MIMO detection method, computer program product, electronic device and medium

Abstract

Disclosed are an MIMO detection method, a computer program product, an electronic device and a medium, which relate to the technical field of communications, and are applied to a receiver to obtain channel information, noise information and a set constellation configuration set, constellation points in the constellation configuration set being fixed and only including constellation points for message update; determining a to-be-processed signal which fails to be decoded; initializing an output symbol and first soft information according to the to-be-processed signal; updating the second soft information based on the first soft information, the to-be-processed signal, the channel information, the noise information and the constellation configuration set; updating the output symbol based on the second soft information, and updating the first soft information based on the output symbol and the second soft information; detecting whether the maximum number of updating iterations is reached or not; if not, returning to execute the step of updating the second soft information based on the first soft information, the channel information, the noise information and the constellation configuration set; and if yes, decoding the to-be-processed signal based on the output symbol. According to the invention, the processing delay is reduced, and the throughput is improved.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and more specifically, to a MIMO detection method, a computer program product, an electronic device, and a medium. Background Art

[0002] Multiple-input, multiple-output (MIMO) detection technology is a key technology in wireless communication systems. As antenna sizes increase to hundreds or even thousands, signal detection becomes increasingly complex with the development of massive MIMO systems.

[0003] Iterative detectors based on Bayesian message passing, such as expectation propagation (EP) and belief propagation (BP), have attracted attention due to their efficient balance between error rate performance and computational complexity. EP detectors approximate the posterior message to a Gaussian distribution through moment matching, achieving near-optimal error rate performance in small- and medium-scale MIMO scenarios. However, EP detectors require matrix inversion during each message iteration, significantly increasing hardware implementation complexity and power consumption. In contrast, BP detectors implement message updates based on a factor graph model (FGM), avoiding matrix inversion during iteration. This allows for better hardware implementation while maintaining near-optimal error rate performance. The belief-selective propagation (BsP) detector, a recently proposed BP detector, not only significantly reduces the computational complexity of BP detectors but also successfully addresses the error floor effect that exists in traditional BP detectors.

[0004] However, BSP detectors perform poorly in coded MIMO systems. Receivers in coded MIMO systems are generally classified into two categories: separated detection and decoding (SDD) and iterative detection and decoding (IDD). Because the information exchange between the detector and decoder effectively improves the confidence level of the information at the receiver, IDD receivers offer superior error reduction performance compared to SDD receivers. However, IDD receivers face application bottlenecks such as high overall complexity, high processing latency, and low throughput.

[0005] In summary, how to reduce the processing delay of a receiver and improve the throughput is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a MIMO detection method that can, to a certain extent, solve the technical problem of how to reduce the processing delay of a receiver and improve the throughput. This application also provides a computer program product, an electronic device, and a computer-readable storage medium.

[0007] In order to achieve the above objectives, this application provides the following technical solutions:

[0008] A MIMO detection method, applied to a receiver, comprising:

[0009] Acquiring channel information, noise information, and a set constellation configuration set, where constellation points in the constellation configuration set are fixed and only include constellation points for message update;

[0010] Determining a signal to be processed that failed decoding;

[0011] Initialize output symbols and first soft information according to the signal to be processed, where the first soft information includes soft information transmitted from the symbol node to the factor node;

[0012] updating second soft information based on the first soft information, the signal to be processed, the channel information, the noise information, and the constellation configuration set, wherein the second soft information includes soft information transmitted from a factor node to a symbol node;

[0013] updating the output symbol based on the second soft information, and updating the first soft information based on the output symbol and the second soft information;

[0014] Check whether the maximum number of update iterations has been reached;

[0015] In response to the number of update iterations being less than the maximum number, returning to the step of updating the second soft information based on the first soft information, the channel information, the noise information, and the constellation configuration set;

[0016] In response to reaching the maximum number of update iterations, the signal to be processed is decoded based on the output symbols to determine a MIMO detection result based on the decoding result.

[0017] Preferably, the updating of the second soft information based on the first soft information, the signal to be processed, the channel information, the noise information, and the constellation configuration set includes:

[0018] generating, based on the first soft information and the constellation configuration set, a priori probability that the symbol node is a constellation point in the constellation configuration set;

[0019] generating multi-user interference for constellation points within the constellation configuration set based on the prior probability;

[0020] Based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference, second soft information corresponding to the constellation configuration set is generated, and the second soft information corresponding to the constellation points other than the constellation configuration set is updated to zero.

[0021] Preferably, the generating, based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference, the second soft information corresponding to the constellation configuration set, and updating the second soft information corresponding to the constellation points other than the constellation configuration set to zero, includes:

[0022] generating second soft information corresponding to the constellation configuration set based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference using a second soft information generation formula, and updating the second soft information corresponding to the constellation points other than the constellation configuration set to zero;

[0023] The second soft information generation formula includes:

[0024] ;

[0025] ;

[0026] ;

[0027] ; ;

[0028] ;

[0029] in, represents the second soft information; j represents the number of the symbol node; i represents the number of the factor node; represents the second soft information transmitted from the i-th factor node to the j-th symbol node; k represents a number; Indicates a constellation point; represents the constellation configuration set corresponding to the j-th symbol node; represents the number of constellation points in the constellation configuration set; Indicates the number of the constellation point, represents the constellation point with the kth highest confidence in the constellation configuration set; Indicates a signal to be processed. represents the signal to be processed corresponding to the i-th factor node; represents the transmission signal of the jth symbol node; n iIndicates that the received symbol corresponding to the i-th factor node receives noise; h represents the channel information, represents the channel information between the i-th factor node and the j-th symbol node; represents the channel information between the i-th factor node and the g-th symbol node; represents the constellation configuration set corresponding to the g-th symbol node; Indicates sending symbol constellation points The prior probability of represents the noise signal; Indicates the multi-user interference received by the signal to be processed; Indicates the number of transmitting antennas; represents the natural exponential function; represents the first soft information transmitted from the g-th symbol node to the i-th factor node; represents the constellation point multi-user interference between the i-th factor node and the j-th symbol node.

