoding iteration; the input information of the decoding module is the cavity information output after the detection module has undergone the second preset number of detection iterations in the first decoding iteration; in the 2nd to L-th decoding iterations, it is the sum of the cavity information output after the detection module has undergone the second preset number of detection iterations and the parity information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; the value of l is an integer from 0 to L - 1, L is the first preset number; t is the iteration number of the current outer loop.

[0102] Specifically, the non-restart mechanism of the decoder is as follows: for the sake of facilitating the explanation, the superscript t represents the (t + 1)-th outer loop, and l represents the (l + 1)-th decoding iteration (assuming the maximum number of decoding iterations is L, that is, the first preset number is L), and the entire DEP process is as follows:

[0103] ① The bit node c uses the parity information of adjacent check nodes and the decoder input information as the bit node information:

[0104] ②The check node S uses the bit node information as input, and updates the check information according to its own constraint relationship:

[0105] ③If l + 1 is equal to L, use the check information updated in ② as the input of the detector to perform the detection iteration for the second preset number of times, and obtain the cavity information as the input information of the decoder Let t = t + 1, l = 0, and jump to ①; otherwise, let l = l + 1 and jump to ①.

[0106] It should be noted that the non - restart mechanism of the decoder can be used alone or in combination with the non - restart mechanism of the detector (referred to as the bilateral non - restart mechanism). Based on the bilateral non - restart mechanism, step 101 and step 102 can be executed simultaneously, which can reduce the system delay while improving the conversion efficiency of information between modules, thereby improving the bit error rate performance.

[0107] In the following text, the receiver that combines the bilateral non - restart mechanism of the detector and the decoder with the expectation - propagation detection algorithm is called DEP - dNRe (iterative receiver based on the bilateral non - restart mechanism of EP detection).

[0108] The detection and decoding method of the LDPC encoding and decoding MIMO system provided by the embodiments of the present invention first gives the factor graph model of double expectation - propagation and the factor graph model of the entire receiver, derives the reasons for the low complexity and high performance of the double expectation - propagation structure from the perspective of the factor graph model, and proposes from the perspective of the factor graph that all message - passing - based decoding algorithms should retain the updated information of their check nodes after a certain number of iterations, which is called the non - restart mechanism. Based on this restart mechanism, the message continuity of the entire receiver is guaranteed, so that the detector can consider the information processed by the decoder while processing information, making the detector and the decoder both in the working state. This design fully mobilizes the mutual utilization degree of information between the detector and the decoder, and can greatly reduce the delay of the entire system on the premise of improving the system performance to a certain extent. Compared with the optimal receiver based on double expectation - propagation detection in the related art, the present invention significantly reduces the decoding complexity and the total number of iterations, and reduces the system delay while partially improving the system bit error rate performance.

[0109] In some embodiments, the error detection and decoding method of the low-density parity-check code encoding and decoding multiple-input multiple-output system further includes: initializing the first outer loop; for the current outer loop, while the detection module performs the second preset number of detection iterations, the decoding module performs the first preset number of decoding iterations in parallel, where the current outer loop is an outer loop other than the first outer loop; the detection module outputs cavity information to the decoding module as input, and the decoding module outputs check information to the detection module as input; when the preset maximum parallel number is reached, the information bits are determined according to the bit soft information output by the decoding module.

[0110] Specifically, without sacrificing performance, although the non-restart mechanism of the decoder mentioned in the foregoing embodiments has greatly reduced the complexity of the decoder, or the bilateral non-restart mechanism has greatly reduced the complexity of the detector and the decoder, its delay is still long. There is still room for further performance improvement in the traditional iterative detection and decoding (IDD) design method, and due to the input-output dependence relationship between the two modules, the two modules of the entire system have a relatively long latency.

[0111] Therefore, based on the bilateral non-restart mechanism mentioned in the foregoing embodiments, an efficient implementation scheme is proposed that reduces the module latency while increasing the information update frequency, thereby improving the bit error rate performance. The specific algorithm process is as follows:

[0112] ① Initialization: equivalent to the first outer iteration of DEP-dNRe.

[0113] ② Parallel loop iteration: Based on the bilateral non-restart mechanism, while the detector performs S iterations, the decoder performs L iterations.

[0114] ③ Information exchange: The detector outputs its extrinsic information to the decoder for use in updating the information of the next round of the decoder, and the decoder outputs its extrinsic information to the detector for use in updating the detector of the next round. Specifically, the detector outputs the cavity information determined by the receiving nodes to the decoder as input, and the decoder outputs the check information determined by the check nodes to the detector as input.