[0030] Preferably, decoding the signal to be processed based on the output symbol includes:

[0031] Get the probability reduction factor;

[0032] Compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor to obtain a first compensated symbol;

[0033] Taking the output symbol corresponding to the last constellation point in the constellation configuration set as the basic symbol;

[0034] Compensating output symbols corresponding to constellation points other than the constellation configuration set based on the basic symbol and the probability decreasing factor to obtain a second compensated symbol;

[0035] The signal to be processed is decoded based on the first compensation symbol and the second compensation signal.

[0036] Preferably, compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor to obtain a first compensated symbol includes:

[0037] Compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor using a first compensation formula to obtain a first compensated symbol;

[0038] The first compensation formula includes:

[0039] ;

[0040] The compensating output symbols corresponding to constellation points other than the constellation configuration set based on the basic symbol and the probability decreasing factor to obtain a second compensated symbol includes:

[0041] Compensating output symbols corresponding to constellation points other than the constellation configuration set based on the base symbol and the probability decreasing factor using a first compensation formula to obtain a second compensated symbol;

[0042] The second compensation formula includes:

[0043] ; ;

[0044] in, represents the probability decreasing factor; The transmitted symbol of the jth symbol node corresponds to The output symbol of each constellation point; express a corresponding first compensation signal; A real number representing the set of complex field constellation points; Represents logarithmic operation; express A corresponding second compensation signal.

[0045] Preferably, before determining the signal to be processed that fails in decoding, the method further includes:

[0046] Get the received signal;

[0047] Preprocessing the received signal to obtain a signal estimation result;

[0048] Converting the signal estimation result into bit symbols;

[0049] Deinterleaving the bit symbols to obtain deinterleaved symbols;

[0050] Decoding the deinterleaved symbols as prior information;

[0051] If the decoding is successful, outputting a bit vector of the received signal based on the decoding result;

[0052] If the decoding fails, the step of determining the signal to be processed that failed in decoding is performed.

[0053] Preferably, converting the signal estimation result into a bit symbol includes:

[0054] Converting the signal estimation result into a bit symbol through a conversion formula;

[0055] The conversion formula includes:

[0056] ;

[0057] in, represents the bit symbol of the bth bit in the transmitted symbol estimated by the jth symbol node; Approximate methods for representing logarithms and addition; represents the constellation set of constellation points whose b-th bit is 1; represents the constellation set of constellation points whose b-th bit is 0; represents the signal estimation result, Represents the signal estimation result corresponding to the transmitted symbol of the j-th symbol node; represents the noise of the signal estimation result, Represents the signal estimation result noise; Represents a constellation set The kth constellation point in .

[0058] A computer program product comprises a computer program / instruction, which implements the steps of any of the above MIMO detection methods when executed by a processor.

[0059] An electronic device, comprising:

[0060] memory for storing computer programs;

[0061] A processor is configured to implement the steps of any of the above MIMO detection methods when executing the computer program.

[0062] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any of the above-mentioned MIMO detection methods.

[0063] The present application provides a MIMO detection method, which is applied to a receiver, and obtains channel information, noise information, and a set constellation configuration set, wherein the constellation points in the constellation configuration set are fixed and only include constellation points for message updates; determines a signal to be processed that fails to be decoded; initializes an output symbol and first soft information based on the signal to be processed, wherein the first soft information includes soft information transmitted from a symbol node to a factor node; updates second soft information based on the first soft information, the signal to be processed, channel information, noise information, and the constellation configuration set, wherein the second soft information includes soft information transmitted from a factor node to a symbol node; updates the output symbol based on the second soft information, and updates the first soft information based on the output symbol and the second soft information; detects whether a maximum number of update iterations has been reached; in response to a number less than the maximum number of update iterations, returns to the step of updating the second soft information based on the first soft information, channel information, noise information, and the constellation configuration set; in response to a maximum number of update iterations being reached, decodes the signal to be processed based on the output symbol, and determines a MIMO detection result based on the decoding result. In this application, a receiver updates the output symbols of a signal to be processed based on a set constellation configuration set. Because the constellation points in the constellation configuration set are fixed and only include constellation points that undergo message updates, this application updates the output symbols based on a limited number of constellation points that undergo message updates only. Compared to the BsP method, this method reduces the number of constellation points in the constellation configuration set and eliminates the impact of constellation points that do not require message updates on the output symbol updates. Furthermore, the constellation configuration set remains unchanged during the iterative update of the output symbols, resulting in a low-complexity method. Furthermore, the output symbols of the signal to be processed are updated only for signals that fail decoding, further reducing computational complexity, thereby reducing receiver processing latency and improving throughput. The computer program product, electronic device, and computer-readable storage medium provided in this application also address corresponding technical issues. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0065] Figure 1 is a schematic diagram of a MIMO communication system;

[0066] Figure 2 Schematic diagram of the BsP detection method;

[0067] Figure 3 A flowchart of a MIMO detection method provided in an embodiment of the present application;

[0068] Figure 4This is a working diagram of the IDD receiver in this application;

[0069] Figure 5 Schematic diagram of the CFBsP detection method based on the IDD architecture;

[0070] Figure 6 This is an uplink scenario diagram for a MIMO system;

[0071] Figure 7 This is a simulation diagram of an uncoded MIMO system in a medium-scale Rayleigh channel environment;

[0072] Figure 8 This is a simulation diagram of an uncoded MIMO system in a very large-scale Rayleigh channel environment;

[0073] Figure 9 This is a simulation diagram of a coded MIMO system in a Rayleigh channel environment;

[0074] Figure 10 This is a simulation diagram of an uncoded MIMO system in a 3GPP channel environment;

[0075] Figure 11 This is a simulation diagram of the coded MIMO system in the GPP channel environment;

[0076] Figure 12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0077] Figure 13 This is another structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0078] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0079] In order to facilitate understanding of the MIMO detection solution provided in this application, the IDD receiver is first described.