[0115] ④ If the set maximum parallel number is reached, a hard decision output is made according to the bit soft information output by the decoder; otherwise, jump to ②.

[0116] In this embodiment, the iterative process of the scheme is divided into the first round of outer iteration and parallel loop iteration. In the first round of outer iteration, since there is no information to be retained at the receiving nodes of the detector and the check nodes of the decoder at this time, the information processing processes of the detector and the decoder are equivalent to the first round of outer iteration process of DEP at this time, and it is called the initialization process. To distinguish DEP and parallel loop iteration, Figure 3 is a schematic diagram of the message passing between DEP and PDD-EP provided by the embodiment of the present invention, Figure 3 where (a) and (b) in it represent the information processing flow graphs in the two algorithms.

[0117] Due to the existence of the bilateral non-restart mechanism, there is extrinsic information left in the previous round of iteration available for use in both the detector and the decoder. Therefore, the detector / decoder can use this information as input to simultaneously update the prior information / bit node information in the next round. After the two modules have updated their information respectively, the detector / decoder respectively uses the extrinsic information of the receiving node / check node as an exchange as the input of the other module, and then performs the next round of parallel iteration until the preset maximum number of parallel iteration times is reached.

[0118] For the sake of convenience of explanation, this iterative process is called the parallel loop iteration process, and the optimized iterative detection and decoding receiver is called the parallel iterative detection and decoding algorithm based on the expectation propagation detection algorithm (EP-based Parallel Iterative Detection and Decoding, PDD-EP).

[0119] The detection and decoding method of the low-density parity-check code encoding and decoding multi-input multi-output system provided by the embodiment of the present invention, based on the bilateral non-restart mechanism, through the parallel implementation of the iterative detection and decoding algorithm, reduces the latency time caused by the dependency relationship between the input and output of the two modules, and at the same time improves the information update frequency, thereby improving the bit error rate performance.

[0120] The following further illustrates the technical solution provided by the present invention with a specific embodiment.

[0121] Figure 4 is a timing analysis diagram in the LDPC-coded MIMO system provided by the embodiment of the present invention. As Figure 4 shown, in an LDPC-coded, 256-QAM, 8×16 MIMO system with a code length of 1280 and a code rate of 0.5625, it is assumed that the channel state can be perfectly obtained. For the sake of convenience of explanation, the number of clock cycles is used as an index to measure the latency of the algorithm.

[0122] Based on the basic hardware implementation principle, except that division takes 5 clock cycles, other arithmetic modules only take 1 clock cycle. For each parallel loop process, it includes one complete expected propagation detection and one complete decoding, and is composed of Figure 5 it can be seen that the detector takes 177 clock cycles and the decoder takes 182 clock cycles.

[0123] After parallel optimization, two parallel loops altogether take 461 clock cycles. For the same number of processing times, DEP needs to take 693 clock cycles, including 395 clock cycles for two detection iterations and one conversion of extrinsic information, and 298 clock cycles for two decoders.

[0124] In addition, since in the DEP algorithm, the message passing of the decoder follows the restart principle, its performance will be greatly reduced. To ensure that the DEP scheme has good performance, the maximum number of decoding iterations is set to 20. In this way, DEP needs to take 2781 clock cycles to achieve the performance of DEP-dNRe. However, such a delay is unacceptable. For DEP-dNRe, although its performance has been improved, its delay and complexity still do not achieve a good compromise. Only after the use of parallel optimization, the entire receiver can achieve better bit error rate performance with lower delay.

[0125] Furthermore, for the PDD-EP scheme provided in the embodiments of the present invention, as well as the existing separated detection and decoding receiver based on the double expected propagation algorithm (DEP), the iterative detection and decoding receiver based on MMSE detection (MMSE-IDD), and the iterative detection and decoding receiver based on the bilateral non-restart mechanism of the double expected propagation structure (DEP-dNRe), simulations of performance and computational complexity are carried out.

[0126] Figure 5 is one of the performance and complexity comparison diagrams of different detection and decoding receivers provided in the embodiments of the present invention. As Figure 5 shown, it gives the relationship between the bit error rate (BER) performance and the number of inner loop decoding iterations with the change of the number of iterations in the LDPC coding with code length 1280, code rate 0.5625, 256-order quadrature amplitude modulation, and 32×64 MIMO system for DEP, DEP-dNRe, and PDD-EP. Among them, the abscissa T is time, the ordinate is BER, and D is the number of decoding iterations.