[0080] like Figure 1 As shown, assuming that transmit antennas and An uplink MIMO communication system with multiple receiving antennas. At the transmitting end, the length is The information bit vector First input into the LDPC encoder to generate a length of The codeword c, the encoding rate is Then, the codeword goes through the interleaver After interleaving, the input modulation order is M. During the modulation process, every M bits are divided into a group and mapped to a set of complex domain constellation points through quadrature amplitude modulation (QAM). On the other hand, the number of constellation points in the set is The symbols are divided into G groups, each consisting of For the sake of simplicity and generality, we only focus on the channel transmission of a set of modulation symbols, namely , at the receiving end, the received symbol vector It can be expressed as:

[0081] (1);

[0082] in is the channel matrix, is complex additive white Gaussian noise (AWGN), Each element in has a mean of 0 and a variance of Independent Gaussian distribution. Further simplifying formula (1), the complex signal transmission model can be converted into an equivalent real signal transmission model:

[0083] (2);

[0084] in, , , , and Represent the operations of taking the real part and imaginary part respectively, yes The real version of , whose set size is The real channel matrix 𝑯 is obtained by the following mapping:

[0085] (3);

[0086] Assuming that the channel state information (CSI) is fully known at the receiver, in an IDD receiver, the received data is first input into a preprocessor for initialization. The preprocessor typically uses a LMMSE detector to preprocess the received data and generate an initial estimate of the transmitted symbol. The specific calculation formula is:

[0087] (4);

[0088] in Is the size of Then, through the information exchange between the detector and the decoder, the IDD receiver can gradually improve the confidence of the log-likelihood ratios (LLRs) of the codeword bits. In the detection module, the output symbol LLR is expressed as , which describes the jth transmitted symbol corresponding to the constellation set The kth constellation point in To implement the decoding process, the symbol LLR needs to be converted to the bit LLR using the following formula:

[0089] (5);

[0090] in, represents the bit LLR corresponding to the bth bit in the jth transmitted symbol, and denote the constellation set of constellation points whose b-th bit is 1 and 0 respectively, Indicates a constellation point The output symbol is obtained; and the bit LLR converted by formula (5) needs to be detected before decoding. In order to reduce the computational complexity, formula (5) usually uses the approximate method of logarithmic sum addition (LogSumExP, LSE) ( ) is simplified, and the simplified results are as follows:

[0091] (6);

[0092] Subsequently, Bit LLR is fed into the deinterleaver The deinterleaver deinterleaves the bit LLR to generate the bit LLR As a priori information input to the LDPC decoder. In the reverse processing stage, the LDPC decoder calculates and outputs the bit LLR , and then passed to the interleaver. The interleaver re-interleaves the bit LLR to obtain , the bit LLR is then remapped to symbol LLR, which is used as the prior information of the detector in the next detection process. The calculation formula is:

[0093] (7);

[0094] in represents a real-valued constellation point The b-th bit of , a real value, that is, a value of type real number.

[0095] When the LDPC decoder output bits meet the early stopping criterion or the IDD receiver reaches its maximum number of iterations When , the IDD receiver will stop iterating and output the final bit vector result .

[0096] Based on the above-mentioned IDD receiver, the MIMO detection scheme proposed in this invention is based on the BSP detection method. To facilitate understanding, the algorithm principle of the BSP detection method is first introduced. The BSP detection method models the MIMO system through FGM and uses factor graph theory to estimate the marginal posterior probability of each transmitted symbol. The factor graph model contains two types of nodes: factor nodes (FNs) and symbol nodes (SNs), which are respectively used and To express. Figure 2 The BsP detection method based on FGM in the real field is presented. Transfer to The second soft information (symbol LLR) is used Indicates that from Transfer to The first soft information (symbol LLR) is used In order to achieve iterative update of soft information with lower complexity, the BsP detection method proposes two effective optimization strategies: symbol-based truncation (ST) and edge-based simplification (ES). The core idea of ​​the ST strategy is to truncate the low reliability of the sorted Message, to simplify The computational complexity of the message. During message calculation, only the largest Symbol LLRs ( message) and its corresponding constellation points, thus significantly reducing the complexity. The ES strategy further reduces the computational complexity by retaining only the largest symbol LLR ( Message) and its corresponding constellation points are used to calculate the message. In the factor graph model, each factor node is connected to edges, any of which The ST strategy is applied to the edges (called selected edges), and the rest The ES strategy is applied to each edge (called simplified edge). With these two strategies, the BsP detection method defines the configuration set , whose expression is:

[0097] (8);

[0098] in, Indicates except The sending symbol vector other than . Subvector The corresponding sent symbol sub-vector composed of the selected edges, To simplify the edge composition, the sub-vector of the sent symbols is used. Based on the configuration set , soft information The calculation formula is:

[0099] (9);

[0100] in, Indicates from arrive The corresponding Soft information; express The transmitted symbol vector at time ; express It should be noted that t and j in formula (9) and the following formulas represent the numbers of symbol nodes and have the same meaning. According to the above description, the message update steps of the BsP detection method are as follows:

[0101] The BsP detector is initialized using the output of the preprocessor Message, the calculation formula is:

[0102] (10);

[0103] in, Output results for the preprocessor The jth element of and its corresponding variance The calculation method is . Further, using Message initialization Message, that is ;

[0104] FNs to SNs Message passing: Calculated using formula (9) information;

[0105] SNs to FNs Messaging: The message update rules are: (11);

[0106] Information output: When the number of iterations reaches the maximum value When , the BsP detector outputs the symbol LLR of each transmitted symbol, specifically: (12).