[0127] From Figure 5It can be seen that DEP has good performance when the number of decoding iterations is set to 20 times, but its delay is unacceptable. For DEP-dNRe, setting the number of decoding iterations to 4 times can exceed the performance of DEP itself, fully demonstrating the excellent performance of the bilateral non-restart mechanism.

[0128] In addition, PDD-EP can obtain better performance at the cost of the same number of iterations, further illustrating the performance advantage of PDD-EP. Specifically, the number of decoding iterations spent by the three are 33, 32, and 160 respectively when the number of outer iterations is 8, and 50, 48, and 240 respectively when the number of outer iterations reaches 12. To sum up, PDD-EP is superior to DEP both in terms of delay and performance.

[0129] Figure 6 It is the second figure of the performance and complexity comparison of different detection and decoding receivers provided by the embodiments of the present invention. Figure 6 It gives the performance comparison of PDD-EP, MMSE-IDD, SDD-EP, and DEP-dNRe in a MIMO system with various antenna ratios (transmit antennas / receive antennas) under LDPC coding with a code length of 4096 and a code rate of 0.5, and 16-QAM and 256-QAM. Among them, the abscissa Average Received SNR is the average received signal-to-noise ratio (Signal Noise Ratio), and the ordinate BER is the bit error rate; the antenna ratios are 16×32 (corresponding to Figure 6 (a) in Figure 6 (b) in Figure 6 (c) in respectively. For the sake of comparison, the theoretical performance limit of the iterative receiver in the case of single-input single-output (SISO) antenna ratio in an additive white Gaussian noise (AWGN) channel is also given.

[0130] Figure 7 It is the third figure of the performance and complexity comparison of different detection and decoding receivers provided by the embodiments of the present invention. Figure 7 It gives the performance comparison of PDD-EP, MMSE-IDD / SDD-EP, and DEP-dNRe in a MIMO system with various antenna ratios under LDPC coding with a code length of 4096 and a code rate of 0.5, and 256-QAM. Among them, the abscissa is AverageReceived SNR which is the average received signal-to-noise ratio, and the ordinate BER is the bit error rate; the antenna ratios are 64×128 (corresponding to Figure 7 (a) in Figure 7in (b)) and 64×192 (corresponding to Figure 7 in (c)). For the sake of easy comparison, the theoretical performance limit of the iterative receiver in the case of SISO antenna ratio under the AWGN channel is also given.

[0131] According to Figure 6 and Figure 7 It can be seen that there is a gap of about 2 dB between all the algorithms and the theoretical performance limit of the iterative receiver in the case of SISO antenna ratio under the AWGN channel, and this gap varies with different iterative detection algorithms.

[0132] Both DEP-dNRe and PDD-EP have a SNR lead of 0.2 to 0.5 dB over SDD-EP in terms of performance, and are more than 1 dB better than MMSE IDD. For DEP-dNRe, its performance converges after 12 outer loops.

[0133] When the number of outer loops is 7 or 8, PDD-EP has already reached the limit performance of DEP-dNRe. It can be seen that PDD-EP has excellent convergence performance. If considering the same complexity situation, PDD-EP has a SNR lead of 0.2 to 0.25 dB. When the number of outer loops reaches 15, PDD-EP gradually reaches the convergence performance.

[0134] In summary, the above simulations show that PDD-EP has better performance advantages, and this advantage becomes gradually obvious with the increase of the antenna ratio. This performance advantage mainly comes from the fact that PDD-EP has a faster information exchange rate, enabling the updated information to be used by another module in a timely manner.

[0135] Furthermore, since the present invention mainly adopts the min-sum decoding algorithm, whose complexity can be ignored compared with the detector, the differences among several algorithms mainly come from the different numbers of detection iterations. Therefore, in terms of complexity detection, the complexity of the detection module and the information conversion module (the mutual conversion process between bit-level information and symbol-level information) is mainly considered. In addition, the sum of the complexities of real-number multiplication, division, and exponential operations is also considered.

[0136] Figure 8 is a schematic diagram of the complexity comparison of different MIMO receivers under a given bit error rate provided by an embodiment of the present invention. Figure 8 shows the relationship between the complexity (Complexity) and the signal-to-noise ratio for different antenna ratios for a given BER performance in an MIMO system with LDPC coding of code length 4096 and code rate 0.5, and 16-QAM and 256-QAM, where the antenna ratios are 16×32 (corresponding to Figure 8in (a)) and 64×92 (corresponding to Figure 8 in (b)).