[0107] Based on the above, to facilitate understanding of the decoder of an IDD receiver, we use LDPC codes as an example in a coded MIMO system and introduce the relevant decoders used. The LDPC code decoder uses the offset min-sum (OMS) decoding algorithm. The OMS decoding algorithm is widely used for its good balance between computational complexity and decoding performance. Its core idea is to use the received prior information ( ), the iterative message passing mechanism between variable nodes (VNs) and check nodes (CNs) updates the bit LLR and finally outputs the decoding result. For the convenience of description, we define , where the kth element in the bit LLR vector represents the LLR of the b-th bit in the j-th symbol ( ),then, Used for VNs initialization. Definition Indicates that from the jth VN ( ) is passed to the i-th CN ( ) news, Indicates from Pass to The message update strategy for LDPC code decoding is as follows:

[0108] (13);

[0109] in, Represents Connect but not include VN collection, Represents the offset factor. When the maximum number of decoding iterations is reached, the LDPC decoder outputs bit LLR as external information, which is expressed as follows:

[0110] (14);

[0111] in, Represents The set of all connected CNs.

[0112] When the IDD receiver meets the output conditions, the output bits are estimated using hard decision:

[0113] (15);

[0114] Otherwise, the LDPC decoder will output the bit LLR Remap to , and returns it to the detector according to formula (7) for the next receiver iteration update.

[0115] Although IDD receivers offer excellent bit error performance, their high latency limits their practical application in MIMO systems. Directly incorporating BsP detection into IDD receivers significantly increases the overall complexity of the receiver and also results in poor bit error performance. To address these issues, this application proposes a CFBsP (constellation-fixed BsP) detection method and a CFBsP IDD receiver based on this method, thereby improving the system's bit error performance and meeting practical application requirements.

[0116] See also Figure 3 , Figure 3 A flowchart of a MIMO detection method provided in an embodiment of the present application.

[0117] An embodiment of the present application provides a MIMO detection method, which is applied to a receiver and may include the following steps:

[0118] Step S101: Acquire channel information, noise information, and a set constellation configuration set. The constellation points in the constellation configuration set are fixed and only include constellation points for message update.

[0119] In practical applications, the channel information and noise information between the transmitter and receiver in the MIMO system can be obtained first, and a pre-set constellation configuration set is required. This is then used to detect the signal based on the channel information, noise information, and constellation configuration set, and to generate output symbols for decoding the signal. The receiver in this application can be an SDD receiver or an IDD receiver. To facilitate the description of the solution, the IDD receiver will be used as an example.

[0120] It should be noted that in the BsP detection method, the configuration set The parameter selection and set update of have a great impact on the computational complexity. Specifically, the largest computational overhead in the BsP detection method comes from the message update of FNs. Each FN needs to calculate according to its configuration set according to formula (9) Optimizing the configuration set and message update strategy can reduce the complexity of the BSP detection method. Therefore, this application proposes a constellation fixed configuration set.

[0121] It should also be noted that in a coded MIMO system, the constellation-fixed configuration set can be determined based on the a priori information returned by the decoder in the IDD receiver, expressed as:

[0122] (16);

[0123] in, represents a constellation fixed configuration set, To correspond to Configuration set ( ), represents the transmitted signal corresponding to the u-th receiving antenna, is the symbol LLRs generated by formula (7) Selected the most reliable The most reliable constellation point can have the highest transmission accuracy and the fastest transmission efficiency, that is, the largest value is selected. The constellation point used is the most reliable constellation point. In the non-coded MIMO system, the constellation fixed configuration set can be determined based on the estimation result provided by the preprocessor. And in terms of message update strategy, the configuration set in the CFBsP detection method is different from the BsP detection method. Only the constellation points that need message updates are included. The included constellation points do not participate in the calculation and update of the message. In addition, the configuration set It is determined before the CFBsP detection iteration and remains unchanged during the entire detection process. Therefore, compared with the BsP detector, the configuration set Message updating can effectively reduce the complexity of the CFBsP detection method.

[0124] Step S102: Determine the signal to be processed that has failed decoding.

[0125] In practical applications, if the decoding is successful, the IDD receiver can directly output it. If the decoding fails, it can be detected by the detector and then decoded again. Therefore, the IDD receiver in this application can only detect the signal to be processed that failed to be decoded. In this way, the working process of the IDD receiver is as follows: According to the received signal, channel information and noise information, the received signal vector is obtained. , channel matrix , and the noise variance The preprocessor uses the received data to obtain the preprocessing result, and directly inputs it as a priori information to the decoder for decoding; if the decoder result is wrong, the output LLR message Feedback is sent to the CFBsP detector to activate the receiver's IDD mechanism. If the decoder's decoding result is correct, the decoding result is directly output. That is, before determining the signal to be processed that failed decoding, the received signal can also be obtained; the received signal is preprocessed to obtain a signal estimation result; the signal estimation result is converted into bit symbols; the bit symbols are deinterleaved to obtain deinterleaved symbols; the deinterleaved symbols are used as prior information for decoding; if decoding is successful, the bit vector of the received signal is output based on the decoding result; if decoding fails, the steps of determining the signal to be processed that failed decoding are executed.

[0126] In an exemplary embodiment, to further reduce the computational complexity of the IDD receiver and fully utilize the preprocessing results provided by the preprocessor, a decode-first IDD receiver architecture is proposed based on the CFBsP detector. In this IDD receiver, the estimation results output by the preprocessor are directly converted into bit LLRs. That is, in the process of converting the signal estimation results into bit symbols, the signal estimation results can be converted into bit symbols through a conversion formula.

[0127] The conversion formula includes:

[0128] ;

[0129] in, represents the bit symbol of the bth bit in the transmitted symbol estimated by the jth symbol node; Approximate methods for representing logarithms and addition; represents the constellation set of constellation points whose b-th bit is 1; represents the constellation set of constellation points whose b-th bit is 0; represents the signal estimation result, represents a signal estimation result corresponding to the transmitted symbol of the j-th symbol node; represents the noise of the signal estimation result, Represents the signal estimation result noise; Represents a constellation set The bit symbol obtained by this conversion formula can be directly decoded without being processed by the detector. It will be directly transmitted to the decoder as a priori information and bypass the CFBsP detector. If the decoder result is correct, the IDD receiver will directly output the estimated bit vector Otherwise, the IDD receiver will transmit the decoding result to the CFBsP detector, starting the information exchange between the detector and the decoder, such as Figure 4 shown.

[0130] Step S103: Initialize output symbols and first soft information according to the signal to be processed, where the first soft information includes soft information transmitted from the symbol node to the factor node.

[0131] In practical applications, in the process of detecting a signal to be processed that fails in decoding, the output symbol and the first soft information can be initialized according to the signal to be processed, and the first soft information includes soft information transmitted from the symbol node to the factor node, for example, Initialize the output symbols by Initialize the first soft information, etc.

[0132] Step S104: updating second soft information based on the first soft information, the signal to be processed, the channel information, the noise information and the constellation configuration set, where the second soft information includes soft information transmitted from the factor node to the symbol node.

[0133] Step S105: updating the output symbol based on the second soft information, and updating the first soft information based on the output symbol and the second soft information.

[0134] In practical applications, after initializing the output symbol and the first soft information according to the signal to be processed, the second soft information can be updated based on the first soft information, the signal to be processed, the channel information, the noise information and the constellation configuration set, and the second soft information includes the soft information transmitted from the factor node to the symbol node; the output symbol is updated based on the second soft information, and the first soft information is updated based on the output symbol and the second soft information.

[0135] In the exemplary embodiment, although the above-mentioned fixed constellation point configuration set can significantly reduce the complexity of message updates, the use of this configuration set may lead to a decrease in detection performance due to the truncation of constellation points. Therefore, it is necessary to improve the message update accuracy of the constellation points in the configuration set to reduce the potential performance loss. To this end, a multi-user interference approximation strategy is proposed to update Specifically, the received signal corresponding to the i-th factor node is It can be expressed as:

[0136] (17);

[0137] in, Indicates the sending symbol outside, In fact, the more accurate the multi-user interference calculation is, the better the detection performance is. However, in theory, the most accurate multi-user interference calculation requires searching through all transmitted symbols, which will lead to high computational complexity. To solve this problem, only the configuration set The constellation points within approximate multi-user interference. Specifically, use To express the approximate multi-user interference, the calculation formula is:

[0138] (18);

[0139] in, Is the sending symbol constellation points The prior probability of is calculated as follows:

[0140] (19);

[0141] To further reduce To improve the computational complexity of the message, a unilateral update strategy can also be used. The incoming prior information is ignored when the message is received, and after introducing approximate multi-user interference, The message update formula can be simplified as follows:

[0142] (20);

[0143] in .therefore, News updates are limited to The constellation symbol in . Other messages that are not updated remain zero. Eventually, The message can be represented as:

[0144] (twenty one).

[0145] Therefore, in the process of updating the second soft information based on the first soft information, the signal to be processed, the channel information, the noise information, and the constellation configuration set, a priori probability that the symbol node is a constellation point in the constellation configuration set can be generated based on the first soft information and the constellation configuration set; multi-user interference of the constellation points in the constellation configuration set can be generated based on the priori probability; second soft information corresponding to the constellation configuration set can be generated based on the signal to be processed, the channel information, the noise information, and the multi-user interference of the constellation points, and the second soft information corresponding to the constellation points outside the constellation configuration set can be updated to zero. Furthermore, in the process of generating the second soft information corresponding to the constellation configuration set based on the signal to be processed, the channel information, the noise information, and the multi-user interference of the constellation points, and updating the second soft information corresponding to the constellation points outside the constellation configuration set to zero, the second soft information corresponding to the constellation configuration set can be generated based on the signal to be processed, the channel information, the noise information, and the multi-user interference of the constellation points using a second soft information generation formula, and the second soft information corresponding to the constellation points outside the constellation configuration set can be updated to zero.

[0146] The second soft information generation formula includes:

[0147] ;

[0148] ;

[0149] ;

[0150] ; ;

[0151] ;

[0152] in, represents the second soft information; j represents the number of the symbol node; i represents the number of the factor node; represents the second soft information transmitted from the i-th factor node to the j-th symbol node; k represents a number; Indicates a constellation point; represents the constellation configuration set corresponding to the j-th symbol node; represents the number of constellation points in the constellation configuration set; Indicates the number of the constellation point, represents the constellation point with the kth highest confidence in the constellation configuration set; Indicates a signal to be processed. represents the signal to be processed corresponding to the i-th factor node; represents the transmission signal of the jth symbol node; n i Indicates that the received symbol corresponding to the i-th factor node receives noise; h represents the channel information, represents the channel information between the i-th factor node and the j-th symbol node; represents the channel information between the i-th factor node and the g-th symbol node; represents the constellation configuration set corresponding to the g-th symbol node; Indicates sending symbol constellation points The prior probability of represents the noise signal; Indicates the multi-user interference received by the signal to be processed; Indicates the number of transmitting antennas; represents the natural exponential function; represents the first soft information transmitted from the g-th symbol node to the i-th factor node; represents the constellation point multi-user interference between the i-th factor node and the j-th symbol node.