[0137] PDD-EP can exceed the convergence performance of DEP-dNRe with complexities of 67% and 66.7%. When PDD-EP has 12 outer iteration times, its performance has a gain of nearly 0.2 dB. The slightly higher performance of PDD-EP (T = 12) than that of DEP-dNRe is due to the additional complexity of information conversion by a factor of two caused by parallel implementation.

[0138] Furthermore, Figure 9 is a comparison diagram of latency / complexity of different MIMO receivers under a given bit error rate provided by an embodiment of the present invention. Figure 9 It shows the relationship between the simulation latency (Latency) and the signal-to-noise ratio, and the relationship between the complexity (Complexity) and the signal-to-noise ratio of PDD-EP and DPE-dNRe in an 8×16 MIMO system with LDPC coding of code length 1280 and code rate 0.5625, 256-order quadrature amplitude modulation, under a given bit error rate (BER = 4×10 -4 ), when the number of iterations increases from 8 to 15 (corresponding to Figure 9 in (a)), and the relationship between the complexity (Complexity) and the signal-to-noise ratio (corresponding to Figure 9 in (b)).

[0139] It can be seen that, on the premise of considering the same performance, the simulation latency of PDD-EP is much lower than that of DPE-dNRe, and its complexity is also greatly saved. When considering the same complexity, the latency cost of PDD-EP is 60.3% of that of DEP-dNRe, and the performance improvement is 0.2 dB.

[0140] Next, the decoding and encoding device of the low-density parity-check code encoding and decoding multi-input multi-output system provided by the present invention will be described. The decoding and encoding device of the low-density parity-check code encoding and decoding multi-input multi-output system described below can be mutually referred to the decoding and encoding method of the low-density parity-check code encoding and decoding multi-input multi-output system described above.

[0141] Figure 10 is a schematic structural diagram of the decoding and encoding device of the LDPC encoding and decoding MIMO system provided by an embodiment of the present invention. As Figure 10 shown, the device at least includes:

[0142] A detection module 1001, configured to perform detection based on expected propagation in a multi-input multi-output system of low-density parity-check code encoding and decoding and multi-carrier quadrature amplitude modulation. For the current outer loop, according to the sum of the parity-check information output after the decoding module performs the first preset number of decoding iterations in the previous outer loop, after the second preset number of detection iterations, update and output cavity information;

[0143] A decoding module 1002, configured to update and output check information after the first preset number of decoding iterations according to the cavity information output by the detection module after the second preset number of detection iterations, and use the updated and output check information as the input for the first detection iteration of the detection module in the next outer loop.

[0144] In some embodiments, the decoding module 1002 is further configured to:

[0145] In the (l + 1)-th decoding iteration, update and output check information according to the bit node information and the check constraint The bit node information is the check information and the input information of the decoding module Iterate the step of updating and outputting check information according to the bit node information and the check constraint until the value of (l + 1) reaches L, and output the check information as the input for the first detection iteration of the detection module in the next outer loop; where the check information is the check information updated in the l-th decoding iteration; the input information of the decoding module is the cavity information output by the detection module after the second preset number of detection iterations in the first decoding iteration; in the 2nd to L-th decoding iterations, it is the sum of the cavity information output by the detection module after the second preset number of detection iterations and the check information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; the value of l is an integer from 0 to L - 1, and L is the first preset number; t is the iteration number of the current outer loop. In the first decoding iteration, it is the cavity information output by the detection module after the second preset number of detection iterations; in the 2nd to L-th decoding iterations, it is the sum of the cavity information output by the detection module after the second preset number of detection iterations and the check information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; the value of l is an integer from 0 to L - 1, and L is the first preset number; t is the iteration number of the current outer loop.

[0146] In some embodiments, the detection module 1001 is further configured to:

[0147] In the (s + 1)-th detection iteration, obtain the prior information PrDI [t][s] through moment matching of the extrinsic information EDI and the bit check node information [t][s] ; update and output the cavity information EDI [t][s] through moment matching of the prior information PrDI [t][s+1] ; for the moment matching of the extrinsic information EDI [t][s] and the bit check node information to obtain the prior information PrDI [t][s] and the moment matching of the prior information PrDI [t][s] to update and output the cavity information EDI [t][s+1]Iterate the steps until the value of s + 1 reaches S, and output the cavity information EDI [t][S] , as the input for the first decoding iteration of the decoding module in the current outer loop; wherein, the extrinsic information EDI [t][s] is the cavity information updated in the s-th detection iteration; the bit check node information is the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop; the value of s is an integer from 0 to S - 1, and S is the second preset number; t is the iteration number of the current outer loop.

[0148] In some embodiments, the device further includes a parallel module for:

[0149] Initialize the first outer loop; for the current outer loop, while the detection module performs the second preset number of detection iterations, the decoding module performs the first preset number of decoding iterations in parallel; the current outer loop is the outer loop other than the first outer loop; the detection module outputs the cavity information to the decoding module as input, and the decoding module outputs the check information to the detection module as input; when the preset maximum parallel number is reached, determine the information bits according to the bit soft information output by the decoding module.

[0150] Figure 11 is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. As Figure 11 shown, the electronic device may include: a processor 1110, a communication interface 1120, a memory 1130, and a communication bus 1140. Among them, the processor 1110, the communication interface 1120, and the memory 1130 complete mutual communication through the communication bus 1140. The processor 1110 can call the logical instructions in the memory 1130 to execute a decoding method for a multi-input multi-output system with low-density parity-check code encoding and decoding, and the method includes:

[0151] In a multi-input multi-output system with low-density parity-check code encoding and decoding and multi-carrier quadrature amplitude modulation, perform detection based on expected propagation. For the current outer loop, the detection module updates and outputs the cavity information after the second preset number of detection iterations according to the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop;

[0152] The decoding module updates and outputs the check information after the first preset number of decoding iterations according to the cavity information output by the detection module after the second preset number of detection iterations, as the input for the first detection iteration of the detection module in the next outer loop.

[0153] In addition, when the logical instructions in the above-mentioned memory 1130 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0154] On the other hand, the present invention also provides a computer program product. The computer program product includes 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 execute the decoding and encoding methods provided by the above-mentioned various methods for a multi-input multi-output system using low-density parity-check codes, and this method includes:

[0155] In a multi-input multi-output system using low-density parity-check code encoding / decoding and multi-level quadrature amplitude modulation, detection is performed based on expected propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations according to the sum of the parity-check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

[0156] The decoding module updates and outputs parity-check information after the first preset number of decoding iterations according to the cavity information output by the detection module after the second preset number of detection iterations, and uses it as the input for the first detection iteration of the detection module in the next outer loop.

[0157] On yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the decoding and encoding methods provided by the above-mentioned various methods for a multi-input multi-output system using low-density parity-check codes, and this method includes:

[0158] In a multiple-input multiple-output system with low-density parity-check code encoding and decoding and multi-carrier quadrature amplitude modulation, detection is performed based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity-check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

[0159] The decoding module updates and outputs parity-check information after the first preset number of decoding iterations based on the cavity information output by the detection module after the second preset number of detection iterations, and uses it as the input for the first detection iteration of the detection module in the next outer loop.

[0160] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0161] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A decoding method for a low - density parity - check code encoding and decoding multiple - input multiple - output system, characterized in that, Including: In a multi - input multi - output system with low - density parity - check code encoding / decoding and multi - carrier quadrature amplitude modulation, detection is performed based on expectation propagation. For the current outer loop, the detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity - check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop; The decoding module updates and outputs parity - check information after the first preset number of decoding iterations based on the cavity information output by the detection module after the second preset number of detection iterations, and uses it as the input for the first detection iteration of the detection module in the next outer loop; The updating and outputting of the parity - check information after the first preset number of decoding iterations includes: In the (l + 1)-th decoding iteration, parity information is updated and output according to bit node information and parity constraints The bit node information is the parity information and the input information of the decoding module The check information is the check information updated in the l-th decoding iteration, and the input information of the decoding module is the cavity information output by the detection module after the second preset number of detection iterations in the first decoding iteration; t is the iteration number of the current outer loop; The detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity - check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop, including: In the (s + 1)-th detection iteration, the prior information PrDI is obtained by moment matching based on the extrinsic information EDI [t][s] and the bit check node information ; [t][s] ; Update and output the cavity information EDI according to the moment matching of the prior information PrDI [t][s] ; [t][s+1] ; Among them, the extrinsic information EDI [t][s] is the cavity information updated in the s-th detection iteration; the bit check node information is the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

2. The decoding method of the low-density parity-check code encoding and decoding multi-input multi-output system according to claim 1, characterized in that The updating and outputting of the parity - check information after the first preset number of decoding iterations includes: Update and output the parity information according to the bit node information and the parity check constraint, and iterate the steps until the value of l+1 reaches L, and output the parity information as the input for the first detection iteration of the detection module in the next outer loop; Among them, the input information of the decoding module In the 2nd to the Lth decoding iteration, it is the sum of the cavity information output by the detection module after the second preset number of detection iterations and the parity information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; the value of l is an integer from 0 to L - 1, and L is the first preset number.