[0153] Step S106: Detect whether the maximum number of update iterations is reached; in response to being less than the maximum number of update iterations, return to executing the step of updating the second soft information based on the first soft information, channel information, noise information, and constellation configuration set; in response to being reached the maximum number of update iterations, execute step S107.

[0154] Step S107: Decode the signal to be processed based on the output symbol.

[0155] In practical applications, after updating the first soft information based on the output symbol and the second soft information, it is possible to detect whether the maximum number of update iterations has been reached; in response to the number being less than the maximum number of update iterations, the process returns to executing the step of updating the second soft information based on the first soft information, channel information, noise information, and constellation configuration set, so as to perform the next iterative update; in response to the maximum number of update iterations being reached, the processed signal can be decoded based on the output symbol to determine the MIMO detection result based on the decoding result, that is, a specific bit decoding result is obtained after decoding, and the result can be fed back to the detector as prior information to further increase the accuracy of the detection.

[0156] In an exemplary embodiment, although the approximate multi-user interference strategy can reduce the detector performance loss caused by symbol truncation to a certain extent, due to the lack of certain symbol information, information loss is inevitable during information exchange between the detector and the decoder, thereby reducing the error performance of the IDD receiver. To solve this problem, the present application proposes an information compensation strategy to reduce performance loss by compensating for the missing information of symbols outside the configuration set. Specifically, when the CFBsP detection method reaches the maximum number of iterations, the symbol LLR of each constellation point in the configuration set will be output, and the expression is as follows:

[0157] (twenty two);

[0158] in , this part of the symbol LLRs needs to be compensated. Specifically, we can first reasonably assume that the last The probability of a constellation point decreases as its reliability decreases. Secondly, a probability reduction factor is introduced To control the extent of probability reduction, specifically expressed as:

[0159] (twenty three);

[0160] in Finally, we can assume that the sum of the probabilities of all constellation points is 1. Based on the above conditions, we can deduce that:

[0161] (twenty four);

[0162] in , Indicates the constellation symbol assuming no truncation of symbols Therefore, for the configuration set The CFBsP detection method can be the most reliable symbols are treated as untruncated symbols and the The LLR information of the symbols is used to compensate for the remaining symbol.

[0163] In other words, in the process of decoding the signal to be processed based on the output symbol, a probability decreasing factor can be obtained; the output symbol corresponding to the constellation configuration set is compensated based on the probability decreasing factor to obtain a first compensation symbol; the output symbol corresponding to the last constellation point in the constellation configuration set is used as the base symbol; based on the base symbol and the probability decreasing factor, the output symbols corresponding to constellation points other than the constellation configuration set are compensated to obtain a second compensation symbol; and the signal to be processed is decoded based on the first compensation signal and the second compensation symbol. Furthermore, in the process of compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor to obtain the compensation symbol, the output symbol corresponding to the constellation configuration set can be compensated based on the probability decreasing factor using a first compensation formula to obtain the first compensation symbol; the first compensation formula includes:

[0164] ;

[0165] Based on the basic symbol and the probability reduction factor, compensating the output symbols corresponding to the constellation points other than the constellation configuration set to obtain a second compensated symbol includes:

[0166] Compensating output symbols corresponding to constellation points other than the constellation configuration set based on the basic symbol and the probability decreasing factor using the first compensation formula to obtain a second compensated symbol;

[0167] The second compensation formula includes:

[0168] ; ;

[0169] in, represents the probability decreasing factor; The transmitted symbol of the jth symbol node corresponds to The output symbol of each constellation point; express a corresponding first compensation signal; A real number representing the set of complex field constellation points; Represents logarithmic operation; express The corresponding second compensation signal. In this way, the process of the CFBsP detection method based on the IDD architecture in this application can be as follows Figure 5 shown.

[0170] The present application provides a MIMO detection method, which is applied to a receiver, to obtain channel information, noise information and a set constellation configuration set, wherein the constellation points in the constellation configuration set are fixed and only include constellation points for message update; to determine a signal to be processed that has failed decoding; to initialize an output symbol and first soft information based on the signal to be processed, wherein the first soft information includes soft information transmitted from a symbol node to a factor node; to update second soft information based on the first soft information, the signal to be processed, channel information, noise information and the constellation configuration set, wherein the second soft information includes soft information transmitted from a factor node to a symbol node; to update the output symbol based on the second soft information, and to update the first soft information based on the output symbol and the second soft information; to detect whether a maximum number of update iterations has been reached; in response to a number less than the maximum number of update iterations, returning to the step of updating the second soft information based on the first soft information, channel information, noise information and the constellation configuration set; and in response to reaching the maximum number of update iterations, decoding the signal to be processed based on the output symbol. In the present application, the receiver updates the output symbols of the signal to be processed based on a set constellation configuration set. Since the constellation points in the constellation configuration set are fixed and only include constellation points for message updates, the present application updates the output symbols based on a limited number of constellation points that are only updated with messages. Compared with the BsP method, the number of constellation points in the constellation configuration set is reduced and the impact of constellation points that do not require message updates on the update of output symbols is eliminated. The constellation configuration set does not change during the iterative update process of the output symbols, and the method complexity is low. In addition, the output symbols of the signal to be processed are only updated for decoding failures, which further reduces the computational complexity, thereby reducing the processing delay of the receiver and improving the throughput.

[0171] To demonstrate the performance advantages of this application's CFBsP detection method and the IDD receiver based on it, simulation scenarios were considered for Rayleigh channels and actual 3GPP channels in uncoded and LDPC-coded MIMO systems. The channel data for the actual 3GPP channel was captured from an actual 6G scalable distributed massive MIMO (CF-mMIMO) system. Figure 6 An uplink scenario of the MIMO system is shown, which includes three user equipment (UEs) supporting four-stream data and eight distributed remote radio units (RRUs) on the base station side, each equipped with four receive antennas.

[0172] In the uncoded MIMO system, it is compared with LMMSE, RD-GAI-BP, BsP and channel hardening-exploiting message passing (CHEMP) detectors. At the same time, the performance of the EP detector is also presented to provide near-optimal performance results. In the coded MIMO scenario, the performance of the BsP SDD / IDD receiver, the CHEMP SDD / IDD receiver and the CFBsP SDD / IDD receiver are compared. The specific parameter configuration is that the damping coefficient of the CHEMP and RD-GAI-BP detectors is set to δ = 0.33; the configuration set of the BsP detector is ; The configuration set used by the CFBsP detector is (uncoded MIMO system) and (Coded MIMO system), the reduction factor is set to θ = 0.8.

[0173] Based on the channel environment and whether channel coding is added, the simulated MIMO systems can be divided into the following four categories:

[0174] (1) Uncoded MIMO system in Rayleigh channel environment

[0175] In this simulation scenario, a medium-scale 8×4 (8 receive antennas and 4 transmit antennas), {16,64,256}-QAM MIMO system ( Figure 7 ) and larger scale (16 receiving antennas and 8 transmitting antennas), (32 receiving antennas and 12 transmitting antennas), (256 receiving antennas and 64 transmitting antennas) MIMO system ( Figure 8 ). From the simulation results, in the 8×4 MIMO scenario, BER= When the signal-to-noise ratio (SNR) is 0.000, the CFBsP detection method achieves approximately 1.5dB and 0.7dB performance gains over LMMSE and BsP, respectively. In massive MIMO scenarios, while message passing algorithms all exhibit near-optimal error performance, CFBsP achieves a 0.5dB performance gain over BsP at lower signal-to-noise ratios. Compared to LMMSE, CFBsP achieves a performance gain of over 2.0dB.

[0176] (2) Coded MIMO system in Rayleigh channel environment

[0177] In this simulation scenario, a 32×12, 64-QAML DPC coded MIMO scenario with a code length of 1440 and code rates of 0.8 and 0.5 is considered. Figure 9 As shown. Figure 9 In the high bit rate scenario shown in (a), the CFBsP SDD receiver has a performance gain of 0.9dB and 0.3dB compared to the BsP SDD receiver and the CHEMP SDD receiver. The CFBsP IDD receiver has a performance gain of 0.7dB and 0.3dB compared to the BsP IDD receiver and the CHEMP IDD receiver. Figure 9 (b) shows that in the medium bit rate scenario, the CFBsP IDD receiver has a performance gain of 0.4dB and 0.6dB compared to the BsP IDD receiver and CHEMP IDD receiver.

[0178] (3) Uncoded MIMO system in 3GPP channel environment

[0179] In this simulation scenario, a 32×12, {16,64,256}-QAM MIMO system is considered ( Figure 10 The CFBsP detector has performance advantages of 2.2dB and 0.5dB over the LMMSE and BsP detection methods, respectively. Compared to RD-GAI-BP, the CFBsP detector has a performance advantage of 0.8dB in 256-QAM modulation scenarios.

[0180] (4) Coded MIMO system in 3GPP channel environment

[0181] In this simulation scenario, a 32×12, 64-QAML DPC coded MIMO scenario with a code length of 1440 and code rates of 0.8 and 0.5 is considered. Figure 11 As shown. Figure 11 In the high bit rate scenario shown in (a), the CFBsP SDD receiver has a performance gain of 2.5dB and 1.5dB compared to the LMMSE SDD receiver and the BsP SDD receiver, while the CFBsP IDD receiver has a performance advantage of 1.8dB compared to the BsP IDD receiver. Figure 11 (b) shows that in the medium bit rate scenario, the CFBsP SDD receiver has a performance gain of 2.2dB and 1.9dB compared to the BsP SDD receiver and CHEMP SDD receiver. The CFBsP IDD receiver has a performance gain of 1.9dB and 0.8dB compared to the BsP IDD receiver and CHEMP IDD receiver.

[0182] Judging from the above simulation results, the CFBsP detection method and the CFBsP-based IDD receiver proposed in this application demonstrate superior performance and robustness in multiple MIMO scenarios.

[0183] The present application also provides an electronic device and a computer-readable storage medium, both of which have the corresponding effects of the MIMO detection method provided in the embodiments of the present application. Figure 12 , Figure 12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0184] An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program, and when the processor 202 executes the computer program, the steps of the MIMO detection method described in any of the above embodiments are implemented.

[0185] See also Figure 13 Another electronic device provided in an embodiment of the present application may further include: an input port 203 connected to the processor 202 for transmitting commands inputted from the outside to the processor 202; a display unit 204 connected to the processor 202 for displaying the processing results of the processor 202 to the outside world; and a communication module 205 connected to the processor 202 for enabling communication between the electronic device and the outside world. The display unit 204 may be a display panel, a laser scanning display, etc. The communication method adopted by the communication module 205 includes but is not limited to Mobile High-Definition Link (MHL), Universal Serial Bus (USB), High-Definition Multimedia Interface (HDMI), wireless connection: Wireless Fidelity (WiFi), Bluetooth communication technology, Bluetooth low energy communication technology, and communication technology based on IEEE802.11s.

[0186] An embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the MIMO detection method described in any of the above embodiments.

[0187] An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the MIMO detection method described in any of the above embodiments are implemented.

[0188] The computer-readable storage medium involved in this application includes random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs (Compact Disc Read-Only Memory), or any other form of storage medium known in the technical field.

[0189] For the description of the relevant parts of the computer program product, electronic device, and computer-readable storage medium provided in the embodiments of the present application, please refer to the detailed description of the corresponding parts in the MIMO detection method provided in the embodiments of the present application, and no further description is given here. In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive elaboration.