3. The decoding method of the low-density parity-check code encoding and decoding multiple-input multiple-output system according to claim 1 or 2, characterized in that The detection module updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity - check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop, including: For the extrinsic information EDI [t][s] and the bit check node information perform moment matching to obtain the prior information PrDI [t][s] , and, for the prior information PrDI [t][s] perform moment matching update and output the cavity information EDI [t][s+1] Iterate the steps until the value of s + 1 reaches S, and output the cavity information EDI [t][S] , which is used as the input for the first decoding iteration of the decoding module in the current outer loop; where s takes integer values from 0 to S - 1, and S is the second preset number.

4. The decoding method of the low-density parity-check code encoding and decoding multi-input multi-output system according to claim 3, characterized in that The method further includes: Initializing the first outer loop; For the current outer loop, while the detection module performs the second preset number of detection iterations, the decoding module performs the first preset number of decoding iterations in parallel; the current outer loop is an outer loop other than the first outer loop; The detection module outputs the cavity information to the decoding module as input, and the decoding module outputs the parity - check information to the detection module as input; When the preset maximum parallel number is reached, decision - making on information bits is performed according to the bit - soft information output by the decoding module.

5. A decoding device for a low-density parity-check code encoding and decoding multiple-input multiple-output system, characterized in that, Including: A detection module for performing detection based on expectation propagation in a multi - input multi - output system with low - density parity - check code encoding / decoding and multi - carrier quadrature amplitude modulation. For the current outer loop, it updates and outputs cavity information after a second preset number of detection iterations based on the sum of the parity - check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop; A decoding module for updating and outputting parity - check information after the first preset number of decoding iterations based on the cavity information output by the detection module after the second preset number of detection iterations, and using it as the input for the first detection iteration of the detection module in the next outer loop; The decoding module is specifically used for: In the (l + 1)-th decoding iteration, update and output the parity information according to the bit node information and the parity constraint The bit node information is the parity information and the input information of the decoding module The verification information is the verification information updated in the l-th decoding iteration, and the input information of the decoding module is the cavity information output by the detection module after the second preset number of detection iterations in the first decoding iteration; t is the number of iterations of the current outer loop; The detection module is specifically used for: In the (s + 1)-th detection iteration, the prior information PrDI [t][s] is obtained by moment matching based on the extrinsic information EDI and the bit-check node information [t][s] ; Based on the moment matching of the prior information PrDI [t][s] update and output the cavity information EDI [t][s+1] ; Among them, the extrinsic information EDI [t][s] is the cavity information updated in the s-th detection iteration; the bit check node information is the sum of the check information output by the decoding module after the first preset number of decoding iterations in the previous outer loop.

6. The decoding device for a low-density parity-check code encoding and decoding multiple-input multiple-output system according to claim 5, characterized in that, The decoding module is further used for: For the step of updating and outputting check information according to the bit node information and the parity check constraint, perform iteration until the value of l+1 reaches L, and output the check information as the input for the first detection iteration of the detection module in the next outer loop; ​ Among them, the input information of the decoding module In the 2nd to the Lth decoding iteration, it is the sum of the cavity information output by the detection module after the second preset number of detection iterations and the parity information updated and output after the first preset number of decoding iterations in the previous outer loop iteration; l takes an integer value from 0 to L - 1, and L is the first preset number.

7. The decoding device for a low-density parity-check code encoding and decoding multiple-input multiple-output system according to claim 5 or 6, characterized in that The detection module is further used for: Regarding the extrinsic information EDI [t][s] and the bit check node information obtain the prior information PrDI through moment matching [t][s] , and, regarding the prior information PrDI [t][s] update through moment matching and output the cavity information EDI [t][s+1] iterate the steps until the value of s + 1 reaches S, and output the cavity information EDI [t][S] , as the input for the first decoding iteration of the decoding module in the current outer loop; s takes integer values from 0 to S - 1, and S is the second preset number.

8. A receiver, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the encoding / decoding method for the low - density parity - check code encoding / decoding multi - input multi - output system according to any one of claims 1 to 4.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the encoding and decoding method of the low-density parity-check code encoding and decoding multiple-input multiple-output system according to any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the encoding and decoding method of the low-density parity-check code encoding and decoding multiple-input multiple-output system according to any one of claims 1 to 4.

Images