[0190] It should also be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0191] The above description of the disclosed embodiments will enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A MIMO detection method, characterized in that: Applications in receivers include: Acquiring channel information, noise information, and a set constellation configuration set, where constellation points in the constellation configuration set are fixed and only include constellation points for message update; Determining a signal to be processed that failed decoding; Initialize output symbols and first soft information according to the signal to be processed, where the first soft information includes soft information transmitted from the symbol node to the factor node; updating second soft information based on the first soft information, the signal to be processed, the channel information, the noise information, and the constellation configuration set, wherein the second soft information includes soft information transmitted from a factor node to a symbol node; updating the output symbol based on the second soft information, and updating the first soft information based on the output symbol and the second soft information; Check whether the maximum number of update iterations has been reached; In response to the number of update iterations being less than the maximum number, returning to the step of updating the second soft information based on the first soft information, the channel information, the noise information, and the constellation configuration set; In response to reaching the maximum number of update iterations, the signal to be processed is decoded based on the output symbols to determine a MIMO detection result based on the decoding result.

2. The method according to claim 1, characterized in that The updating of the second soft information based on the first soft information, the signal to be processed, the channel information, the noise information, and the constellation configuration set includes: generating, based on the first soft information and the constellation configuration set, a priori probability that the symbol node is a constellation point in the constellation configuration set; generating multi-user interference for constellation points within the constellation configuration set based on the prior probability; Based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference, second soft information corresponding to the constellation configuration set is generated, and the second soft information corresponding to the constellation points other than the constellation configuration set is updated to zero.

3. The method according to claim 2, characterized in that The generating, based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference, the second soft information corresponding to the constellation configuration set, and updating the second soft information corresponding to the constellation points other than the constellation configuration set to zero, includes: generating second soft information corresponding to the constellation configuration set based on the signal to be processed, the channel information, the noise information, and the constellation point multi-user interference using a second soft information generation formula, and updating the second soft information corresponding to the constellation points other than the constellation configuration set to zero; The second soft information generation formula includes: ； ； ； ； ； ； in, represents the second soft information; j represents the number of the symbol node; i represents the number of the factor node; represents the second soft information transmitted from the i-th factor node to the j-th symbol node; k represents a number; Indicates a constellation point; represents the constellation configuration set corresponding to the j-th symbol node; represents the number of constellation points in the constellation configuration set; Indicates the number of the constellation point, represents the constellation point with the kth highest confidence in the constellation configuration set; Indicates a signal to be processed. represents the signal to be processed corresponding to the i-th factor node; represents the transmission signal of the jth symbol node; n i Indicates that the received symbol corresponding to the i-th factor node receives noise; h represents the channel information, represents the channel information between the i-th factor node and the j-th symbol node; represents the channel information between the i-th factor node and the g-th symbol node; represents the constellation configuration set corresponding to the g-th symbol node; Indicates sending symbol constellation points The prior probability of represents the noise signal; Indicates the multi-user interference received by the signal to be processed; Indicates the number of transmitting antennas; represents the natural exponential function; represents the first soft information transmitted from the g-th symbol node to the i-th factor node; represents the constellation point multi-user interference between the i-th factor node and the j-th symbol node.

4. The method according to claim 3, characterized in that The decoding of the signal to be processed based on the output symbol includes: Get the probability reduction factor; Compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor to obtain a first compensated symbol; Taking the output symbol corresponding to the last constellation point in the constellation configuration set as the basic symbol; Compensating output symbols corresponding to constellation points other than the constellation configuration set based on the basic symbol and the probability decreasing factor to obtain a second compensated symbol; The signal to be processed is decoded based on the first compensation symbol and the second compensation signal.

5. The method according to claim 4, characterized in that The compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor to obtain a first compensated symbol includes: Compensating the output symbol corresponding to the constellation configuration set based on the probability decreasing factor using a first compensation formula to obtain a first compensated symbol; The first compensation formula includes: ； The compensating output symbols corresponding to constellation points other than the constellation configuration set based on the basic symbol and the probability decreasing factor to obtain a second compensated symbol includes: Compensating output symbols corresponding to constellation points other than the constellation configuration set based on the base symbol and the probability decreasing factor using a first compensation formula to obtain a second compensated symbol; The second compensation formula includes: ； ； in, represents the probability decreasing factor; The transmitted symbol of the jth symbol node corresponds to The output symbol of each constellation point; express a corresponding first compensation signal; A real number representing the set of complex field constellation points; Represents logarithmic operation; express A corresponding second compensation signal.

6. The method according to claim 5, characterized in that Before determining the signal to be processed that fails in decoding, the method further includes: Get the received signal; Preprocessing the received signal to obtain a signal estimation result; Converting the signal estimation result into bit symbols; Deinterleaving the bit symbols to obtain deinterleaved symbols; Decoding the deinterleaved symbols as prior information; If the decoding is successful, outputting a bit vector of the received signal based on the decoding result; If the decoding fails, the step of determining the signal to be processed that failed in decoding is performed.

7. The method according to claim 6, characterized in that Converting the signal estimation result into a bit symbol includes: Converting the signal estimation result into a bit symbol through a conversion formula; The conversion formula includes: ； in, represents the bit symbol of the bth bit in the transmitted symbol estimated by the jth symbol node; Approximate methods for representing logarithms and addition; represents the constellation set of constellation points whose b-th bit is 1; represents the constellation set of constellation points whose b-th bit is 0; represents the signal estimation result, represents a signal estimation result corresponding to the transmitted symbol of the j-th symbol node; represents the noise of the signal estimation result, Represents the signal estimation result noise; Represents a constellation set The kth constellation point in .

8. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the MIMO detection method according to any one of claims 1 to 7 are implemented.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the MIMO detection method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the MIMO detection method according to any one of claims 1 to 7 are implemented.

Images